Phase-dependent interpretation of quantitative hepatitis B surface antigen and hepatocellular carcinoma risk in chronic hepatitis B infection

Abstract

Quantitative hepatitis B surface antigen (qHBsAg) is one of the known and commonly utilized blood tests that shows the total hepatitis B virus (HBV) antigen production from 2 distinct sources: (1) Transcriptionally active covalently closed circular DNA; and (2) The HBV DNA that is already integrated to the host genome. Its use has resulted in inconsistent results across patients with different stages of hepatitis B and therefore has limited its clinical utility. For this reason, this marker is yet to be adopted on a larger scale for routine hepatocellular carcinoma (HCC) cancer prediction. Emerging evidence suggests that the prognostic meaning of qHBsAg is heavily impacted by the phase of HBV infection and the underlying source of the antigen presence. For instance, in patients with hepatitis B e antigen+ chronic infection, elevated qHBsAg levels do indicate a state of active covalently closed circular DNA transcription and active viral replication, and higher HBsAg levels (> 10000 IU/mL) are paradoxically associated with delayed HCC development. In contrast, in patients with negative hepatitis B e antigen and a low viral load (< 2000 IU/mL), high qHBsAg (> 1000 IU/mL) suggests replication-independent HBsAg production from the integrated HBV DNA and is strongly associated with increased risk of HCC with an adjusted hazard ratio of 13.7 (95%CI: 4.8-39.3). This minireview analyzes the current data and presents a phase-dependent framework for interpreting qHBsAg in untreated chronic hepatitis B virus patients. We also discuss how it can be used in clinical practice as a complementary test to stratify the risk of developing HCC in those patients.

Key Words: Quantitative hepatitis B surface antigen; Hepatitis B; Hepatocellular carcinoma; Disease phase; Hepatitis B virus integration; Risk stratification; Phase-dependent interpretation

Core Tip: Quantitative hepatitis B surface antigen can serve as a complementary predictor of development of hepatocellular carcinoma among chronic hepatitis B patients. However, its interpretation is largely dependent on the phase of chronic hepatitis B at which it was measured. This phase-dependence arise from inherent differences in the pathophysiology of viral replication (from covalently closed circular DNA in early stages to integration-driven antigen production in later stages) and evolving host immune control. Clinically, quantitative hepatitis B surface antigen offers the most predictive value in hepatitis B e antigen negative patients with low viremia (< 2000 IU/mL). In this group, persistently elevated marker level (> 1000 IU/mL) can identify patients with significantly increased risk of developing hepatocellular carcinoma who warrant closer monitoring or earlier treatment discussion. Recognizing this in context is essential for appropriate clinical use.

INTRODUCTION

Chronic hepatitis B (CHB) remains a widespread disease affecting hundreds of millions of people worldwide and continues to be one of the leading causes of liver cirrhosis and hepatocellular carcinoma (HCC). According to World Health Organization, approximately 254 million people had CHB in 2022, while approximately 1.1 million hepatitis B virus (HBV)-related deaths are reported annually in eight studies[1-8]. HBV is the main culprit for HCC cases in hyperendemic areas as it’s responsible for more than 50% of all cases, with even higher proportions reported in endemic regions of Asia and sub-Saharan Africa[9,10]. Despite well-established surveillance strategies, identifying the CHB at highest risk for cancer remains difficult, especially for patients who don’t meet treatment criteria.

Quantitative hepatitis B surface antigen (qHBsAg) is gaining more visibility as a valid blood test that captures the total intrahepatic viral transcriptional activity and antigen production. In contrast to HBV DNA, which can only measure replicating viruses, qHBsAg integrates signals from two biologically diffrerent sources within the infected hepatocytes, which are the transcriptionally active covalently closed circular DNA (cccDNA) and HBV DNA that has already integrated into the host genome[1,2,11].

The relative contribution of each source shifts across different phases of chronic infection. In addition to reflecting cccDNA and intrahepatic DNA levels, qHBsAg also meaasures HBsAg that arises from the integrated DNA, thereby decreasing its specificity as a biomarker for viral replication[11]. Integration of HBV DNA into the host genome starts early after the inital infection and continues to accumulate over time through repeated cycles of hepatocyte death and regeneration. Even in disease phases where viral replication is suppressed, integrated HBV DNA continues to produce HBsAg and may directly contribute to the development of HCC[2,3,12].

Current literature show multiple observational studies that examined the relationship between qHBsAg and the risk of HCC. However, those produced inconsistent results. Some studies show a positive association, some showed neutral findings, and even a few studies suggested an inverse relationship depending on the population studied and the phase of infection[5,6,13]. A meta analysis of 10 studies pooled 12541 patients found that at the 1000 IU/mL threshold, the pooled odds ratio (OR) for HCC at high vs low HBsAg was 2.46 (95%CI: 2.15-2.83) with low heterogeneity, but at the 100 IU/mL threshold, heterogeneity was substantial (I² = 79%)[13]. These discrepancies indicate that qHBsAg does not carry the same prognostic meaning for all patients. This minireview aims to reconcile these conflicting findings by synthesizing data from large natural history cohorts and emerging mechanistic studies. We propose that the prognostic significance of qHBsAg is fundamentally phase-dependent, reflecting shifts in both the biologic source of HBsAg production and host immune control over the course of chronic HBV infection. This minireview offers three contributions to the existing literature. First, we propose a phase-dependent framework that reconciles previously conflicting results by demonstrating that the qHBsAg-HCC association varies systematically by disease phase. Second, we introduce the concept of persistently elevated qHBsAg (≥ 1000 IU/mL sustained for one year or longer) as a clinical surrogate marker for integration burden – a parameter not directly measurable in routine clinical practice. Third, we provide a simplified clinical algorithm (Figure 1) that translates this framework into actionable practice. Current American Association for the Study of Liver Diseases (AASLD) and European Association for the Study of the Liver (EASL) guidelines do not provide qHBsAg-specific algorithms for HCC risk stratification[3,7,14].

Figure 1 This schematic illustrates the evolving risk of hepatocellular carcinoma in untreated chronic hepatitis B, progressing from left (hepatitis B e antigen-positive) to right (hepatitis B e antigen-negative) phases. The dominant biology shifts from covalently closed circular DNA-driven replication to integration-driven, replication-independent oncogenesis. Fibrosis stage (F0-2 vs F3-4) is the primary determinant of absolute risk. The clinical utility of quantitative hepatitis B surface antigen (qHBsAg) varies by phase: It parallels hepatitis B virus DNA and offers limited prognostic value in hepatitis B e antigen-positive and high-replication states, but serves as a key risk-refinement tool in hepatitis B e antigen-negative patients with low viremia (< 2000 IU/mL), where elevated qHBsAg (> 1000 IU/mL sustained over ≥ 1 year) reflects an increased integration burden and clonal hepatocyte expansion, signaling elevated hepatocellular carcinoma risk independent of viral load. Integration burden is not directly measurable in clinical practice; persistently elevated qHBsAg serves as its surrogate marker. cccDNA: Covalently closed circular DNA; HBeAg: Hepatitis B e antigen; HBV: Hepatitis B virus; HCC: Hepatocellular carcinoma; qHBsAg: Quantitative hepatitis B surface antigen.

METHODOLOGY

We conducted a literature search of PubMed, Embase, and Cochrane Library databases from January 2000 to February 2026. The search strategy combined the following terms: (“quantitative HBsAg” OR “hepatitis B surface antigen“ OR “qHBsAg“) AND (“hepatocellular carcinoma“ OR “HCC“ OR “liver cancer“) AND (“chronic hepatitis B“ OR “HBV“). Additional manual searches of reference lists from key articles and recent reviews were performed.

Inclusion criteria: (1) Original cohort studies (prospective or retrospective) with ≥ 100 untreated CHB patients; (2) Minimum follow-up of 5 years; reported quantitative HBsAg levels and HCC incidence; and (3) Provided adjusted hazard ratios (HRs) or sufficient data for calculation.

Exclusion criteria: (1) Studies limited to treated patients only (we focused on natural history); (2) Cross-sectional designs; conference abstracts without full publications; and (3) Non-English language articles. While we prioritized large cohort studies for quantitative risk assessment, we also included smaller mechanistic studies in the Biological Basis section to support claims regarding cccDNA biology, HBV integration, and clonal hepatocyte expansion.

We acknowledge that exclusion of non-English language articles may introduce language bias, particularly for regions with high HBV burden where important studies may be published in local languages. This limitation is further discussed in the Limitations section.

BIOLOGICAL BASIS OF HBSAG PRODUCTION AND PHASE-DEPENDENT SIGNALING

Circulating HBsAg originates from two biologically distinct intracellular sources: Transcriptionally active cccDNA and HBV DNA that has integrated into the host genome[1,2]. Understanding how the contribution of each source varies across disease phases is essential for proper interpretation of qHBsAg levels. Following hepatocyte infection, relaxed circular HBV DNA travels to the nucleus and converts into cccDNA, which functions as a stable episomal minichromosome and serves as the main transcriptional template for all viral RNAs. cccDNA persists for years, resists immune clearance, and continuously produces viral proteins – including HBsAg – even without active liver inflammation[4,15]. In hepatitis B e antigen (HBeAg)-positive chronic infection, high-level viral replication and abundant cccDNA transcription drive markedly elevated circulating HBsAg levels (typically above 10000 IU/mL) that generally parallel serum HBV DNA[11]. In this setting, qHBsAg mostly reflects the replicative viral burden rather than cumulative cancer risk.

By contrast, HBV DNA integration into the host genome is a replication-independent process that begins early after infection and accumulates over time through cycles of hepatocyte turnover in six studies[15-20]. Although integrated HBV DNA cannot produce infectious virions, many integrants retain functional surface gene sequences and continue to express HBsAg independently of cccDNA activity[1,2,19]. A recent study of paired liver biopsies from persons with CHB on nucleos(t)ide analogue (NA) therapy demonstrated that with increased treatment duration, cells producing HBsAg mRNAs shifted their transcription from chiefly cccDNA to chiefly integrated DNA[3]. In HBeAg-negative CHB, 78% of patients had greater than 50% integrated DNA of total HBV DNA in their livers, and the high frequency of DR2-DR1 integration distribution suggests selection advantage and clonal expansion of these integrants[11,21-24].

Importantly, integration burden is not currently quantifiable in routine clinical practice. No commercially available assay measures integrated HBV DNA directly. In the low-viremic setting, persistently elevated qHBsAg serves as a surrogate marker for integration burden, reflecting replication-independent HBsAg production from integrated templates. However, qHBsAg is an indirect measure; a high qHBsAg level does not prove the presence of oncogenic integrations, nor does a low qHBsAg level exclude them.

Integration events are not biologically inert. Integrated viral fragments can promote genomic instability, insertional mutagenesis, dysregulated oncogene expression, and clonal hepatocyte expansion, thereby contributing directly to liver cancer development[12,17,18].

A genomic analysis of HCC samples demonstrated that HBV integrations cause both local structural variants and distant oncogenic alterations, which include TERT promoter rearrangements and CCND1 amplifications[18,22,23]. Clonal populations of hepatocytes carrying advantageous integration events can progressively dominate the liver parenchyma, leading to persistent HBsAg production that is uncoupled from viral replication but associated with increased malignant potential[15,16].

Based on above findings, there appears to be a clear clinical implication: Identical qHBsAg levels carry different prognostic meanings and should be interpreted with respect to the underlying source. In early disease phases with high viral replication, a high qHBsAg would primarily reflect cccDNA transcription and might occur with no true hepatocellular injury. However, in HBeAg and low viremia, persistence of qHBsAg elevation may signal integration-driven antigen production and therefore increased oncogenic risk. Therefore; the biologic source of HBsAg, rather than merely its concentration, should determine its clinical significance.

RECONCILING CONFLICTING EVIDENCE

The heterogeneity in reported qHBsAg-HCC associations can be largely explained by differences in disease phase composition across study populations. Studies enrolling predominantly HBeAg-positive patients (e.g., immune-tolerant cohorts) tend to find no association or even an inverse association between qHBsAg and HCC, whereas studies focused on HBeAg-negative patients with low viremia consistently report positive associations[6,7,19]. Mixed-population studies produce intermediate or null results depending on the proportion of patients in each phase[5,13].

The ERADICATE-B/REVEAL-HBV combined analysis of 6139 non-cirrhotic Taiwanese patients (median follow-up 21.7 years) directly demonstrated this phase-dependent divergence: HBsAg levels were positively associated with HCC risk in HBeAg-negative patients but negatively associated in HBeAg-positive patients[6]. This single study reconciles much of the prior conflicting literature by showing that the direction of the qHBsAg-HCC association reverses depending on HBeAg status.

A meta-analysis confirmed that at the 1000 IU/mL cutoff, the pooled OR for HCC was 2.46 (95%CI: 2.15-2.83) with low heterogeneity (I² not significant), suggesting that when an appropriate threshold is used, the association is consistent[13]. The higher heterogeneity observed at the 100 IU/mL cutoff (I² = 79%) likely reflects the inclusion of studies with mixed HBeAg-positive and HBeAg-negative populations.

QUANTITATIVE HBsAg AND HCC RISK BY DISEASE PHASE

Table 1 summarizes the phase-dependent interpretation of qHBsAg and its association with HCC risk across the natural history of CHB[1-6,8,9,23,25].

Table 1 Phase-dependent interpretation of quantitative hepatitis b surface antigen for hepatocellular carcinoma risk stratification.

Disease phase	Typical qHBsAg	Primary HBsAg source	The qHBsAg-HCC association	Key evidence	Clinical action beyond guidelines	Ref.
HBeAg-positive chronic infection	> 10000 IU/mL	cccDNA (replication-driven)	Inverse or neutral; higher HBsAg associated with delayed HCC	ERADICATE-B/REVEAL-HBV (n = 6139; 21.7 years f/u): HBsAg ≥ 10000 associated with delayed HCC in HBeAg-positive	The qHBsAg not recommended; HBV DNA is dominant predictor	Kumada et al[1], Tseng et al[4], Yang et al[6]
HBeAg-negative, low viremia (DNA < 2000 IU/mL)	100 to > 1000 IU/mL	Integrated HBV DNA	Strongly positive; HR 13.7 (4.8-39.3) for ≥ 1000 vs < 1000	ERADICATE-B (n = 2688; 14.7 years); Shanghai OR 2.21 (1.10-4.43); meta-analysis OR 2.46 (2.15-2.83)	Measure qHBsAg; if ≥ 1000 sustained ≥ 1 year: Enhanced surveillance and consider treatment	Seto et al[2], Terrault et al[3], Tseng et al[4]
Indeterminate phase (“grey zone“)	Variable	Mixed	Positive; biomarkers stratify heterogeneous risk	HBcrAg stratified risk (HR = 4.47, 2.62-7.63); annual HCC 0.32%	The qHBsAg > 1000 may identify higher-risk subset	Tseng et al[4], Ghany et al[25]
Intermediate viremia (DNA 2000-20000 IU/mL)	Variable	Mixed	Attenuated; HBcrAg may be more informative	HBcrAg ≥ 10 KU/mL: HR = 6.29 (2.27-17.48)	Modest adjunctive value; HBcrAg preferred	Zhao et al[23]
High viremia (DNA > 20000 IU/mL)	> 10000 IU/mL	cccDNA	Neutral; HBV DNA dominates	REVEAL-HBV: HR of 10.7 for DNA ≥ 200000 vs undetectable	The qHBsAg not recommended; manage per guidelines	Yucuma et al[5], Grudda et al[9]
Cirrhosis/advanced fibrosis (F3-F4)	Variable	Variable	Neutral; fibrosis dominates	Annual HCC 3%-5%; AASLD recommends surveillance for all	Fibrosis supersedes viral biomarkers	Tseng et al[4], Mahajan et al[8]

Integration burden is inferred from persistently elevated quantitative hepatitis B surface antigen in low-viremic patients; it is not directly measurable in clinical practice. Hazard ratios represent estimates from individual cohort studies and meta-analyses; individual patient risk varies. Fibrosis stage remains the dominant predictor of hepatocellular carcinoma risk in patients with advanced fibrosis/cirrhosis and supersedes viral biomarkers in clinical decision-making. Quantitative hepatitis B surface antigen cutoffs (100 IU/mL, 1000 IU/mL) were derived from Asian cohorts with hepatitis B virus genotypes B and C and have not been validated in non-Asian populations. AASLD: American Association for the Study of Liver Diseases; cccDNA: Covalently closed circular DNA; HBcrAg: Hepatitis B core-related antigen; HBeAg: Hepatitis B e antigen; HBsAg: Hepatitis B surface antigen; HBV: Hepatitis B virus; HCC: Hepatocellular carcinoma; HR: Hazard ratio; OR: Odds ratio; qHBsAg: Quantitative hepatitis B surface antigen.

HBeAg-positive chronic infection (historically “immune-tolerant“)

In patients with HBeAg-positive chronic infection [normal alanine aminotransferase (ALT), high HBV DNA], qHBsAg levels typically exceed 10000 IU/mL and reflect active cccDNA transcription without significant immune-mediated liver injury[6,11]. The combined ERADICATE-B/REVEAL-HBV analysis of 6139 patients (median follow-up 21.7 years) found that among HBeAg-positive immune-tolerant patients, HBsAg levels ≥ 10000 IU/mL were associated with delayed HCC development, validated both internally and externally by an independent Japanese cohort[6]. The landmark REVEAL-HBV study similarly demonstrated that among HBeAg-positive patients, HBV DNA – not qHBsAg – was the dominant predictor of HCC[9]. In this clinical setting, qHBsAg provides minimal additional prognostic information beyond demographic and fibrosis-related factors.

HBeAg-negative chronic infection with low viremia (HBV DNA < 2000 IU/mL)

In HBeAg-negative patients with low HBV DNA levels (< 2000 IU/mL), qHBsAg demonstrates its strongest and most consistent association with HCC risk. Multiple longitudinal cohorts have shown that higher qHBsAg levels in this subgroup are independently associated with increased HCC incidence, even after adjustment for viral load and biochemical activity[25,26].

In the ERADICATE-B cohort (2688 Taiwanese HBsAg-positive patients, mean follow-up 14.7 years), among HBeAg-negative patients with HBV DNA < 2000 IU/mL, the adjusted HR for HCC in patients with HBsAg ≥ 1000 IU/mL vs < 1000 IU/mL was 13.7 (95%CI: 4.8-39.3)[7]. This finding was corroborated by the Lancet Seminar on CHB, which cited the same relative risk of 13.7 (4.8-39.9) for HBsAg ≥ 1000 IU/mL in this subgroup[10]. A nested case-control study from Shanghai (211 liver cancer cases, 221 HBsAg-positive controls) confirmed a dose-response relationship: Compared with HBsAg 100 IU/mL, the adjusted OR for HCC was 1.82 (95%CI: 0.90-3.68) for HBsAg 100-999 IU/mL and 2.21 (95%CI: 1.10-4.43) for HBsAg ≥ 1000 IU/Ml[19]. A separate population-based study from Shanghai demonstrated that HBsAg levels were positively associated with liver cancer risk in a dose-response manner, with the adjusted OR reaching 47.33 (95%CI: 23.50-95.34) at the highest HBsAg level (≥ 1000 IU/mL) in men[20].

In this phase, elevated qHBsAg likely reflects replication-independent expression from integrated HBV DNA and serves as a surrogate marker of cumulative oncogenic exposure and clonal hepatocyte expansion[15].

HBeAg-negative indeterminate phase (the “grey zone“)

A substantial proportion of HBeAg-negative patients do not fit neatly into traditional categories, presenting with modestly elevated ALT or HBV DNA levels that do not meet treatment thresholds. This “indeterminate phase“ affects up to 40% of adults with CHB[7]. The AASLD 2026 guideline reported a pooled annual HCC incidence of 0.32% (95%CI: 0.21%-0.48%) in the indeterminate phase, with antiviral treatment associated with lower HCC incidence (adjusted incidence rate ratio 0.36, 95%CI: 0.16-0.81)[7]. In a territory-wide Hong Kong study of 17287 HBeAg-negative patients, those in the indeterminate phase had significantly higher HCC risk than those in chronic infection (adjusted HR = 1.587, 95%CI: 1.262-1.995)[21].

In the ERADICATE-B cohort, hepatitis B core-related antigen (HBcrAg) level of 10000 U/mL stratified HCC risk in HBeAg-negative indeterminate-phase patients[27], with 10-year HCC cumulative incidence of 0.51% vs 5.33% for low vs high HBcrAg groups (HR = 4.47, 95%CI: 2.62-7.63)[28-32]. Although this study focused on HBcrAg rather than qHBsAg, it demonstrates that viral biomarkers can meaningfully stratify risk in this heterogeneous population. By analogy, qHBsAg > 1000 IU/mL may similarly identify a higher-risk subset and inform decisions about closer monitoring or earlier treatment.

Intermediate viremia (HBV DNA 2000-20000 IU/mL)

Among patients with intermediate HBV DNA levels (2000-20000 IU/mL), the prognostic role of qHBsAg is attenuated. Both viral replication and host factors contribute to disease progression, and qHBsAg provides modest incremental risk information when interpreted alongside HBV DNA trends[10]. In the ERADICATE-B cohort, HBcrAg level ≥ 10 KU/mL identified patients with intermediate viral load who were at increased risk for hepatocellular cancer (HR = 6.29, 95%CI: 2.27-17.48)[33-35], suggesting that cccDNA-derived markers are potentially more accurate and informative than qHBsAg in this subgroup. Note that the variability seen in reported thresholds is likely a reflection of the heterogeneity in the relative contribution of cccDNA and integrated HBV DNA.

High viremia (HBV DNA > 20000 IU/mL)

In persistently high-viremic states, HBV DNA remains the dominant determinant of HCC risk as shown by multiple studies: The REVEAL-HBV study demonstrated a dose-response relationship, with HR for HCC of 2.7 for HBV DNA 2000-19999 IU/mL, 8.9 for 20000-199999 IU/mL, and 10.7 for ≥ 200000 IU/mL compared with undetectable HBV DNA[9,28]. A 2026 systematic review and meta-analysis confirmed approximately a five-fold increase in HCC incidence at HBV DNA ≥ 20000 IU/mL compared with 2000 IU/mL[22]. Therefore, it can be hypothesized that in this disease phase, qHBsAg is mainly a reflection of viral replication and therefore offers little additional benefit in prognostication.

Cirrhosis and advanced fibrosis (F3-F4)

In patients with established cirrhosis or advanced fibrosis (F3-F4), structural liver disease remains the dominant determinant of HCC risk, with a high annual incidence rate reaching 3%-5%[8]. The AASLD 2026 guideline recommends widespread HCC surveillance for all cirrhotic patients who have a detectable HBV DNA irrespective to their ALT levels[7,26]. In these individuals, viral biomarkers offer limited incremental discrimination, and management should be guided primarily by fibrosis stage and HCC surveillance protocols[29].

Longitudinal dynamics: Beyond single measurements

Based on above findings, serial and longitudinal measurements of qHBsAg dynamics is more likely to provide more valuable prognostic information rather than isolated single measurement. In HBeAg-negative patients, persistently elevated qHBsAg (> 1000 IU/mL over one or more years) likely holds more value for predicting an increased risk of HCC than a single elevated value. The 2026 systematic review by Yucuma et al[5] demonstrated that incidence rates of HCC were much lower in individuals with persistently low HBV DNA (< 2000 IU/mL) across multiple assessments compared with those with low HBV DNA at a single baseline assessment, reinforcing the importance of serial monitoring[22].

Conversely, spontaneous HBsAg seroclearance is associated with a substantially decreased risk of HCC (relative risk = 0.34, 95%CI: 0.20-0.56), with the lifetime cumulative incidence of HCC falling from 14.2% to 4.0%[10]. HCC risk is reduced by approximately 70% following HBsAg seroclearance, particularly when it occurs before age 50 years and prior to development of cirrhosis[8]. However, HCC risk persists after HBsAg loss, especially in those with cirrhosis. These temporal patterns should be incorporated into risk assessment.

QUANTITATIVE HBSAG IN TREATED PATIENTS

While this review focuses on untreated patients to characterize natural history, the role of qHBsAg in NA-treated patients warrants brief discussion. In patients receiving long-term NA therapy, HBsAg production shifts from cccDNA to integrated DNA sources over time[3]. NAs have no effect on integrated HBV DNA, which continues to produce HBsAg during therapy[1]. A recent study developed the HBsAg-HCC score incorporating HBsAg quantification alongside age, sex, hypoproteinaemia, and APRI, achieving 3-year/5-year/7-year area under the receiver operating characteristics curves of 0.867/0.872/0.871 in the training cohort, significantly outperforming PAGE-B, mPAGE-B, and aMAP[35-37].

Following HBsAg seroclearance, HCC risk persists but is substantially reduced. In a territory-wide Hong Kong study of 9769 patients who achieved HBsAg seroclearance[38], 1.1% developed HCC during median follow-up of 4.6 years, with risk concentrated in those who were older, male, and had cirrhosis at the time of HBsAg loss[39-43]. In NA-treated patients who achieved HBsAg seroclearance, HCC risk was further reduced compared with spontaneous seroclearance (adjusted HR = 0.442)[42]. Current guidelines do not recommend qHBsAg for routine treatment monitoring, and this remains an area of active investigation.

COMPARISON WITH OTHER EMERGING VIRAL BIOMARKERS

Although qHBsAg provides unique insight into cumulative viral antigen production, it represents only one component of the complex biology of chronic HBV infection. Several complementary biomarkers have emerged that may further refine risk assessment, each reflecting distinct aspects of the viral life cycle.

HBcrAg

HBcrAg is a composite marker consisting of HBcAg, HBeAg, and a truncated 22-kDa precore protein (p22cr). It correlates closely with intrahepatic cccDNA quantity and transcriptional activity. A meta-analysis of 18 studies (10320 patients) demonstrated that high serum HBcrAg was an independent risk factor for HCC occurrence (adjusted HR = 3.12, 95%CI: 2.40-4.06, I² = 43.2%)[31]. In untreated patients, HBcrAg was superior to HBV DNA in predictive power for HCC development by time-dependent receiver operating characteristics analysis (HR = 5.05, 95%CI: 2.40-10.63 for HBcrAg > 2.9 log U/mL)[30]. In HBeAg-negative patients with intermediate viral load, HBcrAg ≥ 10 KU/mL identified those at high HCC risk (HR = 6.29, 95%CI: 2.27-17.48). In NA-treated HBeAg-negative patients, serum HBsAg level did not correlate with HCC risk, whereas HBcrAg above 2.9 log U/mL was an independent predictor (adjusted HR = 2.13, 95%CI: 1.10-4.14).

HBV RNA

Serum HBV RNA, primarily representing pregenomic RNA encapsidated in virion-like particles, reflects active transcription from cccDNA. In NA-treated patients who achieved viral suppression, quantifiable HBV RNA independently predicted HCC (adjusted HR = 3.313, 95%CI: 1.154-9.507), outperforming HBcrAg in this setting[1]. In HBeAg-negative patients with low viremia, HBV RNA levels are often low or undetectable when qHBsAg remains elevated, confirming the integration-driven origin of HBsAg in this setting[36].

Multimarker panels

Each biomarker reflects distinct aspects of the viral life cycle, and none alone fully captures both replication-dependent and integration-driven processes. Therefore, it is advisable to combine multiple biomarkers to enhance prognostic accuracy. The ERADICATE-B cohort demonstrated that combining ALT, HBV DNA, and HBcrAg levels could reclassify HBeAg-negative patients into high-risk and low-risk groups[27]. Future risk stratification strategies will likely incorporate multimarker panels combining qHBsAg with HBV RNA, HBcrAg, and host factors to improve prediction of long-term outcomes[10].

INTEGRATION WITH HOST FACTORS

The utility of qHBsAg in clinical practice should be viewed alongside other host factors that are known modifiers of HCC risk, those include male sex, age > 40 years, family history of HCC, HBV genotype, coinfection with human immunodeficiency virus/hepatitis C virus/hepatitis delta virus, cigarette smoking, alcohol use, and comorbid conditions including diabetes and metabolic dysfunction-associated steatotic liver disease[8]. Age modifies the interpretation of qHBsAg: A qHBsAg level of 1000 IU/mL in a younger patient may reflect active replication, while the same value in an older patient may reflect accumulated integration burden over decades of chronic infection. Fibrosis stage remains the strongest predictor of absolute HCC risk across all phases and should supersede viral biomarkers in clinical decision-making[7].

CLINICAL IMPLICATIONS AND PRACTICAL FRAMEWORK

Quantitative HBsAg should be viewed as a risk refinement tool rather than a standalone predictor. It carries the most weight in aiding to identify higher-risk patients with otherwise low-risk viral categories, particularly those who are HBeAg negative and have low viremia (< 2000 IU/mL).

When qHBsAg measurement is most informative

The qHBsAg can provide the best clinical value in the instances below: (1) HBeAg-negative patients with low viremia (HBV DNA < 2000 IU/mL): Persistent elevation of the marker above 1000 sustained for at least 1 year indicated a markedly increased HCC risk and therefore warrants enhanced surveillance (e.g., biannual liver ultrasounds) and should be considered for earlier antiviral treatment[7,10,19]; and (2) HBeAg-negative indeterminate phase patients: Elevated qHBsAg may refine risk stratification regarding potential closer monitoring and earlier treatment discussion[7,21].

When qHBsAg adds limited value

The qHBsAg offers little added clinical value in the instances below: (1) HBeAg-positive chronic infection: Most qHBsAg parallels HBV DNA in this stage, and therefore higher levels can paradoxically indicate a delayed HCC onset, and therefore should not be used as a clinical marker in those patients[6,13]; (2) High viremia (HBV DNA > 20000 IU/mL): HBV DNA dominates risk prediction[9]; (3) Cirrhosis or advanced fibrosis (F3-F4): Fibrosis stage determines absolute risk; viral biomarkers do not alter management; and (4) Patients already meeting treatment criteria: Antiviral therapy should be initiated based on guidelines regardless of qHBsAg level.

Practical considerations for clinical implementation

Implementing qHBsAg in clinical practice requires addressing practical challenges.

Assay standardization: Multiple commercial qHBsAg assays (Abbott Architect, Roche Elecsys, Siemens) have decent correlation however vastly different absoulute values. The qHBsAg levels vary by genotype (higher in genotype A) and by presence of preS/S mutants[12,31]. Clinicians should use consistent assays for longitudinal monitoring and be aware that cutoffs may not be directly transferable between platforms. The cutoffs proposed in this review (100 IU/mL, 100-1000 IU/mL, > 1000 IU/mL) were derived from large Asian cohort studies in which > 95% of patients had HBV genotypes B or C[8,9]. Importantly, these cutoff values were not validated in non-Asian populations or other genotypes (A, D, E).

Frequency of testing: For HBeAg-negative patients with low viremia, annual qHBsAg measurement is reasonable as the value hold less clinical utility. Contrast with patients with elevated qHBsAg (> 1000 IU/mL), in whom confirmation with a second measurement at 6-12 months is recommended to assess persistence.

Interpretation pitfalls: A single elevated qHBsAg value in an older patient (> 50 years) may reflect accumulated integration burden over decades, while the same value in a young patient (< 30 years) may indicate active replication, and therefore qHBsAg should be viewed in full clinical context. The qHBsAg fluctuations < 0.5 Log are within assay variability and should not trigger changes in management. In patients with cirrhosis, qHBsAg should not be used to downgrade surveillance intensity. Integration burden is inferred from persistently elevated qHBsAg; it is not directly measurable clinically.

Integration with guidelines: Current AASLD (2025/2026)[3] and EASL (2025) guidelines do not mandate qHBsAg for HCC risk stratification[11,21,25,29,43]. We propose its use as a complementary tool to aid with shared decision making especially in patients that fit the specifications for the testing to hold a higher clinical value (e.g., low-viremic patients near treatment thresholds, indeterminate phase patients).

Proposed definitions and cutoffs

Based on the evidence reviewed, we propose the following operational definitions: (1) Low viremia: HBV DNA < 2000 IU/mL; (2) Intermediate viremia: HBV DNA 2000-20000 IU/mL; (3) High viremia: HBV DNA > 20000 IU/mL; (4) Persistently elevated qHBsAg: The qHBsAg > 1000 IU/mL on ≥ 2 measurements ≥ 1 year apart; (5) Low qHBsAg: < 100 IU/mL; (6) Intermediate qHBsAg: 100-1000 IU/mL; (7) High qHBsAg: > 1000 IU/mL; (8) Significant qHBsAg decline: ≥ 0.5 Log reduction from baseline sustained over ≥ 2 years; and (9) Discordant qHBsAg-HBV DNA: The qHBsAg > 1000 IU/mL with HBV DNA < 2000 IU/mL.

Disease phase terminology follows current AASLD 2025/2026 and EASL 2025 guidelines[10,21,23].

Simplified clinical algorithm

A simplified conceptual algorithm is presented in Figure 1.

HBeAg-negative + HBV DNA < 2000 IU/mL [measure qHBsAg (confirm with repeat at 6-12 months)]: (1) Persistently < 100 IU/mL: Standard surveillance (annual ultrasound); (2) Persistently 100-1000 IU/mL: Standard surveillance, consider repeat qHBsAg annually; and (3) Persistently > 1000 IU/mL: Enhanced surveillance (6-monthly ultrasound), consider treatment discussion (particularly if age > 50, family history of HCC, or fibrosis stage ≥ F2).

HBeAg-negative + HBV DNA 2000-20000 IU/mL (assess fibrosis stage): (1) F0-F1: The qHBsAg may provide adjunctive information (consider if > 1000 IU/mL); and (2) F2-F4: Manage per guidelines regardless of qHBsAg.

HBeAg-positive or HBV DNA > 20000 IU/mL or cirrhosis (manage per guidelines): The qHBsAg not required for decision-making.

LIMITATIONS AND FUTURE DIRECTIONS

Study design limitations of cited cohorts

Most included studies are retrospective or post-hoc analyses of prospectively collected cohorts. As such, they are subject to selection bias, variable loss to follow-up, and incomplete adjustment for confounders. The foundational ERADICATE-B study, for example, determined baseline HBsAg levels retrospectively from stored sera[10]. Publication bias may overestimate associations, as studies with null findings are less likely to be published. The consistency of findings across different study designs and populations suggests the phase-dependent framework is robust, but the precise magnitude of risk associated with qHBsAg thresholds requires prospective validation.

Critical appraisal of study quality

Common limitations across the cited cohort studies include: (1) Adjustment for fibrosis stage in only a minority of studies, despite fibrosis being the dominant predictor of HCC; (2) Reliance on single baseline qHBsAg measurement in most studies, despite longitudinal dynamics having greater prognostic value; and (3) Loss to follow-up rates that are variably reported without sensitivity analysis. These limitations suggest that reported HRs may be biased toward the null (due to non-differential misclassification from single measurements) or away from the null (due to publication bias and incomplete confounder adjustment).

Geographic and genotypic generalizability

Data derive predominantly from Asian cohorts infected with HBV genotypes B and C. The qHBsAg cutoffs proposed in this review (100 IU/mL, 1000 IU/mL) were derived from the ERADICATE-B and REVEAL-HBV cohorts, in which > 95% of patients had genotypes B or C[10]. Whether the same thresholds apply to other genotypes remains uncertain.

HBV genotype significantly influences HCC risk independently of qHBsAg. In the Alaska hepatitis B cohort (1185 persons, median follow-up 35.1 years), HCC incidence per 1000 person-years was 5.73 for genotype F, 4.77 for genotype C, 1.28 for genotype A, 0.47 for genotype D, and 0.00 for genotype B. In sub-Saharan Africa, genotype A1 (southern and east Africa) and genotype E (west Africa) carry increased HCC risk, and genotype-independent preS2 deletion mutations are significantly associated with HCC in African cohorts[39,40]. These genotype-specific differences in HCC risk may alter the prognostic significance of qHBsAg thresholds.

The qHBsAg levels also vary by genotype: Levels are generally higher in genotype A than in genotypes B-D, and are influenced by the presence of preS/S mutants[3]. This means that a qHBsAg level of 1000 IU/mL in a genotype A patient may not carry the same biological significance as the same level in a genotype B or C patient.

Strategies for genotype A/D/E validation

To address the generalizability gap for non-Asian genotypes, we propose the following strategies.

Strategy 1: Multi-center consortium to help validate qHBsAg thresholds in non Aisan populations (genotypes A, D, and E).

Strategy 2: Individual-patient data meta-analysis. Conduct an individual participant data meta-analysis of existing smaller cohorts from non-Asian regions, therefore potentially overcoming the limitations of isolated studies and providing genotype-stratified HRs for HCC risk.

Strategy 3: In silico modeling. Use available sequencing data from HCC samples (TCGA, ICGC) to model genotype-specific integration patterns and predict whether the same qHBsAg thresholds are biologically plausible for genotypes A, D, and E.

Age and cohort effects

In the current literature, many foundational studies took place between the 1980s-2000s, when epidemiology and treatment patterns were significantly different from current practice. For example, the ERADICATE-B cohort recruited patients from 1985 through 2000. It’s possible that contemporary patients had different integration burdens, fibrosis progression rates, and HCC risk due to either earlier treatment initiation or even the dynamic changes in the viral genotype. Age is also an important modifier of qHBsAg, as it accumulates over time by the natural course of disease. Therefore, a high level in a 30-year-old may indicate active replication, but merely reflects accumulated integration burden in a 60-year-old.

Temporal changes in disease phase

Due to the natural history of HBV infection, people move between disease phases overtime, and therefore studies that used baseline HBsAg only might have misclassified risk if patients transition phases during follow-up. For instance, a patient initially classified as HBeAg-positive chronic infection may transition to HBeAg-negative chronic hepatitis, altering the prognostic meaning of their qHBsAg level. Therefore, serial measurements rather than a single value, while reassessing the disease phase at each measurement, is more likely to allow for more prognostic value of qHBsAg. The importance of serial monitoring is reinforced by the finding that incidence rates of HCC were much lower in individuals with persistently low HBV DNA across multiple assessments compared with those with low HBV DNA at a single baseline assessment.

Practical limitations of qHBsAg in clinical settings

Cost: The qHBsAg adds approximately $50-$100 per test, which may not be covered by insurance in some regions. The cost-effectiveness of qHBsAg-guided surveillance has not been formally evaluated in a published cost-effectiveness analysis, and this represents an important gap for future research.

Accessibility: The qHBsAg is not routinely available in low-resource settings where HBV burden is highest, particularly in sub-Saharan Africa where genotypes A, D, and E predominate. Point-of-care qHBsAg tests are in development but not yet clinically validated.

Lack of standardized reporting: No international standard for qHBsAg units (IU/mL vs log IU/mL) across laboratories. Some reports use log-transformed values, complicating clinical interpretation.

Turnaround time: Results may take days, limiting point-of-care use.

Implementation barriers and solutions

Several barriers may limit widespread implementation of qHBsAg-guided risk stratification in clinical practice (Table 2)[3,40].

Table 2 Implementation barriers and proposed solutions for quantitative hepatitis b surface antigen-guided risk stratification.

Barrier	Details	Proposed solution
Cost	$50-$100 per test	Targeted testing in higher-risk subgroups (age > 50, family history of HCC, ≥ F2 fibrosis)
Insurance coverage	Not covered in some regions	Advocate for coverage in high-risk subgroups (HBeAg-negative, low viremia)
Accessibility	Not available in low-resource settings where HBV burden is highest	Point-of-care qHBsAg tests in development; advocate for technology transfer[40]
Assay standardization	Different absolute values across platforms (Abbott, Roche, Siemens)	Use same platform for serial monitoring; cross-platform calibration studies needed
Clinician training	Unfamiliarity with phase-dependent interpretation	Educational materials and clinical decision aids
Turnaround time	Results may take days	Not suitable for same-day decisions; plan testing at routine visits
Genotype applicability	Cutoffs validated only for genotypes B and C	Prospective validation in non-Asian populations required before broad adoption[3,40]

HBV: Hepatitis B virus; HCC: Hepatocellular carcinoma; HBeAg: Hepatitis B e antigen; qHBsAg: Quantitative hepatitis B surface antigen.

Future directions

Prospective validation of phase-dependent qHBsAg thresholds in diverse populations and genotypes, especially in non-Aisan populations with genotypes A, D, and E. Integration of longitudinal qHBsAg dynamics into risk prediction models to facilitate a more comprehensive clinical picture, serial measurements with corresponding disease stage at each. Multimarker panels combining qHBsAg with HBcrAg, HBV RNA, and host factors (age, fibrosis markers, genetics) to improve prediction. The American Gastroenterological Association 2026 Clinical Practice Update notes that multidimensional models incorporating genetic, metabolic, and biomarker data may better distinguish low-risk from high-risk patients, though most require further external validation[41].

Clinical trials to evaluate whether qHBsAg-guided surveillance or earlier treatment improves outcomes compared with standard guideline-based management. Standardization of qHBsAg assays and reporting to facilitate clinical implementation across platforms and regions. Validation in treated patients to define the role of on-treatment qHBsAg monitoring. The HBsAg-HCC Score, which incorporates qHBsAg alongside age, sex, hypoproteinaemia, and APRI, represents a promising approach for NA-treated patients[1]. Cost-effectiveness analyses of qHBsAg-guided surveillance strategies in different healthcare settings and populations.

Inclusion of non-English language databases (CNKI, CiNii, KCI, SciELO, AJOL) in future systematic reviews to ensure comprehensive capture of regional data, particularly for genotypes D and E where much research may be published in local languages.

CONCLUSION

Quantitative HBsAg provides biologically meaningful yet context-dependent information in CHB. Its clinical significance varies according to disease phase and reflects shifting contributions from transcriptionally active cccDNA and replication-independent expression from integrated HBV DNA[1-3]. These mechanistic differences explain the heterogeneous associations between qHBsAg levels and HCC risk reported across cohorts[9-11].

In clinical practice, qHBsAg is most valuable in HBeAg-negative patients with low viremia (< 2000 IU/mL), where persistently elevated levels (≥ 1000 IU/mL sustained over ≥ 1 year) identify individuals with markedly increased HCC risk (adjusted HR = 13.7, 95%CI: 4.8-39.3 in the ERADICATE-B cohort)[10]. When interpreted within the appropriate clinical context, qHBsAg may serve as an important adjunct for refining HCC risk stratification. However, it should only complement rather than replace established predictors such as fibrosis stage and viral load[10].

Prospective studies incorporating longitudinal qHBsAg dynamics into multivariable risk models are needed to define its optimal role in routine practice. Integration with emerging biomarkers (HBcrAg, HBV RNA) and host factors can allow for a more personalized risk prediction models and therefore surveillance strategies for the millions of individuals living with CHB worldwide[1].

For clinicians, the key take-home message is selective use of qHBsAg: Measure it in HBeAg-negative patients with HBV DNA below 2000 IU/mL. In this group, a single measurement is insufficient; confirm persistence with a second measurement at 6-12 months. Persistently elevated levels (≥ 1000 IU/mL) warrant enhanced surveillance (6-monthly ultrasound) and consideration of earlier treatment, particularly in patients over age 50, those with family history of HCC, or those with F2 or greater fibrosis[10]. In all other phases of CHB, qHBsAg adds little to established predictors and should not drive management decisions.