If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Barts Liver Centre, Blizard Institute, Barts and the London School of Medicine and Dentistry, Queen Mary University of London, London, United KingdomDepartment of Medicine, Queen Mary Hospital, School of Clinical Medicine, The University of Hong Kong, Hong Kong
Corresponding author. and contact details: Department of Immunobiology The Blizard Institute, Barts and The London School of Medicine and Dentistry, Queen Mary University of London London, United Kingdom.
First-line antivirals such as Entecavir and Tenofovir remain the mainstay of CHB treatment, but there remains considerable interest in the potential for treatment discontinuation in select patients to maintain partial cure or achieve functional cure.
•
Novel viral markers, HBcrAg and HBV RNA, in conjunction with quantitative hepatitis B surface antigen, have demonstrated utility in predicting off-therapy partial cure, with emerging evidence also supporting a correlation with functional cure.
•
There is a dearth of novel immune marker studies, particularly those analysing HBV-specific T cell responses, which have the potential to predict immune restoration with treatment discontinuation, considered a central tenet of functional cure.
•
Both virus-targeted and immune modulatory agents (where NA discontinuation can be considered an immunomodulatory strategy) should be utilised in concert with the goal of achieving functional cure where the risk of severe clinical relapse is considered low.
Abstract
Background & Aims
Antivirals represent the mainstay of CHB treatment given their efficacy and tolerability, but rates of functional cure remain low during long-term therapy. Treatment discontinuation has emerged as a strategy to maintain partial cure and achieve functional cure in select patient groups. We aimed to evaluate how data from treatment discontinuation studies exploring novel viral and/or immune markers could be applied to the functional cure program.
Methods
Treatment discontinuation studies evaluating novel viral and/or immune markers were identified by a systematic search of the PubMed database through to October 30, 2022. Data extraction focused on information regarding novel markers, including identified cut-off levels, timing of measurement and associated effect on study outcomes of virological relapse (VR), clinical relapse (CR) and HBsAg seroclearance.
Results
From a search of 4492 citations, 33 studies comprising a minimum of 2986 unique patients met the inclusion criteria. Novel viral markers, HBcrAg and HBV RNA, were demonstrated across most studies to be helpful in predicting off-therapy partial cure, with emerging evidence to support a link with functional cure. From novel immune marker studies, we observed that treatment discontinuation has the potential to trigger immune restoration, which may be associated with a transient VR. To this end, these studies support the combination of virus-directing agents with immunomodulator therapies to induce two key steps underlying functional cure: viral antigen load reduction and restoration of the host immune response.
Conclusions
Patients with a favourable profile of novel viral and immune markers stand to benefit from a trial of antiviral treatment discontinuation alongside novel virus-directing agents with the aim of achieving functional cure without excessive risk of severe clinical relapse.
Lay Summary
Select Chronic Hepatitis B patients undergoing nucleoside analogue therapy may benefit from a trial of treatment discontinuation, aiming to maintain partial cure and/or achieve functional cure. We propose a profile of novel viral and immune markers to identify patients who are likely to achieve these goals without excessive risk of hepatic decompensation. Furthermore, treatment discontinuation may also be considered as a therapeutic strategy to trigger immune restoration, which may increase the chance of functional cure when used in conjunction with novel virus-directing agents.
All data analysed during this study are available from the corresponding author on reasonable request.
Conflict of Interest Statement
Georgia Zeng, Apostolos Koffas and Upkar Gill have no conflicts of interest to disclose.
Lung-Yi Mak has served as an advisory board member for Gilead Sciences.
Patrick Kennedy has served as a speaker, a consultant/advisory board member for Abbott Diagnostics, Aligos, Antios Therapeutics, Assembly Biosciences, Gilead Sciences, Janssen, GlaxoSmithKline, Immunocore and Drug Farm, and has received research funding from Gilead Sciences.
Financial Support
None.
Author contributions
GZ.: data curation; drafting; reviewing & editing of manuscript.
AK.: data curation; drafting; reviewing & editing of manuscript.
L-YM: drafting; reviewing & editing of manuscript.
USG: reviewing & editing of manuscript.
PTFK.: conceptualization; reviewing & editing of manuscript; supervision.
Introduction
An estimated 296 million individuals are known to have chronic hepatitis B (CHB) worldwide, with 30% of the global population showing serological evidence of current or past infection.
CHB resulted in an estimated 820,000 deaths in 2019 according to the World Health Organization (WHO), the vast majority of which are attributable to cirrhosis and hepatocellular carcinoma (HCC).
Current CHB treatment aims primarily to prevent disease progression and the sequelae of chronic infection by providing continuous on-treatment viral suppression.
First-line antivirals, Entecavir (ETV), Tenofovir Disoproxil Fumarate (TDF) and Tenofovir Alafenamide (TAF) represent the mainstay of treatment given their efficacy, tolerability and favourable safety profile, moreover they are distinguished by their high barriers to resistance in addition to their ability to reverse liver fibrosis and reduce HCC incidence.
Treatment with nucleoside analogues (NAs) is lifelong in the majority of patients. This is in contrast to treatment with interferon-alfa, the only recognised finite therapy in CHB, employed in a small subset of patients only, due to its recognised systemic side effects. NAs lack the potential to achieve functional cure, defined as sustained off-treatment hepatitis B surface antigen (HBsAg) loss, in the majority of CHB patients. The persistence of HBV infection is attributed to the cccDNA pool in infected hepatocytes; although it reduces naturally over the course of HBV infection
Novel therapeutic approaches for the management of CHB have been under evaluation to overcome these limitations. These comprise a number of promising agents or combination approaches currently being evaluated in pre-clinical or early-phase clinical trials, which target either the viral cycle directly or enhance host immunity. The former group includes viral entry inhibitors, RNA interference, capsid assembly modulators (CAMs), nucleic acid polymers, strategies targeting cccDNA formation or degradation, amongst others. Examples of the latter group include therapeutic vaccines, toll-like receptor agonists, T cell redirection, checkpoint inhibitors, antibodies to HBV and indeed NA discontinuation. Recently, considerable focus has been given to NA discontinuation as a strategy to achieve functional cure. However, there is a lack of consensus between international guidelines
regarding the requirements for safe NA cessation in CHB patients (Supplementary Table 1). Secondly, patients often experience viral relapse (VR), defined as a rebound of HBV DNA levels following treatment cessation; and clinical relapse (CR), defined by VR with an associated biochemical flare. Off-therapy rates of VR and CR vary largely between published studies, likely due to heterogeneity in study participants, relapse definitions and other aspects of study design. In a recent meta-analysis by Hall et al in 2021
which explored rates of partial cure following discontinuation of oral antivirals in HBeAg-negative patients, rates of VR and CR at 12 months were 63% and 35% respectively.
While treatment discontinuation can be considered a therapeutic strategy in its own right with the potential to offer partial and functional cure in some patients, studies of NA discontinuation can also provide unique insights into the virological and immunological conditions required to achieve both partial and functional cure. Several discontinuation studies assessed novel viral markers such as Hepatitis B core-related antigen (HBcrAg) and HBV RNA, and both have been proposed as novel tools to signpost partial and functional cure post NA cessation. Additionally, immune markers, particularly relating to T cell phenotype and function, are differentiated in patient populations who progress to VR and/or CR. Thus, we seek to comprehensively review the data generated to date on novel viral and immune markers in treatment discontinuation studies, aiming to evaluate their potential in providing a roadmap to functional cure.
Methods
Literature Search
We performed a systematic review according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.
In order to retrieve all works of potential relevance, a systematic search of the PubMed/Medline database was performed of all studies through to October 30, 2022. The search used the terms (“Hepatitis B” OR “Chronic Hepatitis B”) AND (“Antiviral” OR “Treatment” OR “Therapy” OR “Lamivudine” OR “Adefovir” OR “Entecavir” OR “Telbivudine” OR “Tenofovir”) AND (“End” OR “Discontinuation” OR “Withdrawal” OR “Cessation” OR “Off-treatment”) which were searched as text words and as exploded medical subject headings where possible, with no language restrictions. The reference lists of relevant articles were also searched for appropriate studies. We requested full texts from authors where we found relevant paper abstracts and conference abstracts. A search for unpublished literature was not performed.
Inclusion criteria
We included randomised or observational studies that met the following inclusion criteria: (1) studies including adult CHB patients who ceased NA only if fulfilling the following standards: HBeAg seroconversion and a minimum mean/median of 6 months of consolidation therapy following virological suppression for initial HBeAg-positive populations, and a minimum median/mean of 12 months of consolidation therapy following virological suppression for initial HBeAg-negative populations, without HBsAg seroclearance; (2) studies providing data in the form of virological and/or clinical relapse rates; (3) studies providing data relating to novel viral and/or immune markers; (4) studies with a minimum follow up of 6 months; (5) studies with a minimum of 10 patients; (6) studies available in English as full papers.
Exclusion criteria
We excluded studies with (1) populations co-infected with hepatitis C virus (HCV) or human immunodeficiency virus (HIV); (2) studies with populations with a history of HCC, liver transplants or immunosuppressive therapies, (3) studies with populations co-treated with interferon; (4) studies with populations that have exclusively experienced HBsAg seroclearance.
Data extraction
The baseline characteristics of study cohort including age, gender, type of NA, HBeAg status, HBV genotype, and duration of NA were extracted. For each article included, we recorded the author names, year of publication, country of origin, study design, and duration of follow-up. Study outcomes of VR and CR (both regarded as not achieving partial cure) and HBsAg seroclearance (functional cure) as defined in each article were recorded. Regarding novel viral biomarkers, HBcrAg and HBV RNA, the identified cut-off levels, and timing of measurement were presented alongside the associated effect estimates on study outcomes, expressed as either hazard ratios (HR), odds ratios (OR), or cumulative rate of study outcomes. Data regarding novel immune markers, namely the phenotype and function of peripheral immune cells, were harvested in the form supplied by the authors.
Quality assessment
For viral markers, the Risk of Bias in Non-Randomized Studies of Interventions (ROBINS-I) assessment tool was used to evaluate study quality, and is available in the Supplementary Materials. The judgements within each domain of the tool were carried forward to an overall risk of bias judgement, categorised as low, moderate, serious, or critical. Studies judged to be at critical risk of bias were not included in the analysis. For immune markers, due to the heterogeneous and complex nature of the immunological analyses, no well-established scale could be applied. Authors (GZ) and (AK) screened the abstracts and selected relevant studies after screening the retrieved full articles. Conflicts of study eligibility or quality assessment was resolved by discussion with a senior author (PTFK).
Results
Characteristics of included studies
The search identified 4492 titles and abstracts that were reviewed, with 41 citations being selected for full-text review. Of these, 8 studies were excluded after rigorous review. The stopping criteria in 7 of these studies did not meet the minimum requirements as per our inclusion criteria and we could not source the full text of another study. Therefore, we evaluated 33 studies,
Serum Level of Antibodies Against Hepatitis B Core Protein Is Associated With Clinical Relapse After Discontinuation of Nucleos(t)ide Analogue Therapy.
Pregenomic HBV RNA and Hepatitis B Core-Related Antigen Predict Outcomes in Hepatitis B e Antigen-Negative Chronic Hepatitis B Patients Suppressed on Nucleos(T)ide Analogue Therapy.
Combining Hepatitis B Virus RNA and Hepatitis B Core-Related Antigen: Guidance for Safely Stopping Nucleos(t)ide Analogues in Hepatitis B e Antigen-Positive Patients With Chronic Hepatitis B.
Predictors of hepatitis B surface antigen loss, relapse and retreatment after discontinuation of effective oral antiviral therapy in noncirrhotic HBeAg-negative chronic hepatitis B.
The role of hepatitis B virus core-related antigen in predicting hepatitis B virus relapse after cessation of entecavir in hepatitis B e antigen-negative patients.
Hepatitis B Core-Related Antigen is a Biomarker for off-Treatment Relapse After Long-Term Nucleos(t)ide Analog Therapy in Patients with Chronic Hepatitis B.
Pilot Study Using Machine Learning to Identify Immune Profiles for the Prediction of Early Virological Relapse After Stopping Nucleos(t)ide Analogues in HBeAg-Negative CHB.
Serum hepatitis B virus RNA level is associated with biochemical relapse in patients with chronic hepatitis B infection who discontinue nucleos(t)ide analogue treatment.
Increased NK Cell Function After Cessation of Long-Term Nucleos(t)ide Analogue Treatment in Chronic Hepatitis B Is Associated With Liver Damage and HBsAg Loss.
Immunological biomarkers as indicators for outcome after discontinuation of nucleos(t)ide analogue therapy in patients with HBeAg-negative chronic hepatitis B.
Genetic variations in the CXCR5 gene decrease the risk of clinical relapse after discontinuation of nucleos(t)ide analogue therapy in patients with chronic hepatitis B.
Clinical Utility of SCALE-B to Predict Hepatitis B Virus Relapse, Hepatitis B Surface Antigen Loss After Antiviral Cessation in Asian Patients After 2-Year Follow-up.
extracted data from the CREATE database, which pooled cohorts from previous studies in Asia and Europe that were already included in this meta-analysis.
Pregenomic HBV RNA and Hepatitis B Core-Related Antigen Predict Outcomes in Hepatitis B e Antigen-Negative Chronic Hepatitis B Patients Suppressed on Nucleos(T)ide Analogue Therapy.
Predictors of hepatitis B surface antigen loss, relapse and retreatment after discontinuation of effective oral antiviral therapy in noncirrhotic HBeAg-negative chronic hepatitis B.
Clinical Utility of SCALE-B to Predict Hepatitis B Virus Relapse, Hepatitis B Surface Antigen Loss After Antiviral Cessation in Asian Patients After 2-Year Follow-up.
Combining Hepatitis B Virus RNA and Hepatitis B Core-Related Antigen: Guidance for Safely Stopping Nucleos(t)ide Analogues in Hepatitis B e Antigen-Positive Patients With Chronic Hepatitis B.
Serum Level of Antibodies Against Hepatitis B Core Protein Is Associated With Clinical Relapse After Discontinuation of Nucleos(t)ide Analogue Therapy.
Serum hepatitis B virus RNA level is associated with biochemical relapse in patients with chronic hepatitis B infection who discontinue nucleos(t)ide analogue treatment.
The role of hepatitis B virus core-related antigen in predicting hepatitis B virus relapse after cessation of entecavir in hepatitis B e antigen-negative patients.
. Distinct data on initial e-Antigen positive populations, was provided in 5 studies, 14 studies provided distinct data on initial e-Antigen negative populations and 14 studies provided data on combined e-Antigen positive and negative populations. 22 studies were conducted in Asian-dominant populations, 4 studies were conducted in Caucasian-dominant populations, 3 studies reported on Mediterranean-dominant populations, 1 study was conducted in a Black African-dominant population and 3 studies were conducted in heterogenous populations. Figure 1 displays our study selection process.
The undetectable limit of HBV DNA in the majority of studies was 20 IU/mL (100 copies/ml), but varied from 10 IU/mL to 100 IU/mL. When specified, the definition of VR was set at HBV DNA >2000 IU/mL in all, but one study
Immunological biomarkers as indicators for outcome after discontinuation of nucleos(t)ide analogue therapy in patients with HBeAg-negative chronic hepatitis B.
, which utilised the threshold of HBV DNA >20,000 IU/mL. The definition CR was set as alanine aminotransferase (ALT) >2x upper limit of normal (ULN) in all studies that specified a threshold, but one study specifically looked at severe hepatitis flares, defined as ALT >10x ULN.
The definition of VR and CR in some studies was qualified by multiple time points, for example VR being defined as HBV DNA >2000 IU/mL verified on 2 separate occasions 3 months apart. The main study characteristics of these studies are summarized in Table 1A, Table 1BA and 1B , while the patient and treatment characteristics of these studies are summarised in Table 2A, Table 2BA and 2B .
Table 1AMain Characteristics of Included Studies (n=24) Exploring the Role of Viral Markers in Prediction of Partial Cure.
Pregenomic HBV RNA and Hepatitis B Core-Related Antigen Predict Outcomes in Hepatitis B e Antigen-Negative Chronic Hepatitis B Patients Suppressed on Nucleos(T)ide Analogue Therapy.
Combining Hepatitis B Virus RNA and Hepatitis B Core-Related Antigen: Guidance for Safely Stopping Nucleos(t)ide Analogues in Hepatitis B e Antigen-Positive Patients With Chronic Hepatitis B.
Predictors of hepatitis B surface antigen loss, relapse and retreatment after discontinuation of effective oral antiviral therapy in noncirrhotic HBeAg-negative chronic hepatitis B.
The role of hepatitis B virus core-related antigen in predicting hepatitis B virus relapse after cessation of entecavir in hepatitis B e antigen-negative patients.
Hepatitis B Core-Related Antigen is a Biomarker for off-Treatment Relapse After Long-Term Nucleos(t)ide Analog Therapy in Patients with Chronic Hepatitis B.
Pilot Study Using Machine Learning to Identify Immune Profiles for the Prediction of Early Virological Relapse After Stopping Nucleos(t)ide Analogues in HBeAg-Negative CHB.
Serum hepatitis B virus RNA level is associated with biochemical relapse in patients with chronic hepatitis B infection who discontinue nucleos(t)ide analogue treatment.
Clinical Utility of SCALE-B to Predict Hepatitis B Virus Relapse, Hepatitis B Surface Antigen Loss After Antiviral Cessation in Asian Patients After 2-Year Follow-up.
Increased NK Cell Function After Cessation of Long-Term Nucleos(t)ide Analogue Treatment in Chronic Hepatitis B Is Associated With Liver Damage and HBsAg Loss.
Serum Level of Antibodies Against Hepatitis B Core Protein Is Associated With Clinical Relapse After Discontinuation of Nucleos(t)ide Analogue Therapy.
Immunological biomarkers as indicators for outcome after discontinuation of nucleos(t)ide analogue therapy in patients with HBeAg-negative chronic hepatitis B.
Genetic variations in the CXCR5 gene decrease the risk of clinical relapse after discontinuation of nucleos(t)ide analogue therapy in patients with chronic hepatitis B.
Pilot Study Using Machine Learning to Identify Immune Profiles for the Prediction of Early Virological Relapse After Stopping Nucleos(t)ide Analogues in HBeAg-Negative CHB.
Table 2APatient and Treatment Characteristics in Studies Exploring the Role of Viral Markers in Prediction of Partial and Functional Cure; Stratified by HBeAg Status.
Combining Hepatitis B Virus RNA and Hepatitis B Core-Related Antigen: Guidance for Safely Stopping Nucleos(t)ide Analogues in Hepatitis B e Antigen-Positive Patients With Chronic Hepatitis B.
Hepatitis B Core-Related Antigen is a Biomarker for off-Treatment Relapse After Long-Term Nucleos(t)ide Analog Therapy in Patients with Chronic Hepatitis B.
Pregenomic HBV RNA and Hepatitis B Core-Related Antigen Predict Outcomes in Hepatitis B e Antigen-Negative Chronic Hepatitis B Patients Suppressed on Nucleos(T)ide Analogue Therapy.
Predictors of hepatitis B surface antigen loss, relapse and retreatment after discontinuation of effective oral antiviral therapy in noncirrhotic HBeAg-negative chronic hepatitis B.
The role of hepatitis B virus core-related antigen in predicting hepatitis B virus relapse after cessation of entecavir in hepatitis B e antigen-negative patients.
Pilot Study Using Machine Learning to Identify Immune Profiles for the Prediction of Early Virological Relapse After Stopping Nucleos(t)ide Analogues in HBeAg-Negative CHB.
Serum hepatitis B virus RNA level is associated with biochemical relapse in patients with chronic hepatitis B infection who discontinue nucleos(t)ide analogue treatment.
Clinical Utility of SCALE-B to Predict Hepatitis B Virus Relapse, Hepatitis B Surface Antigen Loss After Antiviral Cessation in Asian Patients After 2-Year Follow-up.
Abbreviations: ADV: adefovir; ALT: alanine transaminase; CR: clinical relapse; CT: consolidation time; eAg+: e-Antigen positive; eAg-: e-Antigen negative; ETV: entecavir; EOT: end-of-treatment; HBsAg: hepatitis B surface antigen (measured in log IU/mL); HBV DNA: hepatitis B virus deoxyribonucleic acid (measured in log copies/mL); SR: sustained response; TDF: tenofovir; TTT: total treatment time (measured in months); ULN: upper limit of normal; VR: viral relapse; “.”: data not specified. Note #1: Fan, R., et al 2020A & Fan, R. et al 2020B use overlapping patient cohorts.
Increased NK Cell Function After Cessation of Long-Term Nucleos(t)ide Analogue Treatment in Chronic Hepatitis B Is Associated With Liver Damage and HBsAg Loss.
Serum Level of Antibodies Against Hepatitis B Core Protein Is Associated With Clinical Relapse After Discontinuation of Nucleos(t)ide Analogue Therapy.
Immunological biomarkers as indicators for outcome after discontinuation of nucleos(t)ide analogue therapy in patients with HBeAg-negative chronic hepatitis B.
Genetic variations in the CXCR5 gene decrease the risk of clinical relapse after discontinuation of nucleos(t)ide analogue therapy in patients with chronic hepatitis B.
Predictors of hepatitis B surface antigen loss, relapse and retreatment after discontinuation of effective oral antiviral therapy in noncirrhotic HBeAg-negative chronic hepatitis B.
Pilot Study Using Machine Learning to Identify Immune Profiles for the Prediction of Early Virological Relapse After Stopping Nucleos(t)ide Analogues in HBeAg-Negative CHB.
Abbreviations: ADV: adefovir; ALT: alanine transaminase; CR: clinical relapse; CT: consolidation time; eAg+: e-Antigen positive; eAg-: e-Antigen negative; ETV: entecavir; EOT: end-of-treatment; HBsAg: hepatitis B surface antigen (measured in log IU/mL); HBV DNA: hepatitis B virus deoxyribonucleic acid (measured in log copies/mL); LdT: telbivudine; SR: sustained response; TDF: tenofovir; TTT: total treatment time (measured in months); ULN: upper limit of normal; VR: viral relapse; “.”: data not specified. Note #1: Honer Zu Siederdissen, C., et al 2016, Rinker, F., et al 2018 and Zimmer, C.L., et al 2018 use the same patient cohort.
Hepatitis B Core-Related Antigen is a Biomarker for off-Treatment Relapse After Long-Term Nucleos(t)ide Analog Therapy in Patients with Chronic Hepatitis B.
Hepatitis B Core-Related Antigen is a Biomarker for off-Treatment Relapse After Long-Term Nucleos(t)ide Analog Therapy in Patients with Chronic Hepatitis B.
(including both HBeAg positive and negative patients) reported that lower EOT HBcrAg levels were significantly associated with higher rates of virological remission/response with multivariate OR 0.73 per log U/mL (0.62-0.86, p<0.001). Seven studies reported the cumulative rates of VR stratified by the level of baseline or EOT HBcrAg, with varying observation periods and cut-off levels of HBcrAg (Figure 2a). For instance, Tseng et al
demonstrated significantly different 5-year VR rates of 23.8% vs 53% in patients with baseline HBcrAg <4 log U/mL and >4 log U/mL respectively (p=0.001), yet returning no significant findings when exploring EOT HBcrAg, in a majority HBeAg negative population. Huang et al
The role of hepatitis B virus core-related antigen in predicting hepatitis B virus relapse after cessation of entecavir in hepatitis B e antigen-negative patients.
also found that baseline HBcrAg at the cut-off 4 log U/mL was a significant predictor for VR in their HBeAg negative population, while EOT HBcrAg was not.
Serum HBV DNA is quantified using the COBAS Taqman HBV test, with a lower limit of detection of 20 IU/mL. This study specifically made a distinction between patients with undetectable HBV DNA (48.5%) and patients with HBV DNA <20 IU/mL (43.8%), which was not performed in other studies.
Cumulative rates of VR, CR or functional cure stratified by the biomarker(s) of interest were summarized in Figures 2a-c.
Cumulative incidence of 4-year CR: no p value DNA negative & RNA <3 log U/mL: 8% DNA positive or RNA >3 log U/mL: 37% MV HR (DNA + or RNA >3 log U/mL vs DNA - & RNA <3 log U/mL): 11.10 (2.69-45.80) p=0.00
Combining Hepatitis B Virus RNA and Hepatitis B Core-Related Antigen: Guidance for Safely Stopping Nucleos(t)ide Analogues in Hepatitis B e Antigen-Positive Patients With Chronic Hepatitis B.
Hepatitis B Core-Related Antigen is a Biomarker for off-Treatment Relapse After Long-Term Nucleos(t)ide Analog Therapy in Patients with Chronic Hepatitis B.
EOT HBcrAg & EOT HBsAg The 3 out of 15 patients with HBsAg loss demonstrated a >1 log HBsAg reduction over median 33 (12-50) month follow-up and had a strong increase in HBV DNA (>4 x 105IU/mL) and >90-fold increase in HBcrAg at 4-8 weeks post EOT
EOT HBcrAg and HBV-RNA More frequently undetectable in patients who achieved HBsAg loss than in patients who did not HBcrAg: 75% vs. 42%; p = 0.12 HBV-RNA: 88% vs. 47%; p = 0.053
Predictors of hepatitis B surface antigen loss, relapse and retreatment after discontinuation of effective oral antiviral therapy in noncirrhotic HBeAg-negative chronic hepatitis B.
The role of hepatitis B virus core-related antigen in predicting hepatitis B virus relapse after cessation of entecavir in hepatitis B e antigen-negative patients.
Pilot Study Using Machine Learning to Identify Immune Profiles for the Prediction of Early Virological Relapse After Stopping Nucleos(t)ide Analogues in HBeAg-Negative CHB.
Cumulative rates of VR, CR or functional cure stratified by the biomarker(s) of interest were summarized in Figures 2a-c.
EOT HBcrAg & EOT HBV RNA HBcrAg >2 log U/mL & RNA positive: p=0.042 47% of patients with VR and 18% of patients without VR HR (either positive vs both negative): not significant EOT HBsAg, EOT HBcrAg & EOT HBV RNA HBsAg > 3 log IU/mL & HBcrAg >2 log U/mL & RNA positive: p=0.209 58% of patients with VR and 35% of patients without VR HR (any positive vs both negative): not significant
Cumulative rates of VR, CR or functional cure stratified by the biomarker(s) of interest were summarized in Figures 2a-c.
EOT HBcrAg & EOT HBV RNA HBcrAg >2 log U/mL & RNA positive: p=0.07 59% of patients with CR and 29% of patients without CR HR (either positive vs both negative): not significant EOT HBsAg, EOT HBcrAg & EOT HBV RNA HBsAg > 3 log IU/mL & HBcrAg >2 log U/mL & RNA positive: p=0.097 71% of patients with CR and 42% of patients without CR HR (any positive vs both negative): not significant
EOT HBV RNA HR (positive vs negative): Not significantly associated with clearance
Cumulative rates of VR, CR or functional cure stratified by the biomarker(s) of interest were summarized in Figures 2a-c.
EOT HBcrAg & EOT HBV RNA HBcrAg >2 log U/mL & RNA positive: p=0.009 0% of patients with HBsAg loss and 46% of patients without HBsAg loss HR (either positive vs both negative): not significant EOT HBsAg, EOT HBcrAg & EOT HBV RNA HBsAg > 3 log IU/mL & HBcrAg >2 log U/mL & RNA positive: p=0.003 0% of patients with HbsAg loss and 61% of patients without VR HR (any positive vs both negative): not significant
EOT HBcrAg Median HBcrAg at VR: 3.76 log U/mL Significantly higher than at EOT (p=0.005) EOT HBV RNA RNA remained undetected in all but 1 patient after VR
Serum hepatitis B virus RNA level is associated with biochemical relapse in patients with chronic hepatitis B infection who discontinue nucleos(t)ide analogue treatment.
Clinical Utility of SCALE-B to Predict Hepatitis B Virus Relapse, Hepatitis B Surface Antigen Loss After Antiviral Cessation in Asian Patients After 2-Year Follow-up.
Cumulative rates of VR, CR or functional cure stratified by the biomarker(s) of interest were summarized in Figures 2a-c.
Cumulative incidence of 3-year HBsAg loss: p<0.001 HBcrAg <2 log U/mL: 14.6% HBcrAg >2 log U/mL: 3.5% MV HR (per log U/mL): 0.718 (0.593-0.869) p=0.001 EOT HBsAg & EOT HBcrAg Among patients with HBsAg <1 log IU/mL, no additive value of HBcrAg (HR = 1.08, p=0.833) Among patients with HBsAg 10-100 IU/mL: MV HR (<2 vs >2 log U/mL): 3.397 p=0.001 Among patients with HBsAg >2 log IU/mL: MV HR (<2 vs >2 log U/mL): 3.702 p<0.001
Abbreviations: AUROC: area under region of curve; CR: clinical relapse; EOT: end of treatment; HBcrAg: hepatitis B core-related antigen; HBsAg: hepatitis B surface antigen; HBV RNA: hepatitis B virus ribonucleic acid; HR: hazard ratio; MV: multivariate; OR: odds ratio; UV: univariate; VR: virological relapse. Note #1: In the first column, mean/median levels of explored biomarkers for each cohort are provided where available.
a Cumulative rates of VR, CR or functional cure stratified by the biomarker(s) of interest were summarized in Figures 2a-c.
b Serum HBV DNA is quantified using the COBAS Taqman HBV test, with a lower limit of detection of 20 IU/mL. This study specifically made a distinction between patients with undetectable HBV DNA (48.5%) and patients with HBV DNA <20 IU/mL (43.8%), which was not performed in other studies.
Figure 2a. Cumulative rates of VR, CR or functional cure stratified by HBcrAg. Each row represents a single study. The bar charts demonstrate the cumulative rate, shown as %, of the specific outcome (VR, CR or HBsAg seroclearance) stratified by whether HBcrAg was above the specified cut-off level. For instance, in Fan R et al 2020B, the 4-year cumulative rate of CR was 39.5% and 7.3% when the end-of-treatment HBcrAg was >4 log and <4log, respectively (p<0.01). b. Cumulative rates of VR, CR or functional cure stratified by HBV RNA. Each row represents a single study. The bar charts demonstrate the cumulative rate, shown as %, of the specific outcome (VR, CR or HBsAg seroclearance) stratified by whether HBV RNA was above the specified cut-off level. For instance, in Fan R et al 2020A, the 4-year cumulative rate of VR was 37.0% and 8.0% when the end-of-treatment HBV RNA was >3 log and <3 log, respectively (no p value). c. Cumulative rates of VR, CR or functional cure stratified by HBcrAg + HBV RNA. Each row represents a single study. The bar charts demonstrate the cumulative rate, shown as %, of the specific outcome (VR, CR or HBsAg seroclearance) stratified by whether HBcrAg and/or HBV RNA was above the specified cut-off level. For instance, in Fan R et al 2020B, the 4-year cumulative rate of HBsAg seroclearance was 16.1% when the end-of-treatment HBcrAg was <4 log + HBV RNA <3 log, compared to 1.3% when HBcrAg was >4 log + HBV RNA >3 log (p<0.01).
Figure 2a. Cumulative rates of VR, CR or functional cure stratified by HBcrAg. Each row represents a single study. The bar charts demonstrate the cumulative rate, shown as %, of the specific outcome (VR, CR or HBsAg seroclearance) stratified by whether HBcrAg was above the specified cut-off level. For instance, in Fan R et al 2020B, the 4-year cumulative rate of CR was 39.5% and 7.3% when the end-of-treatment HBcrAg was >4 log and <4log, respectively (p<0.01). b. Cumulative rates of VR, CR or functional cure stratified by HBV RNA. Each row represents a single study. The bar charts demonstrate the cumulative rate, shown as %, of the specific outcome (VR, CR or HBsAg seroclearance) stratified by whether HBV RNA was above the specified cut-off level. For instance, in Fan R et al 2020A, the 4-year cumulative rate of VR was 37.0% and 8.0% when the end-of-treatment HBV RNA was >3 log and <3 log, respectively (no p value). c. Cumulative rates of VR, CR or functional cure stratified by HBcrAg + HBV RNA. Each row represents a single study. The bar charts demonstrate the cumulative rate, shown as %, of the specific outcome (VR, CR or HBsAg seroclearance) stratified by whether HBcrAg and/or HBV RNA was above the specified cut-off level. For instance, in Fan R et al 2020B, the 4-year cumulative rate of HBsAg seroclearance was 16.1% when the end-of-treatment HBcrAg was <4 log + HBV RNA <3 log, compared to 1.3% when HBcrAg was >4 log + HBV RNA >3 log (p<0.01).
Figure 2a. Cumulative rates of VR, CR or functional cure stratified by HBcrAg. Each row represents a single study. The bar charts demonstrate the cumulative rate, shown as %, of the specific outcome (VR, CR or HBsAg seroclearance) stratified by whether HBcrAg was above the specified cut-off level. For instance, in Fan R et al 2020B, the 4-year cumulative rate of CR was 39.5% and 7.3% when the end-of-treatment HBcrAg was >4 log and <4log, respectively (p<0.01). b. Cumulative rates of VR, CR or functional cure stratified by HBV RNA. Each row represents a single study. The bar charts demonstrate the cumulative rate, shown as %, of the specific outcome (VR, CR or HBsAg seroclearance) stratified by whether HBV RNA was above the specified cut-off level. For instance, in Fan R et al 2020A, the 4-year cumulative rate of VR was 37.0% and 8.0% when the end-of-treatment HBV RNA was >3 log and <3 log, respectively (no p value). c. Cumulative rates of VR, CR or functional cure stratified by HBcrAg + HBV RNA. Each row represents a single study. The bar charts demonstrate the cumulative rate, shown as %, of the specific outcome (VR, CR or HBsAg seroclearance) stratified by whether HBcrAg and/or HBV RNA was above the specified cut-off level. For instance, in Fan R et al 2020B, the 4-year cumulative rate of HBsAg seroclearance was 16.1% when the end-of-treatment HBcrAg was <4 log + HBV RNA <3 log, compared to 1.3% when HBcrAg was >4 log + HBV RNA >3 log (p<0.01).
Combining Hepatitis B Virus RNA and Hepatitis B Core-Related Antigen: Guidance for Safely Stopping Nucleos(t)ide Analogues in Hepatitis B e Antigen-Positive Patients With Chronic Hepatitis B.
demonstrated a multivariate HR of 5.70 (1.37-23.67; p=0.017) between patients with EOT HBcrAg >4 log U/mL and <4 log U/mL, with 5 other studies encompassing HBeAg positive, HBeAg negative and combined populations also producing significant multivariate HRs at EOT HBcrAg cut-off levels ranging from 2 to 4 log U/mL.
Combining Hepatitis B Virus RNA and Hepatitis B Core-Related Antigen: Guidance for Safely Stopping Nucleos(t)ide Analogues in Hepatitis B e Antigen-Positive Patients With Chronic Hepatitis B.
Hepatitis B Core-Related Antigen is a Biomarker for off-Treatment Relapse After Long-Term Nucleos(t)ide Analog Therapy in Patients with Chronic Hepatitis B.
who reported that lower EOT HBcrAg levels were significantly associated with lower rates of CR with multivariate OR 1.29 per log U/mL (1.08-1.54, p=0.005). Eight studies reported the cumulative rates of CR stratified by various cut-off levels of HBcrAg at baseline or EOT (Figure 2a). Once again, both Tseng et al
The role of hepatitis B virus core-related antigen in predicting hepatitis B virus relapse after cessation of entecavir in hepatitis B e antigen-negative patients.
Predictors of hepatitis B surface antigen loss, relapse and retreatment after discontinuation of effective oral antiviral therapy in noncirrhotic HBeAg-negative chronic hepatitis B.
found that EOT HBcrAg was significantly associated with retreatment (a composite endpoint suggestive of CR) in a HBeAg negative population, where the 2-year retreatment rates were 45% and 17% in patients with EOT HBcrAg >2 and <2 log U/mL respectively, alongside a multivariate HR of 3.64 (1.23-10.75; p=0.019) regarding retreatment with this cut-off level.
Overall, HBeAg positive populations demonstrate higher mean/median EOT HBcrAg levels, and the EOT cut-off of HBcrAg of 4 log U/mL is a reliable predictor of both VR and CR. HBeAg negative and combined populations necessitate a lower cut-off level, ranging from 2 to 3.3 log U/mL in the included studies (noting that the validated lower limit of detection is 3 log U/mL). The risk of VR/CR in populations that have a mean/median EOT HBcrAg level at or below 3 log U/mL may be better distinguished by a baseline HBcrAg cut-off of 4 log U/mL.
HBcrAg and functional cure
14 studies evaluated the association between HBcrAg and rates of HBsAg loss, with most studies not returning significant results (Table 3). Five studies reported the cumulative rates of functional cure stratified by baseline or EOT HBcrAg levels (Figure 2a). Only 4 of 12 patients in Liao et al’s study
Hepatitis B Core-Related Antigen is a Biomarker for off-Treatment Relapse After Long-Term Nucleos(t)ide Analog Therapy in Patients with Chronic Hepatitis B.
found a near-significant difference in 48-week HBsAg seroclearance rates, 5.9% vs 0% in patients with EOT HBcrAg <3 log U/mL and >3 log U/mL respectively (p=0.062). Interestingly, Honer Zu Siederdissen et al
demonstrated that HBsAg reduction and seroclearance was associated with the degree of virological relapse. The extent of increase in HBcrAg (in parallel with HBV DNA rebound) at weeks 4-8 post treatment cessation correlated with HBsAg decline and were followed by HBsAg loss in 3 of 15 patients. Carey et al
Pregenomic HBV RNA and Hepatitis B Core-Related Antigen Predict Outcomes in Hepatitis B e Antigen-Negative Chronic Hepatitis B Patients Suppressed on Nucleos(T)ide Analogue Therapy.
also found that a steeper HBsAg decline post-treatment correlated with lower baseline HBcrAg levels rather than EOT levels, observing transiently resolving elevations of HBcrAg after NA cessation. Recently, a multicentre study
comprising 1216 patients demonstrated that EOT HBcrAg was significantly associated with the probability of HBsAg loss (multivariate HR per log U/mL 0.729, 0.603-0.882, p=0.001).
HBV RNA and partial cure
8 studies explored the association between HBV RNA levels and rates of VR (Table 3). Five studies reported the cumulative rates of VR at various observation periods stratified by EOT RNA levels (Figure 2b). For instance, Kaewdech et al
initially reported significantly different 48-week VR rates of 50% and 72% in patients with EOT HBV RNA <2 and >2 log U/mL respectively (p=0.048), yet the effect of HBV RNA on both VR and CR was found to be statistically insignificant in their subsequent publication with longer follow-up (median 35.5 months) when adjusted for SCALE-B strata.
Clinical Utility of SCALE-B to Predict Hepatitis B Virus Relapse, Hepatitis B Surface Antigen Loss After Antiviral Cessation in Asian Patients After 2-Year Follow-up.
demonstrated a multivariate HR of 2.96 (1.78-4.93; p=0.001) between combined HBeAg positive and negative patients at RNA cut-off level of 1.65 log U/mL. Papatheodoridi et al
found that detectability of EOT HBV RNA was significantly associated with VR in their HBeAg negative population, quoting a HR of 3.20 (1.10-9.32 p=0.033), as was detectability of HBV RNA detection at one month post-EOT (HR 3.23, 1.57-6.67, p=0.001). Similarly, Xie et al
demonstrated a multivariate OR of 3.453 (1.387-8.597; p=0.008) between patients with positive vs negative RNA detection in their HBeAg positive population. Furthermore, Lai et al,
who demonstrated high VR rates in patients with undetectable cccDNA and RNA, found that all but one patient continued to exhibit undetectable HBV RNA levels after relapse.
8 out of 10 relevant studies affirmed a significant association between HBV RNA and CR (Table 3). Six studies reported the cumulative rates of CR at various observation periods stratified by EOT RNA levels (Figure 2b). For instance, Fan et al
Combining Hepatitis B Virus RNA and Hepatitis B Core-Related Antigen: Guidance for Safely Stopping Nucleos(t)ide Analogues in Hepatitis B e Antigen-Positive Patients With Chronic Hepatitis B.
demonstrated a multivariate HR of 3.58 (1.26-10.14; p=0.017) between HBeAg positive patients with EOT RNA >3 and <3 log U/mL, alongside significantly different 4-year CR rates of 12.9% vs 40.1% according to that cut-off (p=0.004). Liu et al
also reported significantly different 2-year CR rates of 17.5% vs 38.3% in patients who were HBV RNA negative and positive respectively (combined HBeAg positive and negative population). Papatheodoridi et al
Pregenomic HBV RNA and Hepatitis B Core-Related Antigen Predict Outcomes in Hepatitis B e Antigen-Negative Chronic Hepatitis B Patients Suppressed on Nucleos(T)ide Analogue Therapy.
demonstrated transient elevations in HBV RNA after NA cessation and found that 3/4 patients who demonstrated CR had RNA levels >1.65 log U/mL (75% sensitivity, 100% specificity, 100% PPV).
In conclusion, HBV RNA demonstrates utility in predicting both VR and CR in the majority of publications. There is less of a distinction in mean EOT RNA levels and preferred RNA cut-offs between HBeAg positive, negative and combined populations when compared to the corresponding HBcrAg findings. This is in part due to lack of standardization of RNA assays between different study groups.
HBV RNA and functional cure
The relationship between HBV RNA and HBsAg loss was not reported in most studies. Only two studies reported the cumulative rates of functional cure stratified by EOT RNA levels (Figure 2b). Kaewdech et al
found that EOT HBV RNA was more frequently undetectable in patients who achieved HBsAg loss than in patients who did not (88% vs 47%, p=0.053). Xia et al
Serum hepatitis B virus RNA level is associated with biochemical relapse in patients with chronic hepatitis B infection who discontinue nucleos(t)ide analogue treatment.
also found that cumulative incidence of 6-year HBsAg clearance rates was 30.9% versus 1.6% in patients with EOT HBV RNA <3 versus >3 log U/mL respectively (p=0.007).
derived the SCALE-B score for CR, consisting of the 5 predictors: EOT HBsAg, EOT HBcrAg, age, ALT and use of TDF. Stratifying patient risk, they demonstrated a significant difference in 3-year CR rates of 86.2%, 55.6% and 17.2% in the high-, intermediate- and low-risk subgroups respectively (p=0.0001). Furthermore, all patients achieving functional cure were drawn from the low-risk subgroup and demonstrated EOT HBsAg levels <2 log IU/mL and EOT HBcrAg levels below 3 log U/mL. Later, Papatheodoridi et al
Predictors of hepatitis B surface antigen loss, relapse and retreatment after discontinuation of effective oral antiviral therapy in noncirrhotic HBeAg-negative chronic hepatitis B.
also demonstrated a significant multivariate HR of 0.93 (0.87-0.98; p=0.012) per 1000 points increase in the SCALE-B score for HBsAg seroclearance. Lower SCALE-B score was again associated higher rates of partial and functional cure in Kaewdech et al’s Thailand study
Clinical Utility of SCALE-B to Predict Hepatitis B Virus Relapse, Hepatitis B Surface Antigen Loss After Antiviral Cessation in Asian Patients After 2-Year Follow-up.
found that combining HBV RNA status and EOT HBsAg level was superior to EOT HBsAg level alone in predicting partial cure, with a 2-year VR rate of 10% in patients with EOT HBsAg <2 log IU/mL and EOT HBV RNA negativity. Seto et al
similarly demonstrated that a combination of undetectable EOT HBV RNA level and HBsAg <10 IU/mL was associated with a 1-year VR rate of 9.1%. Lastly, Xie et al
found that the combination of EOT HBsAg <100 IU/mL and EOT HBV RNA undetectability had the highest AUROC for VR or partial cure, with an AUROC of 0.698 that was superior to other singular and combined parameters.
Combining HBcrAg with HBV RNA
The cumulative rates of VR, CR or functional cure stratified by a combination of EOT HBcrAg and EOT HBV RNA were reported in 4 studies (Figure 2c). Xie et al
Combining Hepatitis B Virus RNA and Hepatitis B Core-Related Antigen: Guidance for Safely Stopping Nucleos(t)ide Analogues in Hepatitis B e Antigen-Positive Patients With Chronic Hepatitis B.
demonstrated that the combination of EOT HBcrAg and EOT HBV RNA was most predictive of subsequent CR with an AUROC of 0.742 (0.64-0.84, p < 0.001), indeed superior to qHBsAg alone with an AUROC of 0.609 (0.49-0.73, p=0.089). In Papatheodorididi et al’s study
, although more patients who did not develop VR/CR or achieved HBsAg seroclearance had undetectable HBcrAg and HBV RNA, a combination of detectable HBV RNA and/or HBcrAg at EOT was not significantly associated with partial or functional cure.
Genetic variations in the CXCR5 gene decrease the risk of clinical relapse after discontinuation of nucleos(t)ide analogue therapy in patients with chronic hepatitis B.
have been explored in the context of HBV treatment discontinuation, but the specificity and clinical significance of these findings remains uncertain. For example, Wu et al
Genetic variations in the CXCR5 gene decrease the risk of clinical relapse after discontinuation of nucleos(t)ide analogue therapy in patients with chronic hepatitis B.
found that the rs676925 ‘GC’ genotype of the CXCR5 gene was associated with decreased risk of CR, but failed to demonstrate a corresponding difference in percentage of CXCR5-positivity or expression of CXCL13 ligand between genotype groups. With respect to whole genome gene expression analysis, Kranidioti et al
Immunological biomarkers as indicators for outcome after discontinuation of nucleos(t)ide analogue therapy in patients with HBeAg-negative chronic hepatitis B.
Pilot Study Using Machine Learning to Identify Immune Profiles for the Prediction of Early Virological Relapse After Stopping Nucleos(t)ide Analogues in HBeAg-Negative CHB.
Pilot Study Using Machine Learning to Identify Immune Profiles for the Prediction of Early Virological Relapse After Stopping Nucleos(t)ide Analogues in HBeAg-Negative CHB.
found that levels of almost all SIMs increased after treatment cessation, significantly so for TNF, IL-12p70 and IL-10 at week 4-post EOT and for TNF and CXCL10/IP10 at week 8-post EOT. The increase in SIM levels was associated with VR and HBcrAg rebound, and subsequent decline and loss of HBsAg. Papatheodoridi et al reported that higher EOT IP10 levels at 1 month-post EOT identified patients more likely to achieve HBsAg loss, without mention of whether they underwent transient VR and CR first.
Predictors of hepatitis B surface antigen loss, relapse and retreatment after discontinuation of effective oral antiviral therapy in noncirrhotic HBeAg-negative chronic hepatitis B.
Increased NK Cell Function After Cessation of Long-Term Nucleos(t)ide Analogue Treatment in Chronic Hepatitis B Is Associated With Liver Damage and HBsAg Loss.
studied changes in the NK cell response in HBV patients following treatment cessation. Stopping NA treatment significantly boosted CD56dim NK cell natural cytotoxicity responses, correlating with increased NK cell functional responses and ALT levels at weeks 8 and 12 post-EOT. The subgroup of patients who cleared HBsAg experienced higher ALT levels at week 12 post-EOT and demonstrated higher expression of CD38 on CD56dim NK cells, with increased NK cell functionality. Furthermore, Hall et al
reported that severe hepatitis flares were associated with upregulation of the innate immune response, demonstrated by increased activity of TLR2-8 and TLR9 signalling in PBMCs and upregulation of TLR2 and TREM-1 receptor expression on NK cells at peak flare, with no such change from baseline in patients without flares. There was no significant correlation between TLR signalling activity and HBsAg decline or clearance.
HBV-Specific Immune Markers
The association between EOT anti-HBc levels and rates of partial cure after treatment cessation appear unclear (Table 4). Chi et al
Serum Level of Antibodies Against Hepatitis B Core Protein Is Associated With Clinical Relapse After Discontinuation of Nucleos(t)ide Analogue Therapy.
found that anti-HBc was not significantly associated with VR, but reported a multivariate HR of 0.31 (0.15-0.65; p=0.002) in predicting 4-year CR. Patients with anti-HBc >3 log IU/mL demonstrated 4-year CR rates of 21%, while those with anti-HBc <2 log IU/mL demonstrated 4-year CR rates of 85%. Similar studies did not find a significant association between EOT anti-HBc levels and VR or CR.
Genetic variations in the CXCR5 gene decrease the risk of clinical relapse after discontinuation of nucleos(t)ide analogue therapy in patients with chronic hepatitis B.
CXCR5 (rs676925) GC vs CC genotype: MV OR: 0.25 (0.07-0.95) p=0.0042 CXCR 5 (rs676925) non-CC vs CC genotype:MV HR: 0.34 (0.12-0.96) p=0.041 No difference in # or MFI of CXC5-positive cells or plasma CXCL13 levels between genotype groups
Immunological biomarkers as indicators for outcome after discontinuation of nucleos(t)ide analogue therapy in patients with HBeAg-negative chronic hepatitis B.
Predictors of hepatitis B surface antigen loss, relapse and retreatment after discontinuation of effective oral antiviral therapy in noncirrhotic HBeAg-negative chronic hepatitis B.
EOT IP10 levels HR per 10 pg/ml: 1.03 (0.99-1.07) p=0.01 1 month post-EOT IP10 levels HR per 10 pg/ml: 1.10 (1.02-1.19) p=0.01 Compared to EOT, IP10 levels were higher at months 2 (p<0.001) and 3 (p=0.024), similar at month 6 (p=0.195) and lower at months 9 and 12 (p<0.004)
Pilot Study Using Machine Learning to Identify Immune Profiles for the Prediction of Early Virological Relapse After Stopping Nucleos(t)ide Analogues in HBeAg-Negative CHB.
Serum Level of Antibodies Against Hepatitis B Core Protein Is Associated With Clinical Relapse After Discontinuation of Nucleos(t)ide Analogue Therapy.
Patients achieving HBsAg loss demonstrated a T cell phenotype with lowly expressed PD-1 and KLRG1, and an increase in expression of Ki-67 and CD38 at week 12 post-EOT. Baseline HBsAg was positively correlated with PD-1+ CD8+ T cells, and fold decline of HBsAg at month 12 post-EOT was associated with frequency of Ki-67+ CD38+ T cells at week 12 post-EOT.
The HBV-specific T cell response, mainly targeting core and polymerase proteins, was at least not superior in patients who flared Patients who did not flare demonstrated increased expression of the most differentially expressed gene, PD-1 (p=0.009) in CD8+ T cells
Patients remaining off-therapy had functional HBV-specific CD8+ T cell responses against epitopes from multiple HBV proteins, (68% vs 20%, p=0.048 for IFNγ production and 77% vs 40% p=0.099 for CD107a expression) The percentage of degranulating CD8+ T cells (CD107a) was higher at EOT and week 12 post-EOT in patients remaining off therapy (p=0.039 and p=0.0093) when stimulated with core proteins. The percentage of polyfunctional core specific CD8+ T cells (co-expressing IFNγ and TNFa) was higher among patients remaining off-therapy (p=0.031) and this increase persisted for more than a year post-EOT (p=0.01).
Increased NK Cell Function After Cessation of Long-Term Nucleos(t)ide Analogue Treatment in Chronic Hepatitis B Is Associated With Liver Damage and HBsAg Loss.
Patients achieving HBsAg loss experienced higher ALT levels and higher CD56dimNK cell expression of CD38 at 12 weeks post-EOT Patients achieving HBsAg loss experienced elevated responses upon K562 stimulation at 12 weeks post-EOT, particularly CD56dimNK cell IFNγ, TNF and GM-CSF responses
Hepatitis flares were associated with significant increases in TNF, IL-6 and IL-8 cytokine production after PBMC TLR signalling with stimulation from TLR ligands, whereas patients who did not flare demonstrated no significant changes to baseline Hepatitis flare was associated with increased expression of TREM-1 and TLR2 on NK-bright & NK-T cells, and increased expression of TLR2 alone on NK-dim cells, whereas patients who did not flare demonstrated no significant changes to baseline
Abbreviations: AUROC: area under region of curve; CR: clinical relapse; EOT: end of treatment; HBsAg: hepatitis B surface antigen; HR: hazard ratio; MV: multivariate; OR: odds ratio; UV: univariate; VR: virological relapse.
Several studies examined HBV-specific and global T cell populations in patients undergoing treatment cessation. We previously reported that frequencies of in vitro-expanded HBV-specific T cells both during and after discontinuation of therapy were consistently and significantly higher in patients without hepatic flares post treatment-cessation, in particular the responses against core and polymerase proteins.
Patients who did not develop a biochemical flare upon treatment cessation demonstrated increased gene expression encoding for PD-1 in CD8+ T cells. Garcia-Lopez et al
found that patients who did not require retreatment demonstrated a higher percentage of degranulating CD8+ T cells (CD107a) in addition to polyfunctional CD8+ T cells co-producing IFNγ/TNFa. HBV-specific T cell responses did not augment following treatment withdrawal, and were not associated with the development of clinically relevant flares or HBsAg loss. Conversely, Rinker et al
found a significant increase in HBV core-specific multifunctional T cell responses at 8 and 12 weeks post-EOT, while no significant changes were observed following stimulation with polymerase- or envelope-specific peptides. Patients experiencing functional cure demonstrated a less exhausted and more activated T cell phenotype, with increases in Ki-67 and CD38 expression at week-12 post-EOT. HBV-specific CD4+ and CD8+ T cell responses were also significantly enhanced by PD-L1 blockade at weeks 4 and 8 post-EOT. The findings from studies of HBV-specific immune markers are summarized in Table 4.
Discussion
Rates of functional cure remain low in CHB patients who remain on antiviral therapy, as evidenced by an 8-year cumulative incidence of 1.69% in ETV-treated patients and 1.34% in TDF-treated patients in a recent, large multi-ethnic study.
Treatment cessation in CHB patients has emerged as a possible strategy to achieve functional cure in select patients, but remains a controversial approach given concerns around safety of treatment withdrawal. Overall, the included studies report wide variations in off-therapy outcomes, owing to the heterogeneity of patient populations and stopping criteria. Patient factors, such as ethnicity,
Off-Therapy Response After Nucleos(t)ide Analogue Withdrawal in Patients With Chronic Hepatitis B: An International, Multicenter, Multiethnic Cohort (RETRACT-B Study).
Incidence and predictors of hepatitis B surface antigen seroclearance after cessation of nucleos(t)ide analogue therapy in hepatitis B e antigen-negative chronic hepatitis B.
The Incidence of Hepatitis B Surface Antigen Loss Between Hepatitis B E Antigen-Negative Noncirrhotic Patients Who Discontinued or Continued Entecavir Therapy.
in achieving functional cure after treatment discontinuation. However, there is interest in leveraging additional factors to forcast off-therapy HBsAg loss with more certainty. Given the limitations of current treatment strategies in CHB, the aim of this review was to evaluate how the data from treatment discontinuation studies could be applied to the functional cure program to better predict treatment response and ultimately HBsAg loss.
Reliable biomarkers are essential to identify individuals where NA therapy can be discontinued safely and functional cure achieved. It is well established that the correlation between serum HBsAg level and cccDNA exists only in the HBeAg-positive phase of CHB infection.
Serum hepatitis B surface antigen and hepatitis B e antigen titers: disease phase influences correlation with viral load and intrahepatic hepatitis B virus markers.
Following HBeAg seroconversion, there is continued production of HBsAg, partly from integrated HBV DNA in hepatocytes, and in fact the fraction of integrated HBV DNA as a fraction of total intrahepatic HBV DNA is significantly higher in HBeAg-negative patients compared to HBeAg-positive patients.
To this end, HBcrAg, which represents the combined antigenic reactivity of e-antigen, core antigen and defective core-related protein p22cr, has been shown to more strongly correlate with cccDNA quantity in both patients that are treatment naïve
In situations where serum HBV DNA has become undetectable, the presence of HBcrAg indicates continued secretion of viral-end products. Conversely, serum HBV RNA reflects the amount of virion-like encapsidated particles in which pgRNA was non or partially reverse transcribed.
Undetectable HBV RNA despite the persistence of cccDNA in most patients with HBsAg loss after treatment cessation may demonstrate a functional reduction in cccDNA transcriptional activity.
Our review suggests that EOT HBV RNA and EOT HBcrAg are both strong predictors for sustained partial cure across the included studies, but both markers also have their limitations. The decline in HBcrAg across treatment may result in an EOT level that falls below the accepted LLoD, especially among HBeAg-negative patients, and as a result this assay may not be able to reflect very low but persistent levels of cccDNA. Similarly, Liu et al reported that the lack of highly sensitive methods of detection for HBV RNA may result in a low threshold for undetectable RNA levels.
Standardization of cut-offs of viral markers (especially HBV RNA) in terms of method of detection and quantification is of paramount importance to allow fair comparison between various settings. While previous studies have failed to find a strong correlation between either of these biomarkers and functional cure, the CREATE study group
recently pooled multiple large-scale cohorts to conclude that EOT HBcrAg, in isolation or in combination with EOT HBsAg was significantly associated with HBsAg seroclearance. Various combinations of viral markers have also shown potential in predicting off-therapy responses, but the evidence behind SCALE-B score is the most substantial, having been validated for clinical relapse, retreatment and HBsAg loss.
We propose an algorithm, stratified by HBeAg status at NA initiation, based on EOT qHBsAg, in combination with HBcrAg and HBV RNA to decide whether NA should be discontinued in CHB patients (Figure 3). In general, NA should be continued if EOT qHBsAg is ≥2 log. NA cessation can be considered when the EOT qHBsAg <2 log in combination with HBV RNA <3 log or HBcrAg <4 log for initially HBeAg-positive patients. As the sensitivity of HBcrAg in HBeAg-negative patients is lower, undetectable HBcrAg should not be over interpreted in this scenario; NA cessation could only be considered when HBV RNA is <2 log or undetectable.
Figure 3Proposed algorithm to decide whether NA should be discontinued based on qHBsAg, HBcrAg, and HBV RNA stratified by HBeAg status.
The hallmark of CD8+ T cell exhaustion is loss of proliferative capacity, cytotoxicity and cytokine production, which is enhanced through the upregulation of inhibitory pathways with continued antigen and viral load exposure.
Regarding innate immunity, NK cells appear to act in inverse correlation to T cells. Their inhibition of CD4+ T cells is likely necessary to limit persistent T cell activation, yet their reversion to a quiescent phenotype is reflective of restoration of HBV-specific T cell function.
Natural killer cell phenotype modulation and natural killer/T-cell interplay in nucleos(t)ide analogue-treated hepatitis e antigen-negative patients with chronic hepatitis B.
The adaptive humoral response is driven by the role of B cells, which are activated by T-cell dependent and independent pathways to produce disease-specific antibodies. In CHB, HBsAg specific B cells demonstrate defective antibody production and an accumulation of atypical memory B cells with increased expression of inhibitory receptors.
While all exposed individuals mount an antibody response to HBcAg, higher anti-HBc levels in CHB infection may represent a larger number of activated B cells, which in turn modulate CD4+ and CD8+ T cell activity, and augment naïve T-helper cells through their highly potent antigen-presenting function.
Milich DR, Chen M, Schödel F, Peterson DL, Jones JE, Hughes JL. Role of B cells in antigen presentation of the hepatitis B core. Proceedings of the National Academy of Sciences of the United States of America 1997;94:14648-14653.
While the high levels of antigen expression in hepatocytes result in T cell exhaustion and deletion, evidence suggests that long-term antiviral therapy only results in a partial reconstitution of the T cell response. Following in vitro expansion experiments, the HBV-specific polyfunctional T cell response of successfully treated CHB patients (with HBsAg loss) was comparable to patients with spontaneously resolving acute HBV infection. In contrast, in NA-treated CHB patients who were HBV DNA negative but remain HBsAg positive, T-cell responses were notably weaker, compounded by the slow decline in HBsAg on long-term NA.
Therefore, there is growing interest in treatment interruption or discontinuation as a strategy to boost the host immune response to facilitate functional cure.
Although studies that explore SIMs in the context of NA cessation demonstrate low replicability potential and lack of disease specificity, it was observed that VR preceded cytokine upregulation and ALT flare, and subsequent HBsAg decline.
Hepatitis B surface antigen (HBsAg) decrease and serum interferon-inducible protein-10 levels as predictive markers for HBsAg loss during treatment with nucleoside/nucleotide analogues.
This suggests that a transient virological rebound, with or without subsequent clinical relapse, may assist in the immune-mediated killing of infected hepatocytes and non-cytolytic degradation of cccDNA. This is in contrast to previous studies that have shown that sustained off-therapy response is associated with higher chance of functional cure,
Incidence and predictors of hepatitis B surface antigen seroclearance after cessation of nucleos(t)ide analogue therapy in hepatitis B e antigen-negative chronic hepatitis B.
The Incidence of Hepatitis B Surface Antigen Loss Between Hepatitis B E Antigen-Negative Noncirrhotic Patients Who Discontinued or Continued Entecavir Therapy.
and the role of VR in achieving HBsAg loss remains controversial. We previously demonstrated that increased frequencies of HBV core and polymerase-specific T cells were a promising immunological biomarker for patients who did not experience hepatic flares following treatment cessation, and that hepatic flares were in fact not driven by HBV-specific T cell responses.
Recently, a logistic regression model predicting on-treatment presence of functional HBV-specific CD8+ T cell response has demonstrated a positive correlation with off-treatment HBsAg decline and loss.
Model to predict on-treatment restoration of functional HBV-specific CD8(+) cell response foresees off-treatment HBV control in eAg-negative chronic hepatitis B.
Furthermore, treatment cessation itself triggers a new immunological environment that has been shown to increase frequency and functionality of HBV core-specific T cell responses in patients achieving functional cure.
and clinical studies discussed above, a restored immune response against HBV appears to be the prerequisite for a de novo response against HBV during chronic infection. The reappearance of HBV replication after stopping long-term NA treatment could in fact be the necessary trigger for the immune response
A delicate balance exists between the potential immunological benefits of NA discontinuation (i.e., accelerated rates of HBsAg decline/clearance) and the risk of excessive hepatocyte damage and resultant liver failure. To this end, transient VR in absence of a serious clinical flare should be viewed differently from a sustained rise in viraemia levels off-therapy, but this has been poorly delineated in NA discontinuation studies to date, which mostly define VR by a single timepoint.
Patients achieving functional cure following NA discontinuation demonstrate two important events: firstly, a reduction in viral antigen level, and second, evidence of immune recovery. Therefore, any novel approaches intended to enhance functional cure should follow this rule by combining virus-directing agents with immunomodulators. There are no preclinical/clinical studies to date which have evaluated this sequence using novel agents. This combination effect was found to be effective in mice given siRNA (viral antigen knockdown) followed by therapeutic vaccine (composed of recombinant HBV protein), whereby antigen load shift was induced to end the immune tolerance.
The safety and feasibility of NA discontinuation as a strategy for immune recovery should be guided by EOT antigenic loads. In selected patient groups; characterised by low HBsAg levels, low EOT HBcrAg and/or HBV RNA; NA discontinuation could be utilised in conjunction with virus-directed agents in order to achieve functional cure without risk of severe CR. However, the safety of this approach could not be over-emphasized, especially after the incidence of a subacute liver failure case necessitating liver transplantation in the REEF-2 study (patient in the placebo arm was continued on TDF for 48 weeks which was then stopped).
Across all included studies, there was a maximum of 15 decompensation events mentioned in 6 studies, leading to 2 liver-related deaths (taking into consideration overlapping study cohorts). While the consensus between studies is that retreatment almost always leads to re-compensation and renewed viral control, there is still a possibility of hepatic decompensation and its sequelae, such as transplantation and death. This was highlighted by a recent meta-analysis which showed that severe hepatitis flares or decompensation would occur in 1.21% and liver transplantation or death was observed in 0.37% following NA discontinuation.
Regardless of whether NA discontinuation is used as part of a novel combination therapy, close monitoring is essential and standardization of retreatment criteria will be inevitable to minimize the associated risks.
How can viral markers help to predict response to novel agents?
The results of our review showed that the viral biomarkers that act as a surrogate for transcriptionally active cccDNA are helpful to predict off-therapy partial cure and, to some extent, functional cure. In addition, the timing of assessment has implications on the outcome, and the most commonly used and practical timepoint has been EOT (i.e., end of NA therapy). Recent data suggest that early on-treatment profiles of HBcrAg and HBV RNA can help to identify future responders (HBsAg seroclearance or <2 logs) as early as week-4 of NA therapy.
Early biomarker response suggests effective restoration of antiviral immunity, and potentially identifies those likely to achieve HBsAg reduction or seroclearance following treatment. Those with high baseline viral markers or poor viral biomarker response on treatment with novel agents should be continued on NA.
It remains to be determined whether novel agents inducing viral antigen reduction and passive restoration of the immune response will lead to the same sustainable HBsAg seroclearance as observed following long-term NA. Another unanswered question is whether differentiating the HBsAg source (cccDNA vs integrated DNA) would help to predict risk of severe flare following NA discontinuation/novel treatment strategy. Since only hepatocytes containing transcriptionally active cccDNA have the potential to replicate virus and become susceptible to immune attack upon NA discontinuation or immune modulation, those with HBsAg predominantly from integrated DNA have a theoretically lower risk of severe VR or CR if NAs are discontinued or immunomodulators introduced.
Immune assessment – a practical perspective
Ideally, the demonstration of a multi-faceted, poly-cellular immune response together with an assessment of an appropriate panel of inflammatory SIMs would be needed to prove immune restoration. However, as previously discussed, SIMs are heterogenous, non-specific and so far inconclusive. Moreover, we lack robust and reproducible assays to predict pro-inflammatory cytokine production with novel therapeutic approaches. To allow a more accessible and reproducible assessment, we propose that HBV-specific T cells should be the immune marker of choice for predicting functional cure on antiviral treatment. The frequency of these T cells, the level of PD1 expression, and functionality (e.g., CD107a expression, IFNγ production on CD8+ T cells) are relatively specific readouts and efforts are underway to standardize the experimental assays, as well as implement them in all novel clinical trials moving forward.
For patients receiving immunomodulators, one needs to differentiate responses as being target engagement only or reflecting recovery of the immune response. Ideally, the paired assessment of intra-hepatic HBV-specific T cells in the clinical trial setting would be valuable to inform whether peripheral blood T cell responses are sufficiently informative.
The limitations regarding the clinical utility of viral markers are the recognised shortfalls in both sensitivity and standardisation. HBcrAg was measured by the Chemiluminscent Enzyme Immunoassay system (Fujirebio, Inc, Tokyo, Japan) in all studies. While the automated estimation range is quoted as 2 to 7 log U/mL, the validated lower limit of detection is 3 log U/mL. As a result, readings between 2 to 3 log U/mL are not reliable and many studies state that a large proportion of patients return undetectable readings. In the absence of a unified standard, a range of different HBV RNA assays, platforms and lower limits of detection have been adopted in the included studies (Supplementary Table 3). Moreover, most findings from immune marker studies are yet to be comprehensively validated in further independent samples to affirm reproducibility of results. Furthermore, there is also a recognised distinction between in vitro expansion and ex vivo conditions in the generation of immunological data. Unfortunately, we were unable to perform a meta-analysis of the novel markers presented in this review due to the heterogeneity of the data. There was variation in the cut-offs used for viral markers and the timepoints at which VR, CR and HBsAg loss were measured. Furthermore, many of these studies had overlapping (but not identical) cohorts, which would have disproportionately skewed the results of a meta-analysis.
Conclusion
Treatment discontinuation has emerged as a valid therapeutic option to maintain partial cure, and has also been associated with higher rates of functional cure compared to patients who remain on long-term NA therapy when trialled in select populations. Nonetheless, treatment discontinuation remains a blunt tool lacking both precision and certainty as to which patients will safely achieve functional cure. The findings of our systematic review demonstrate that HBV RNA and HBcrAg are synergistic to traditional markers, including qHBsAg, in predicting off-therapy VR and CR. Early changes in these parameters with novel therapies should be explored with regard to clinical outcomes. Evidently, the most useful immune markers consist of HBV-specific T cell responses, and these should be assessed in the correct context with accessible and reproducible assays. The achievement of partial cure should be regarded as an important step towards functional cure, which remains the therapeutic goal of novel agents currently under investigation.
Both virus-targeted and immune modulatory agents (where NA discontinuation can be considered an immunomodulatory strategy) are likely to be required to achieve functional cure, while the best sequence or combination approach needs to be explored further, drawing on the data from NA discontinuation studies.
Acknowledgements
None.
Appendix A. Supplementary data
The following are the supplementary data to this article:
00051-4&id=ga1.jpg){kind=link}
00051-4&id=gr1.jpg){kind=link}
00051-4&id=gr2a.jpg){kind=link}
00051-4&id=gr2b.jpg){kind=link}
00051-4&id=gr2c.jpg){kind=link}
00051-4&id=gr3.jpg){kind=link}