Primary aim: whether sequential bepirovirsen and Peg-IFN reduces relapse and improves response

“The B-Together study investigated if sequential bepirovirsen and pegylated interferon-α-2a (Peg-IFN) therapy can reduce relapse rates and improve response rates.” (opens the source at this quote in a new tab)

Hypothesis: conversion of partial responders and prevention of relapse

“The B-Together study was designed as a proof-of-concept study based on the hypothesis that Peg-IFN sequential therapy could increase response rates by converting bepirovirsen partial responders (those who had a reduction in HBsAg levels but were HBsAg > lower limit of quantification [LLOQ; 0.05 IU/ml] at EoT) to full responders on Peg-IFN treatment and/or by preventing relapse.” (opens the source at this quote in a new tab)

Objective as registered

“This study is intended to evaluate if 12 or 24 weeks of treatment with GSK3228836 followed by up to 24 weeks of pegylated interferon (PegIFN) can increase the rate of hepatitis B virus surface antigen (HBsAg) loss in participants on stable nucleos(t)ide analogue (NA) therapy, and whether virologic response can be sustained once PegIFN treatment is discontinued.” (opens the source at this quote in a new tab)

Questions the protocol was designed to answer

“Does sequential therapy with GSK3228836 (12 or 24 weeks) followed by up to 24 weeks of PegIFN result in improved rates of efficacy in NA-controlled CHB patients compared to GSK3228836 monotherapy?” (opens the source at this quote in a new tab)
“Is the efficacy rate affected by the duration of treatment with GSK3228836 prior to starting PegIFN?” (opens the source at this quote in a new tab)

Secondary and exploratory objectives

“Secondary objectives included efficacy of bepirovirsen and Peg-IFN sequential therapy on HBV biomarkers (HBsAg and HBeAg) and virus-specific antibody responses (anti-HBeAg status, anti-HBs), durability of virological response, and between-treatment comparison for the primary outcome.” (opens the source at this quote in a new tab)
“Exploratory objectives included generating serum proteomic profiles of enrolled participants (supplementary methods).” (opens the source at this quote in a new tab)

Objective of the post hoc biomarker analysis

“We investigated bepirovirsen’s MoA in relation to virological response, hepatocyte death, and ALT increases.” (opens the original PMC page at this quote in a new tab)View highlighted excerpt for this quote

Trial period, sites, and countries

“The trial was conducted from 28 January 2021–17 February 2023 at 52 sites in 11 countries (Table S1).” (opens the source at this quote in a new tab)

Registry record: identifiers, dates, and enrollment

Field	Quoted record
Other study identifier	“209348”
Study start (actual)	“2021-01-28”
Primary completion (actual)	“2023-02-17”
Enrollment (actual)	“108”
Number of listed locations	“This study has 49 locations”
Results first posted	“2024-05-02”

Regulatory identifiers

“EudraCT: 2020-002979-35” (opens the source at this quote in a new tab)
“IND: IND 122685” (opens the source at this quote in a new tab)

Overlap of sites with B-Clear

“73.5% of the sites from B-Clear are the same as those in B-Together” (opens the source at this quote in a new tab)

Phase IIb, multicentre, randomised, open-label design

“B-Together was a phase IIb, multicentre, randomised, open-label trial to assess the efficacy and safety of sequential treatment with bepirovirsen followed by Peg-IFN in participants with chronic HBV infection.” (opens the source at this quote in a new tab)

Randomisation, stratification, and allocation concealment

“Participants were randomised 1:1 to treatment Arm 1 or 2 (Fig. 1). Treatment arms were stratified by HBsAg level (HBsAg ≤1,000/>1,000 IU/ml) and hepatitis B e antigen (HBeAg) status (positive [+]/negative [−]). Investigators and participants were unblinded to study intervention throughout the trial. To protect against foreknowledge of pre-randomisation, allocation was concealed by central blocking.” (opens the source at this quote in a new tab)

Treatment windows and off-treatment follow-up

“Participants received weekly bepirovirsen 300 mg (plus a loading dose on Days 4 and 11) for 24 (Arm 1) or 12 (Arm 2) weeks (bepirovirsen treatment window) followed by weekly Peg-IFN 180 μg for up to 24 weeks (Peg-IFN treatment window), if eligible. Participants were followed for an additional 24 (Arm 1) or 36 (Arm 2) weeks (off-treatment) following planned end of Peg-IFN treatment. Participants continued NA treatment throughout the trial.” (opens the source at this quote in a new tab)

Oversight

“The sponsor, GSK, was involved in the design, conduct, and analysis of the trial. An Independent Data Monitoring Committee reviewed ongoing unblinded safety data.” (opens the source at this quote in a new tab)

Registered design details

Field	Quoted record
Allocation	“Randomized”
Interventional model	“Parallel Assignment”
Masking	“None (Open Label)”
Model description	“Participants will be randomised in a ratio of 1:1 to one of two treatment arms: 24 weeks GSK3228836 followed by up to 24 weeks PegIFN, or 12 weeks GSK3228836 followed by up to 24 weeks PegIFN.”

Rationale for sequential therapy

“Lower levels of baseline HBsAg consistently predict the highest response rate to PegIFN in terms of HBsAg loss [Ren, 2019].” (opens the source at this quote in a new tab)

Rationale for enrolling participants controlled on NA therapy

“Although we believe these flares to be therapeutic in nature and related to immune clearance of HBV from the liver, our intention is to manage the risk of possible exacerbations of hepatitis by first evaluating proof of concept (POC) in participants whose infection is already under control with stable NA therapy.” (opens the source at this quote in a new tab)

Key inclusion and exclusion criteria

“B-Together enrolled adult participants with confirmed chronic HBV infection (were HBsAg positive [+]) for ≥6 months before screening and currently receiving stable NA therapy (defined as no changes to their NA regimen from at least 6 months prior to screening and with no planned changes to the stable regimen over the duration of the study) except telbivudine. Key inclusion criteria were: eligible for Peg-IFN treatment with plasma or serum HBsAg >100 IU/ml (measured using Roche Elecsys HBsAg II quant II), HBV DNA <90 IU/ml, and alanine aminotransferase (ALT) ≤2x the upper limit of normal (ULN). Key exclusion criteria were: HCV, HDV, or HIV co-infection; cirrhosis; hepatocellular carcinoma; vasculitis.” (opens the source at this quote in a new tab)
“Peg-IFN eligibility was assessed at screening and bepirovirsen EoT.” (opens the source at this quote in a new tab)

Age range and registered inclusion criteria

“18 to 75 years of age at the time of signing the informed consent.” (opens the source at this quote in a new tab)
“Participants who are eligible to be treated with PegIFN.” (opens the source at this quote in a new tab)
“Plasma or serum HBsAg concentration >100 International Units per milliliter (IU/mL).” (opens the source at this quote in a new tab)
“Plasma or serum HBV DNA concentration <90 IU/mL.” (opens the source at this quote in a new tab)

Cirrhosis exclusion

“History of or suspected liver cirrhosis and/or evidence of cirrhosis as determined by both: Aspartate aminotransferase (AST)-Platelet Index (APRI) >2 and FibroSure/FibroTest result >0.7.” (opens the source at this quote in a new tab)
“Regardless of APRI of Fibrosure/FibroTest score, if the participant meets one of the following criteria, they will be excluded from the study: a) Liver biopsy (i.e., Metavir Score F4); b) Liver stiffness >12 kilopascals (kPa).” (opens the source at this quote in a new tab)

Exclusions related to interferon, psychiatric history, and prior oligonucleotides

“Pre-existing severe psychiatric condition or a history of severe psychiatric disorders, including severe depression, suicidal ideation and attempted suicide.” (opens the source at this quote in a new tab)
“Participants with prior treatment with PegINF or interferon will be excluded” (opens the source at this quote in a new tab)
“Prior treatment with any oligonucleotide or small interfering ribonucleic acid (siRNA) within 12 months prior to the first dosing day.” (opens the source at this quote in a new tab)

Laboratory exclusions

“Laboratory results as follows: Serum albumin <3.5 grams per deciliter (g/dL); Glomerular filtration rate (GFR) <60 milliliter per minute per 1.73 square meter (mL/min /1.73 m^2) as calculated by the Chronic Kidney Disease Epidemiologic Collaboration (CKD-EPI) formula (for Japan, Japanese Society of Nephrology Chronic Kidney Disease Initiative [JSN-CKDI equation]); International normalized ratio (INR) >1.25; Platelet count <140x10^9 cells/L; Baseline hemoglobin <10 g/dL; Total bilirubin >1.25 times ULN” (opens the source at this quote in a new tab)

Bepirovirsen dose and duration followed by Peg-IFN

“In this phase IIb, multicentre, open-label trial, participants on stable nucleos(t)ide analogue (NA) therapy were randomised 1:1 to bepirovirsen 300 mg once weekly (plus loading dose on Days 4 and 11) for 24 (Arm 1) or 12 (Arm 2) weeks followed by Peg-IFN 180 μg once weekly for up to 24 weeks, with up to 36 weeks follow-up. Participants continued NA therapy throughout.” (opens the source at this quote in a new tab)

Start of Peg-IFN after bepirovirsen

“Eligible participants began Peg-IFN treatment 1 week after the last dose of bepirovirsen; those not immediately able to start Peg-IFN were re-assessed on Weeks 2, 3, 4, 6, 8, 10, and 13 of the Peg-IFN eligibility window. Participants who failed to meet Peg-IFN eligibility criteria did not start Peg-IFN treatment.” (opens the source at this quote in a new tab)

Registered arms

Arm	Quoted description
Arm 1	“Participants on stable NA therapy received 300 milligrams per week (mg/week) GSK3228836 for 24 weeks (plus a loading dose on Day 4 and 11), followed by Pegylated Interferon (PegIFN) 180 microgram per week (mcg/week) up to 24 weeks.”
Arm 2	“Participants on stable NA therapy received 300mg/week GSK3228836 for 12 weeks (plus a loading dose on Day 4 and 11), followed by PegIFN 180 mcg/week up to 24 weeks.”
Background therapy	“Participants will continue to receive their NA therapy for the duration of the study.”

Rationale for 24 weeks of Peg-IFN rather than 48 weeks

“Current PegIFN treatment guidelines recommend 48 weeks treatment. In this study PegIFN will be administered for a maximum of 24 weeks duration.” (opens the source at this quote in a new tab)
“PegIFN administration is associated with numerous adverse effects and can be difficult to tolerate.” (opens the source at this quote in a new tab)

Rationale for 12 or 24 weeks of bepirovirsen

“In the proposed study, we will investigate 2 different durations of treatment with GSK3228836 (12 and 24 weeks at 300 mg/week) prior to starting PegIFN in order to optimise rates of sustained seroclearance.” (opens the source at this quote in a new tab)

Primary outcome definition

“The primary efficacy outcome was the percentage of participants in each treatment arm who achieved HBsAg and HBV DNA levels below the LLOQ (HBsAg: 0.05 IU/ml, HBV DNA: 20 IU/ml) at every visit for 24 weeks after the planned end of sequential treatment, in the absence of newly initiated antiviral rescue therapy.” (opens the source at this quote in a new tab)

Post hoc alternative primary outcome and definition of relapse

“Post hoc analysis evaluated an alternative primary outcome definition assessing the proportion of participants in each arm with HBsAg and HBV DNA <LLOQ at 24 weeks post-EoT. Relapse was defined as any value of HBsAg or HBV DNA ≥LLOQ or newly initiated antiviral rescue therapy.” (opens the source at this quote in a new tab)

Functional cure as the endpoint for new HBV therapies

“Prognosis may be improved by achieving functional cure,2 defined as sustained loss of hepatitis B surface antigen (HBsAg) and unquantifiable HBV DNA 24 weeks following treatment cessation.” (opens the source at this quote in a new tab)
“Functional cure is uncommon (NA only: <3%, Peg-IFN only: <7%) with current treatments.” (opens the source at this quote in a new tab)

Protocol rationale for sustained virologic response

“The rationale for the proposed study is to evaluate the potential of up to 48 weeks of sequential therapy with GSK3228836 and PegIFN to achieve sustained loss of HBsAg and HBV DNA in participants on stable NA therapy. Sustained virologic response is defined as undetectable levels of HBsAg and HBV DNA on treatment and maintaining this for a minimum of 24 weeks off-treatment (GSK3228836 and PegIFN).” (opens the source at this quote in a new tab)
“Recently, a functional cure of CHB infection has been endorsed as the endpoint for new HBV therapies [Kim, 2018].” (opens the source at this quote in a new tab)

Registered primary outcome measure

“Sustained virologic response is defined as undetectable levels of Hepatitis B surface antigen (HBsAg) and Hepatitis-B virus deoxy-ribonucleic acid (HBV DNA) on treatment. The SVR was a composite endpoint defined as HBsAg and HBV DNA levels were less than (<) Lower limit of quantitation (LLOQ) at the planned end of sequential treatment of GSK3228836 and PegIFN treatment which is sustained for 24 weeks post-GSK3228836 and PegIFN treatment in the absence of any rescue medication.” (opens the source at this quote in a new tab)

Safety outcomes and adverse events of special interest

“AESIs were pre-defined as a group of MedDRA preferred terms (Table S3), allowing for a comprehensive review of safety data not limited to a specific preferred term, and included ALT increase, vascular inflammation and complement activation, renal injury, and thrombocytopenia for bepirovirsen, and ALT increase, hematologic events, and psychiatric events for Peg-IFN.” (opens the source at this quote in a new tab)

Planned enrollment and Bayesian sample size basis

“The trial planned to enrol approximately 100 participants. Using Bayesian inference, this sample size was selected to provide a 65% posterior probability of the true response rate exceeding a fixed threshold under a range of assumed values for the threshold of interest and a range of assumed true response rates.” (opens the source at this quote in a new tab)

Planned participants per arm in the protocol

“For the entire study, approximately 100 participants will be enrolled and randomised to achieve approximately 50 participants in each treatment group” (opens the source at this quote in a new tab)

Participants enrolled and analyzed

“The intent-to-treat population included 108 participants (Arm 1, n = 55; Arm 2, n = 53).” (opens the source at this quote in a new tab)

Participants contributing to the biomarker analysis

“This post hoc peripheral longitudinal biomarker exploratory analysis examined serum proteomics and whole blood transcriptomics from peripheral blood mononuclear cell samples from 82 participants.” (opens the original PMC page at this quote in a new tab)View highlighted excerpt for this quote

Efficacy and safety populations

“Efficacy analyses were conducted on the intent-to-treat (ITT) population, which included all randomised participants, analysed based on treatment assignment. Safety analyses were conducted on all randomised participants who received ≥1 dose of study treatment, analysed based on treatment received.” (opens the source at this quote in a new tab)

Primary analysis model and its simplification

“Primary analysis was conducted when the last participant completed their Week 72 visit. Due to convergence issues caused by quasi-complete separation in the data, the planned Bayesian hierarchical model for the primary analysis was simplified to a Bayesian logistic regression model with fixed effects for treatment arm and baseline stratification factors.” (opens the source at this quote in a new tab)

Handling of wide disruptive events such as COVID-19

“For participants who experienced WDEs (e.g. COVID-19 infections or pandemic lockdowns) resulting in discontinuation or significant interruption/non-adherence to the investigational product (a gap of ≥21 days between treatment doses during the on-treatment period except for the Peg-IFN eligibility window), this model utilised multiple imputation (assuming missing at random) to assess the primary outcome.” (opens the source at this quote in a new tab)

Protocol definitions of analysis populations

Population	Quoted description
Intent-to-treat	“All randomised participants.”
Safety	“All participants who were randomised and received at least one dose of study treatment”
Multiplicity	“No adjustments will be made for multiplicity.”

Biomarker analysis population and response subgroups

“All B-Together randomized participants from the rest of the world population (excluding China), who had been assessed for the primary outcome and who received at least one dose of bepirovirsen treatment and ≥ 12 doses of Peg-IFN were included in biomarker analyses [8].” (opens the original PMC page at this quote in a new tab)View highlighted excerpt for this quote
“Null responders were non-responders who did not have two or more consecutive visits with a > 0.4 log10 HBsAg reduction while on treatment.” (opens the original PMC page at this quote in a new tab)View highlighted excerpt for this quote

Withdrawals, Peg-IFN eligibility, and missed visits

“Nine participants (Arm 1, n = 6; Arm 2, n = 3) withdrew from the study, two due to AEs (Fig. S1). All but two participants (one in each Arm) were eligible to begin Peg-IFN treatment following bepirovirsen treatment. COVID-19 infections and lockdowns resulted in missed treatment visits for nine participants (all in Arm 1).” (opens the source at this quote in a new tab)

Registry participant flow

Disposition	Arm 1 (24 weeks bepirovirsen)	Arm 2 (12 weeks bepirovirsen)
Started	“55”	“53”
Completed	“49”	“50”
Not completed	“6”	“3”
Reason not completed: adverse event	“2”	“0”
Reason not completed: lost to follow-up	“1”	“0”
Reason not completed: physician decision	“0”	“1”
Reason not completed: withdrawal by subject	“3”	“2”

Treatment discontinuations for adverse events

“Four participants permanently discontinued treatment due to an AE (1 dermatitis allergic and 2 injection-site reactions during bepirovirsen treatment; 1 depression during Peg-IFN treatment).” (opens the source at this quote in a new tab)

Missed doses due to wide disruptive events

“For nine participants in Arm 1, WDEs resulted in discontinuation of, significant interruption of, or non-adherence to the investigational product. Of these participants, two missed bepirovirsen dosing (2–4 missed visits) and seven missed Peg-IFN dosing (2–13 missed visits; Table S6).” (opens the source at this quote in a new tab)

Baseline HBsAg level and HBeAg status: most participants above 1,000 IU/ml and HBeAg negative

“Baseline demographics and disease characteristics were similar between arms (Table 1).12 Baseline demographics were also similar to the B-Clear study (Table S4). Most participants had baseline HBsAg >1,000 IU/ml (n = 74 [69%]) and were HBeAg− (n = 80 [74%]).” (opens the source at this quote in a new tab)

Baseline demographics and disease characteristics

Characteristic	Arm 1 (n = 55)	Arm 2 (n = 53)	Overall (N = 108)
Age, mean (SD), years	“45.5 (10.7)”	“45.5 (9.4)”	“45.5 (10.0)”
Female, n (%)	“11 (20)”	“20 (38)”	“31 (29)”
Male, n (%)	“44 (80)”	“33 (62)”	“77 (71)”
Asian, n (%)	“30 (55)”	“26 (49)”	“56 (52)”
Black or African American, n (%)	“3 (5)”	“3 (6)”	“6 (6)”
White, n (%)	“22 (40)”	“24 (45)”	“46 (43)”
BMI, mean (SD), kg/m2	“23.7 (3.4)”	“25.0 (4.2)”	“24.3 (3.9)”
HBsAg, mean (SD), IU/ml	“5,231.4 (8,939.8)”	“5,218.5 (8,437.4)”	“5,225.1 (8,656.2)”
HBsAg ≤1,000 IU/ml, n (%)	“17 (31)”	“17 (32)”	“34 (31)”
HBsAg >1,000 IU/ml, n (%)	“38 (69)”	“36 (68)”	“74 (69)”
HBV DNA <LLOQ, n (%)	“49 (89)”	“48 (91)”	“97 (90)”
HBeAg negative, n (%)	“41 (75)”	“39 (74)”	“80 (74)”
HBeAg positive, n (%)	“14 (25)”	“14 (26)”	“28 (26)”
TAF or TDF, n (%)	“33 (61)”	“37 (70)”	“70 (65)”
Entecavir, n (%)	“19 (35)”	“12 (23)”	“31 (29)”
ALT ≤ULN, n (%)	“51 (93)”	“43 (81)”	“94 (87)”
ALT >ULN, n (%)	“4 (7)”	“10 (19)”	“14 (13)”
HBV genotype undetermined, n (%)	“28 (51)”	“26 (49)”	“54 (50)”
METAVIR fibrosis score available, n	“7”	“8”	“15”

Baseline immunological heterogeneity between arms

“Higher expression of 148 proteins was seen in a subset of Arm 2 versus Arm 1 participants (Fig. S1a and S1b).” (opens the original PMC page at this quote in a new tab)View highlighted excerpt for this quote

Primary outcome: 5 of 55 participants in Arm 1 and 8 of 53 in Arm 2

“The primary outcome was achieved in 5 (9%) and 8 (15%) participants in Arm 1 and Arm 2, respectively. The resulting point estimates (95% credible interval) from the Bayesian model were 12% (0, 46%) for Arm 1 and 15% (0, 56%) for Arm 2 (Table 2). Seven (13%) participants in Arm 1 and 8 (15%) in Arm 2 achieved the alternative primary outcome (Table 2).” (opens the source at this quote in a new tab)

Primary outcome by baseline HBsAg and HBeAg stratum

Stratum	Arm 1 n	Arm 1 responders, n (%)	Arm 2 n	Arm 2 responders, n (%)
Stratum 1: HBsAg ≤1,000 IU/ml, HBeAg negative	“16”	“3 (19)”	“14”	“7 (50)”
Stratum 2: HBsAg >1,000 IU/ml, HBeAg negative	“25”	“1 (4)”	“25”	“1 (4)”
Stratum 3: HBsAg ≤1,000 IU/ml, HBeAg positive	“1”	“1 (100)”	“3”	“0”
Stratum 4: HBsAg >1,000 IU/ml, HBeAg positive	“13”	“0”	“11”	“0”

Registry primary outcome and between-arm comparison

Measure	Arm 1	Arm 2
Participants analyzed (ITT)	“55”	“53”
Percentage achieving SVR for 24 weeks off treatment	“9”	“15”

“The point estimate for the difference in SVR and its respective credible interval (CI) were evaluated at 24 weeks off of planned treatment for both arms.” (opens the source at this quote in a new tab)
“-29 to 11” (opens the source at this quote in a new tab)

Baseline HBsAg threshold among responders

“All participants who met the primary outcome had baseline HBsAg ≤3,000 IU/ml.” (opens the source at this quote in a new tab)

Predictors of response in the logistic regression model

“The final model identified low baseline HBsAg level, high BMI, and treatment with tenofovir alafenamide fumarate or tenofovir disoproxil fumarate as statistically significant predictors of response (Table 3).” (opens the source at this quote in a new tab)

Odds ratios for response by baseline HBsAg, treatment arm, and NA type

Comparison	Responders in reference group, n/N (%)	Responders in comparison group, n/N (%)	Odds ratio (95% CI)	p value
Baseline HBsAg >1,000 versus ≤1,000 IU/ml	“11/34 (32.4)”	“2/73 (2.7)”	“0.05 (0.00, 0.26)”	“<0.0001”
Arm 2 versus Arm 1	“5/54 (9.3)”	“8/53 (15.1)”	“1.01 (0.17, 5.96)”	“0.99”
Entecavir versus TAF and TDF	“10/70 (14.3)”	“2/31 (6.5)”	“0.09 (0.01, 0.61)”	“0.0119”
HBeAg positive versus negative	“12/79 (15.2)”	“1/28 (3.6)”	“0.28 (0.01, 3.27)”	“0.3263”

Response and relapse at the end of each treatment window

“At bepirovirsen EoT, 20/55 (36%; Arm 1) and 12/53 (23%; Arm 2) participants achieved HBsAg and HBV DNA <LLOQ. Of these, 8/20 (40%; Arm 1) and 5/12 (42%; Arm 2) relapsed before Peg-IFN EoT. At Peg-IFN EoT, 13/55 (24%; Arm 1) and 9/53 (17%; Arm 2) participants had HBsAg and HBV DNA <LLOQ. Of these, 7/12 (58%; Arm 1) and 0/9 (0%; Arm 2) relapsed during follow-up.” (opens the source at this quote in a new tab)

Peg-IFN did not convert partial responders

“Only two participants in Arm 2 who did not achieve HBsAg and HBV DNA <LLOQ at bepirovirsen EoT had HBsAg and HBV DNA loss following Peg-IFN treatment.” (opens the source at this quote in a new tab)
“Both these participants were likely on their way to HBsAg and HBV DNA <LLOQ irrespective of Peg-IFN (Fig. S3).” (opens the source at this quote in a new tab)

Indirect comparison of relapse with B-Clear

“In B-Clear, 63% and 75% participants relapsed off-treatment after 24-week and 12-week bepirovirsen 300 mg treatment, respectively. Here, 58% (Arm 1) and 0% (Arm 2) of participants who achieved HBsAg and HBV DNA <LLOQ at Peg-IFN EoT relapsed off-treatment.” (opens the source at this quote in a new tab)

Durability of primary outcome response in Arm 2

“All but one participant in Arm 2 who achieved the primary outcome sustained HBsAg and HBV DNA <LLOQ until off-treatment Week 36.” (opens the source at this quote in a new tab)

HBeAg loss and anti-HBe seroconversion

“Of the participants who were HBeAg+ at baseline, 8/14 (57%) and 1/14 (7%) became anti-HBe+ by the end of the bepirovirsen treatment period in Arm 1 and Arm 2, respectively. At the end of the Peg-IFN treatment period, this changed to 5/9 (56%) and 1/9 (11%), respectively.” (opens the source at this quote in a new tab)

Anti-HBs response

“Among participants who achieved HBsAg and HBV DNA <LLOQ by Peg-IFN EoT, anti-HBs status did not correlate with primary outcome achievement; anti-HBs values were highly variable and overlapping in those who relapsed and those who maintained HBsAg <LLOQ (Fig. S8).” (opens the source at this quote in a new tab)

ALT elevations by treatment window

“Transient increases in ALT levels ≥3x ULN were seen in 38% of participants in Arm 1 and 40% in Arm 2; 24% and 19% of participants, respectively, experienced such increases during the bepirovirsen treatment window, and 18% and 25%, respectively, during the Peg-IFN treatment window. During the bepirovirsen treatment window, ALT increases coincided with a decline in HBsAg, whereas there was no association between HBsAg and ALT levels during the Peg-IFN treatment window (Fig. S9).” (opens the source at this quote in a new tab)

Comparison of bepirovirsen treatment durations

“A high degree of variability in both treatment arms, indicated by wide overlapping confidence intervals (CI), prevented a clear conclusion on the optimal bepirovirsen treatment duration to include in sequential therapy with Peg-IFN (Fig. S10).” (opens the source at this quote in a new tab)

Median time to ALT normalization in participants with baseline ALT above ULN

Measure	Arm 1	Arm 2
Participants analyzed	“4”	“10”
Median, weeks	“not estimable (NE)”	“4.1”
95% CI, weeks	“(6.3 to NA)”	“(1.1 to 47.1)”

Peripheral immune activation and markers of hepatocyte death

“Increases from baseline in mean expression of serum proteins with immune effector and apoptotic functions (Week 3) and transcripts associated with immune cell proliferation and activation (Week 5) were observed regardless of arm or response subgroup. By Week 8, serum liver and apoptosis-specific proteins were increased; this was more pronounced in responders than non-responders, with the difference more marked in Arm 1 versus Arm 2. Increased abundance of these proteins was highly correlated with ALT levels, which were often associated with transient hepatitis B virus (HBV) DNA elevations and HBsAg decreases.” (opens the original PMC page at this quote in a new tab)View highlighted excerpt for this quote

Longer-term follow-up of responders in B-Sure

“Up to 12 months’ durability of functional cure with bepirovirsen plus Peg-IFN sequential therapy in B-Together responders: B-Sure study” (opens the source at this quote in a new tab)
“Data will be presented from a sub-analysis from the B-Sure phase II study, evaluating the achievement and duration of a functional cure in CHB patients who responded to treatment with bepirovirsen followed by Peg-IFN, a type of immune-boosting therapy.” (opens the source at this quote in a new tab)

Follow-up of responders planned in B-Sure

“Long-term durability of response will be evaluated in the B-Sure trial (NCT04954859),23 which will follow responders from B-Together and B-Clear for an additional 33 months and allows for cessation of NA therapy and assessment of functional cure.” (opens the source at this quote in a new tab)

Overall adverse events, discontinuations, serious adverse events, and deaths

“The proportion of participants who experienced AEs in the bepirovirsen and Peg-IFN treatment windows was similar between treatment arms (95% vs. 98%), with most AEs considered treatment related (Table 4). Few AEs led to treatment discontinuation (4%; 3 during bepirovirsen and 1 during Peg-IFN treatment).” (opens the source at this quote in a new tab)
“There were no reports of serious liver injury, clinically significant bleeding events, renal injury, or vascular injury linked to alternate complement pathway activation. Nine serious AEs (SAEs) were reported in eight participants, five during bepirovirsen treatment (one considered treatment related) and three in the off-treatment period. No deaths were reported. The most common AEs were injection-site erythema (Table S7).” (opens the source at this quote in a new tab)

Adverse events over the whole study

Event, n (%)	Arm 1 (n = 55)	Arm 2 (n = 53)	Total (N = 108)
Any AE	“52 (95)”	“52 (98)”	“104 (96)”
Any AE related to study treatment	“52 (95)”	“48 (91)”	“100 (93)”
AEs leading to permanent discontinuation	“2 (4)”	“2 (4)”	“4 (4)”
Any grade 3 and 4 event	“31 (56)”	“23 (43)”	“54 (50)”
Any SAE	“6 (11)”	“2 (4)”	“8 (7)”
AESI: ALT increase	“27 (49)”	“11 (21)”	“38 (35)”
AESI: vascular inflammation and complement activation	“19 (35)”	“11 (21)”	“30 (28)”
AESI: thrombocytopenia	“29 (53)”	“19 (36)”	“48 (44)”
AESI: renal injury	“10 (18)”	“7 (13)”	“17 (16)”
Neutropenia	“10 (18)”	“12 (23)”	“22 (20)”
Leukopenia	“10 (18)”	“7 (13)”	“17 (16)”

Adverse events during the bepirovirsen treatment window

Event, n (%)	Arm 1 (n = 55)	Arm 2 (n = 53)	Total (N = 108)
Any AE	“48 (87)”	“42 (79)”	“90 (83)”
Any AE related to study treatment	“47 (85)”	“39 (74)”	“86 (80)”
AEs leading to permanent discontinuation	“1 (2)”	“2 (4)”	“3 (3)”
Any grade 3 and 4 event	“20 (36)”	“16 (30)”	“36 (33)”
Any SAE	“4 (7)”	“1 (2)”	“5 (5)”
AESI: ALT increase	“16 (29)”	“8 (15)”	“24 (22)”
AESI: vascular inflammation and complement activation	“16 (29)”	“6 (11)”	“22 (20)”
AESI: thrombocytopenia	“11 (20)”	“7 (13)”	“18 (17)”
AESI: renal injury	“10 (18)”	“6 (11)”	“16 (15)”

Adverse events during the Peg-IFN treatment window

Event, n (%)	Arm 1 (n = 55)	Arm 2 (n = 53)	Total (N = 108)
Any AE	“46 (84)”	“43 (81)”	“89 (82)”
Any grade 3 and 4 event	“17 (31)”	“8 (15)”	“25 (23)”
AESI: ALT increase	“16 (29)”	“5 (9)”	“21 (19)”
AESI: thrombocytopenia	“21 (38)”	“16 (30)”	“37 (34)”
Neutropenia	“10 (18)”	“11 (21)”	“21 (19)”

ALT increases and the serious adverse event of drug-induced liver injury

“For AESIs associated with increased ALT, a larger proportion of participants reported events in Arm 1 compared with Arm 2 (49% vs. 21%).” (opens the source at this quote in a new tab)
“Increases in ALT levels observed throughout the study were asymptomatic and resolved without increases in bilirubin. One ALT increase was reported as an SAE of drug-induced liver injury. The participant was not hospitalised for the event and reported no signs or symptoms associated with the ALT increase. The treatment was held at Week 7 and withdrawn at Week 9, when the participant met the stopping criteria per protocol (ALT >10x ULN for ≥4 weeks).” (opens the source at this quote in a new tab)

Grade 3 and 4 events and DAIDS laboratory grading

“Grade 3 and 4 events occurred in 33% of participants during the bepirovirsen period and 23% in the Peg-IFN period. Some Grade 3/4 AEs were driven by Division of AIDS Table for Grading the Severity of Adult and Pediatric Adverse Events (DAIDS) grading for laboratory changes and may not have been correlated with clinical severity.” (opens the source at this quote in a new tab)

Platelet count decline

“A decline in platelet count was observed over the course of sequential treatment that improved in the follow-up period; compared with Peg-IFN, the influence of bepirovirsen was minimal.” (opens the source at this quote in a new tab)

Registry serious and selected other adverse events through Week 72

Event, affected / at risk (%)	Arm 1	Arm 2
All-cause mortality	“0/55 (0.00%)”	“0/53 (0.00%)”
Serious adverse events, total	“6/55 (10.91%)”	“2/53 (3.77%)”
Other adverse events, total	“52/55 (94.55%)”	“50/53 (94.34%)”
Injection site erythema	“20/55 (36.36%)”	“22/53 (41.51%)”
Pyrexia	“17/55 (30.91%)”	“14/53 (26.42%)”
Alanine aminotransferase increased	“23/55 (41.82%)”	“6/53 (11.32%)”
Thrombocytopenia	“14/55 (25.45%)”	“11/53 (20.75%)”
Influenza like illness	“12/55 (21.82%)”	“8/53 (15.09%)”
COVID-19 (non-serious)	“15/55 (27.27%)”	“5/53 (9.43%)”

Summary: study limitations

B-Together randomized 108 participants (55 to 24 weeks and 53 to 12 weeks of bepirovirsen, each followed by up to 24 weeks of Peg-IFN) and had no bepirovirsen-only control arm, so the effect of adding Peg-IFN was assessed by indirect comparison with B-Clear. Treatment was open label, the sample size of approximately 100 was selected to provide a 65% posterior probability of the true response rate exceeding a fixed threshold, and the 95% credible intervals for the primary outcome were 0 to 46% in Arm 1 and 0 to 56% in Arm 2. COVID-19 infections and lockdowns caused missed visits for 9 participants, all in Arm 1. All primary outcome responders had baseline HBsAg of 3,000 IU/ml or lower, and the approved label recommends against concomitant administration of bepirovirsen with peginterferon alfa-2a. GSK funded the trial and took part in its design, conduct, and analysis, and authors employed by GSK hold financial equities in GSK.

Absence of a bepirovirsen-only control group

“Firstly, the B-Together study design lacks control groups, i.e. bepirovirsen 300 mg for 12 or 24 weeks alone, which limits comparisons of bepirovirsen and Peg-IFN sequential therapy with bepirovirsen alone. Thus, the study uses indirect comparisons to the B-Clear study.” (opens the source at this quote in a new tab)

Effect of the COVID-19 pandemic on Arm 1

“Secondly, the COVID-19 pandemic disproportionately affected Arm 1 during the Peg-IFN treatment period, with nine participants missing 50 Peg-IFN dosing visits on Arm 1.” (opens the source at this quote in a new tab)
“However, the modelled point estimate may still have underestimated the response rate had this confounding not occurred.” (opens the source at this quote in a new tab)

Conservative primary outcome definition

“Secondly, the primary outcome definition of HBsAg and HBV DNA <LLOQ maintained at all timepoints off-treatment is highly conservative and likely underestimated the proportion of participants with response.” (opens the source at this quote in a new tab)

Small arms and heterogeneity limit between-arm comparison

“However, it is difficult to fully compare the differences in arms, owing to both the relatively small study size and the heterogeneity observed in the study population. Due to the small participant numbers in each arm and a high level of variability (wide and overlapping 95% CIs), there was uncertainty regarding between-treatment differences.” (opens the source at this quote in a new tab)

Hypothesized improvement not shown

“While the study did not show the substantive improvement in bepirovirsen response we hypothesised, further studies of sequential or combination therapy with bepirovirsen and other agents such as HBsAg-lowering agents (e.g. siRNA) or different immunomodulators are being considered to improve response rates and increase the treatable population, and ultimately deliver functional cure to more patients.” (opens the source at this quote in a new tab)

Limitations of the post hoc biomarker analysis

“These analyses were opportunistic and not powered.” (opens the original PMC page at this quote in a new tab)View highlighted excerpt for this quote
“Second, all patients in B-Together received bepirovirsen, and findings could not be validated against a placebo-treated population within this study. Third, due to limited PBMC availability, this study could not demonstrate activation of HBV-specific T and B cells in the liver or periphery.” (opens the original PMC page at this quote in a new tab)View highlighted excerpt for this quote
“Finally, due to operational complexities, this study did not include liver sampling, which would have been required to show direct evidence of killing of infected hepatocytes.” (opens the original PMC page at this quote in a new tab)View highlighted excerpt for this quote

Applicability: label recommendation on peginterferon alfa-2a

“Concomitant administration of HIBSAGO and peginterferon alfa-2a is not recommended due to the potential for an exacerbated risk of increased ALT, decreased platelet count, and immune-mediated effects.” (opens the prescribing information at this quote)

Funding, sponsor role, and author employment

“This trial was funded by GSK (GSK ID: 209348/NCT04676724). The sponsor, GSK, was involved in the design, conduct, and analysis of the trial and wrote the trial report.” (opens the source at this quote in a new tab)
“SD, MP, LF, ML, TG, JL, DL, GMG, HP, ASY, RE, SK and DT are employed by GSK and hold financial equities in GSK.” (opens the source at this quote in a new tab)