Introduction

Hepatitis B virus (HBV) remains a major global health issue, with the World Health Organization estimating 296 million people living with chronic hepatitis B (CHB) and 820,000 annual deaths from related complications like cirrhosis and liver cancer1. Despite successful vaccination and antiviral treatments reducing new CHB cases, many people worldwide have past HBV exposure, indicated by the presence of antibodies to the hepatitis B core antigen (HBcAb). This is especially common in high-endemic regions, affecting 20%-40% or more of the population2,3. Those who are HBsAg-negative but HBcAb-positive are often regarded as having resolved infection or being in a “non-active” state4. However, emerging evidence suggests that this serological marker may indicate occult HBV infection (OBI), characterized by persistent HBV DNA at low replicative and transcriptional levels despite hepatitis B surface antigen (HBsAg) seroclearance5,6,7,8. Understanding the prevalence and characteristics of OBI in the HBsAg-negative, HBcAb-positive population is vital for improving public health strategies and developing targeted screening for high-risk groups.

Despite undetectable HBsAg, OBI poses significant clinical challenges, such as HBV transmission through blood transfusion or transplantation, reactivation during immunosuppression, and progression to liver disease9. Alarmingly, reactivation rates of 20% to 50% have been reported in HBsAg-negative, HBcAb-positive patients receiving immunosuppressive therapy7,8,10. Furthermore, clinical guidelines emphasize precautions to prevent HBV transmission from HBcAb-positive liver graft donors11, recognizing them as a potential source of occult HBV. While the clinical significance of OBI in HBcAb-positive patients are gaining recognition, the prevalence, geographic distribution, clinical spectrum, and factors associated with its occurrence in the broader HBsAg-negative, HBcAb-positive population remain poorly understood, with significant variability across studies.

Given the uncertainties surrounding OBI in the HBcAb-positive population, a systematic review and meta-analysis is urgently needed. This study aims to estimate the global prevalence of OBI in the HBsAg-negative, HBcAb-positive population and examine study-level characteristics such as geographical HBV endemicity, demographics, and diagnostic methods. Furthermore, we will compare OBI prevalence in blood (serum/plasma) versus liver tissue samples to obtain a more accurate estimate of the true prevalence of OBI.

Materials and methods

Literature search

This systematic review and meta-analysis was registered in the International Prospective Register of Systematic Reviews (PROSPERO; registration ID: CRD42022329560) and followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. We conducted a comprehensive search of Pubmed, Web of Science, Embase, and Cochrane databases from database inception to January 12, 2025, for studies examining OBI occurrence in HBsAg-negative populations. The search was restricted to articles published in English. The search terms included: (occult hepatitis B infection OR occult hepatitis B virus infection OR occult HBV infection OR OBI) AND (hepatitis B core antibody OR HBcAb OR anti-HBc). The full search strategy for each database is detailed in Table S1. We also manually examined the reference lists of relevant review and original articles to identify additional studies. Where necessary, we contacted the authors of included studies to obtain missing data. Two authors (YTY, HWX) independently screened titles and abstracts. The final selection was based on a full-text evaluation. If multiple publications reported on the same study population, we included the one with the most comprehensive data. Any discrepancies were resolved through discussion among all authors.

Study selection

For this review, OBI was defined as the detection of HBV DNA in blood or liver tissue from HBsAg-negative individuals. Inclusion criteria were as follows: (1) original research conducted in general populations of HBsAg-negative individuals; (2) reported both HBV DNA and HBcAb status; and (3) published in a peer-reviewed journal.

Studies were excluded if they met any of the following criteria: (1) focused on specific high-risk or immunologically distinct populations (e.g., patients with acute infection, hepatocellular carcinoma, decompensated cirrhosis, or co-infection with HCV, HDV, or HIV); (2) included pediatric populations; (3) had a sample size of fewer than 20 individuals in either the HBcAb-positive or HBcAb-negative group; or (4) were reviews, meeting abstracts, case reports, letters, non-English articles, or studies lacking sufficient data for analysis.

Data collection and extraction

According to a predefined form, Two investigators (YTY, HWX) independently extracted data using a standardized Excel spreadsheet. Any disagreements were resolved by consensus or discussion with a third investigator (YZ). The extracted information included: first author, publication year, year of study, country of data collection, HBV prevalence in the study area, population source, sample size, mean/median age, HBcAb-positive rate, HBV DNA detection method.

Quality assessment and risk of bias

Two investigators (YTY, GJW) independently assessed the methodological quality and risk of bias of the included studies using the Joanna Briggs Institute (JBI) Critical Appraisal Checklist for Studies Reporting Prevalence Data12. This 9-item checklist evaluates aspects such as sample representativeness, recruitment methods, and data analysis. For the purpose of structured assessment, we applied a quantitative scoring system. Each item was scored as ‘Yes’ (2 points), ‘Unclear’ (1 point), or ‘No’ (0 points), resulting in a total score ranging from 0 to a maximum of 18. Based on this total score, studies were categorized as high quality (score ≥ 10), medium quality (score 7–9), or low quality (score < 7). Any disagreements regarding the quality assessment were resolved by consensus or with the input of a third investigator (YZ).

Data synthesis and statistical analysis

All statistical analyses were performed using R software (version 4.1.2; R Foundation for Statistical Computing) with meta package. We used a random-effects model (DerSimonian and Laird) to compute the pooled estimates of OBI (95% confidence intervals [CI]) for HBcAb-positive and HBcAb-negative populations, stratified by sample type (blood or liver tissue). To stabilize the variance of proportions, especially for studies with prevalence near 0% or 100%, an arcsine transformation was applied. A continuity correction of 0.5 was added to studies with zero events to allow for their inclusion in the meta-analysis. We assessed publication bias using funnel plots and Egger’s test. We hypothesized that OBI prevalence might vary according to various factors; therefore, we conducted subgroup analyses based on HBV endemicity (low: HBsAg < 2%; moderate-to-high: HBsAg ≥ 2%), HBcAb-positive rate, HBV DNA detection method (nested PCR vs. qPCR), HBsAb status (positive: ≥10 mIU/mL; negative: <10 mIU/mL), population source (blood donors vs. general population), and publication year. A leave-one-out sensitivity analysis was performed to assess the robustness of the findings. Finally, we used random-effects meta-regression to explore potential sources of heterogeneity. A two-sided p value of ≤ 0.05 was considered statistically significant.

Results

The study selection process is illustrated in the PRISMA flow diagram (Fig. 1). Our initial electronic database search identified 2,490 studies. After removing 1,130 duplicates, 1,360 records were screened based on their titles and abstracts, leading to the exclusion of 1,267 records. The full texts of the remaining 93 articles were then assessed for eligibility. Following a detailed review, 45 studies were excluded for various reasons, including insufficient data, ineligible study populations, and other factors (see Fig. 1 for details). Ultimately, 48 studies fulfilled all eligibility criteria and were included in the final meta-analysis.

Fig. 1
Fig. 1
Full size image

PRISMA flow diagram of the literature search process for eligible researches.

The 48 included studies involved a total of 21,783 participants from 20 different regions. The baseline characteristics of these studies are summarized in Table 1 and Table S2. Based on the quality assessment (Table S3), 30 of the 48 studies (62.5%) were rated as the high quality and 15 (31.2%) as medium quality according.

Table 1 Characteristics of included studies1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48.
Full size table

Meta-analyses of OBI in HBsAg-negative population

Among the 48 included studies, 44 assessed OBI using blood sample13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56, and four used liver tissue35,57,58,59. We pooled the prevalence of OBI for blood and liver tissue samples and compared the results between HBcAb-positive and HBcAb-negative populations. In blood samples, the overall pooled prevalence of OBI in HBsAg-negative population was 5% (95% CI: 3–8%). The pooled prevalence was significantly higher in HBcAb-positive population than in the HBcAb-negative population (7%, 95% CI: 3–11%, I²=98% vs. 2%, 95% CI: 0–4%, I²=99%; p = 0.02) (Fig. 2A).

Fig. 2
Fig. 2Fig. 2
Full size image

(A) Pooled prevalence of OBI in HBcAb-positive and HBcAb-negative group with blood sample detection. (B) Pooled prevalence of OBI in HBcAb-positive and HBcAb-negative group with liver tissue sample detection. (C) Prevalence of occult HBV carrier in HBcAb-positive and HBcAb-negative population with blood sample and liver tissue detection.

In liver tissue samples, which were exclusively sourced from 228 organ donors, the pooled prevalence also differed significantly by HBcAb status. All liver tissue samples were sourced from organ donors, who represent a screened, low-risk population. A prevalence of 66% (95% CI: 36–91%, I²=76%) was observed in the HBcAb-positive group, compared to 0% (95% CI: 0–2%, I²=0%) in the HBcAb-negative group (p < 0.01) (Fig. 2B). Figure 2C provides an overview of the pooled prevalence across all groups and sample types. Analysis of publication bias using a funnel plot revealed asymmetry for the overall OBI prevalence, suggesting a potential underreporting of studies with null or low prevalence (Figure S1-A). However, Egger’s test showed no significant bias within the HBcAb-negative subgroup (P > 0.05, Figure S1-B).

Subgroup analyses of OBI in HBcAb-positive population

We performed subgroup analyses to explore sources of heterogeneity in OBI prevalence in the HBcAb-positive population (Table 2). The prevalence of OBI was significantly higher in regions with moderate-to-high HBV endemicity compared to low-endemicity regions (10%, 95% CI: 4–18%, I²=98% vs. 3%, 95% CI: 1–7%, I²=97%; p = 0.05) (Figure S2-A). Similarly, OBI prevalence increased with the regional HBcAb-positive rate (p = 0.05) (Figure S2-B).

Table 2 Subgroup analyse of OBI in HBcAb-positive population.
Full size table

When comparing detection methods, the OBI rate was numerically higher with nested PCR (9%, 95% CI: 4–17%, I²=96%) than with real-time qPCR (5%, 95% CI: 2–8%, I²=97%), but this difference was not statistically significant (p = 0.17) (Figure S2-C). No significant difference was found based on HBsAb status (p = 0.96) or population source (blood donors vs. other general population; p = 0.82) (Figure S2-D, Figure S2-E). An analysis by publication year showed a non-significant decreasing trend in OBI prevalence over time (p = 0.44) (Figure S2-F).

Subgroup analyses of OBI in HBcAb-negative population

In the HBcAb-negative population, subgroup analyses revealed no significant difference in OBI prevalence based on HBV endemicity (p = 0.30) (Table 3, Figure S3-A). However, a statistically significant difference in prevalence was observed across categories of the regional HBcAb-positive rate (p < 0.01) (Figure S3-B).

Table 3 Subgroup analyse of OBI in HBcAb-negative population.
Full size table

The detection rate with nested PCR (4%, 95% CI: 0–10%, I²=94%) was numerically higher than with real-time qPCR (1%, 95% CI: 0–2%, I²=73%), though the difference was not statistically significant (p = 0.11) (Figure S3-C). No significant differences were observed based on HBsAb status (p = 0.91) or population source (p = 0.39) (Figure S3-D, Figure S3-E). Notably, OBI prevalence in this group decreased significantly over the publication periods (p = 0.04) (Figure S3-F).

Sensitivity analysis

Given the high heterogeneity, we performed a leave-one-out sensitivity analysis. The result showed that the sequential omission of any single study did not significantly alter the pooled prevalence estimates for either the HBcAb-positive or HBcAb-negative populations (Figure S4-A, Figure S4-B), confirming the robustness of our findings.

Meta regression analysis

To further investigate the substantial heterogeneity, we conducted a meta-regression analysis. However, none of the tested study-level covariates (regional HBV prevalence, HBcAb-positive rate, detection method, HBsAb status, publication year, or population source) were identified as significant sources of heterogeneity in either the HBcAb-positive or HBcAb-negative groups (p > 0.05 for all variables; Tables S4 and S5).

Discussion

Despite a global decline in chronic HBV infection prevalence, hepatitis B remains a significant public health challenge60. HBcAb IgG indicates past HBV exposure and often remains after the virus clears. In HBV-endemic regions, up to 40% test positive for HBcAb3,34,52. Better detection and research advances have highlighted a connection between OBI and HBcAb. Although the clinical relevance of OBI is recognized, its prevalence, clinical significance, and clinical implications in the general HBcAb-positive population remain poorly characterized. This knowledge gap highlights the need for further research to clarify its public health impact.

To our knowledge, this is the first systematic review and meta-analysis to comprehensively assess the global prevalence of OBI in the general population with HBcAb positivity. Our findings demonstrate a significant disparity in OBI prevalence based on HBcAb status, with a markedly higher OBI prevalence in HBcAb-positive individuals than in HBcAb-negative controls (7% vs. 2%; p = 0.02). This disparity was most pronounced in liver tissue analyses, with an OBI prevalence of a remarkable 66% in HBcAb-positive individuals versus 0% in HBcAb-negative controls. This striking finding, primarily from rigorously screened organ donor cohorts (universally HBsAg-negative and often serum HBV DNA-negative), is not paradoxical but biologically expected. This observation powerfully illustrates the known biology of HBV persistence: the liver serves as the primary viral sanctuary. Even after HBsAg clearance, the viral template can persist within hepatocytes as a stable reservoir4. Blood-based screening is designed to detect circulating virus, but it cannot “see” the silent HBV DNA sequestered within the liver. Consequently, intrahepatic OBI prevalence exceeds those in peripheral blood, where viral loads may fluctuate below detection thresholds9. This highlights that the high prevalence found in liver samples may more accurately reflect the true burden of persistent HBV infection in individuals with resolved hepatitis B. Therefore, HBsAg-negative/HBcAb-positive population represents a key group with a substantial burden of OBI. In high-risk clinical settings, such as immunosuppressive therapy planning, solid organ transplantation, or management of chronic liver disease, routine OBI screening and careful evaluation are warranted for this population to mitigate potential HBV reactivation or transmission.

Our subgroup analysis also revealed a noteworthy finding regarding population sources. The prevalence of OBI did not significantly differ between blood donors and the general population cohorts (p = 0.82), a result that might seem counterintuitive given that donors undergo screening. A key explanation for this is that the blood donor population in our meta-analysis is a heterogeneous mix, likely dominated by or heavily influenced by first-time donors whose OBI risk profile has not yet been shaped by prior screening and thus more closely mirrors that of the general population. This insight underscores that the underlying burden of OBI is substantial even in presumably low-risk settings and highlights the importance of including donor populations in epidemiological surveillance.

The prevalence of OBI exhibits significant regional heterogeneity, driven primarily by local HBV endemicity, which necessitates enhanced surveillance and targeted screening strategies in high-risk areas. While HBcAb screening is important for identifying populations with a higher OBI burden, its clinical utility is limited by diagnostic uncertainty, as positivity cannot definitively confirm OBI and may lead to the unnecessary exclusion of safe donors. To address these challenges, two promising diagnostic advancements warrant consideration. First, emerging quantitative HBcAb assays could offer more nuanced information for guiding screening strategies58,61,62. Second, the use of highly sensitive HBsAg assays is vital, as a number of studies have shown their ability to detect minute levels of HBsAg in individuals previously classified as OBI63,64, thereby resolving their infectious status and reducing diagnostic ambiguity. Despite these advances, the reality remains that approximately 30% of countries lack even standardized HBcAb testing in blood donor screening programs65, highlighting the challenge of balancing transfusion safety with blood supply. Therefore, developing optimized OBI prevention protocols that strategically incorporate these newer, more sensitive assays while considering regional cost-effectiveness is a key future priority.

The absence of standardized diagnostic protocols presents a substantial obstacle in the accurate detection of OBI, a challenge that extends far beyond the general choice between nested or real-time PCR. In this study, while the p-value was not significant (likely due to low statistical power from high heterogeneity), the numerical difference (e.g., 9% vs. 5%) is substantial and clinically relevant, highlighting the impact of methodological sensitivity. The true sensitivity of any assay is profoundly influenced by a range of critical, yet often underreported, technical parameters. For instance, pre-analytical factors such as the DNA extraction method and the initial sample volume can dramatically impact the ability to detect the low viral loads characteristic of OBI. Furthermore, the assay design itself is paramount, including the choice of the target gene region and the strategic use of multiple replicates66. The inconsistent reporting of these details in the literature, including the frequent lack of a standardized limit of detection (LoD), not only contributes significantly to the high statistical heterogeneity observed in our analysis but also makes a direct quantitative comparison of assay sensitivities unreliable. This methodological variability underscores why liver tissue remains the gold standard and highlights the urgent need for highly sensitive blood-based assays. While emerging technologies like ddPCR offer enhanced precision58, their ultimate effectiveness still hinges on these fundamental parameters. Therefore, a multimodal diagnostic approach is essential, and moving forward, it must be coupled with a transparent reporting standard for NAT assays. This is essential for navigating the inherent tension between achieving maximal analytical sensitivity and ensuring a standard has meaningful clinical utility and global applicability, especially in high-endemic, resource-limited regions. We therefore strongly advocate that future studies report not just the method used, but also the validated LoD in IU/mL and full details of the assay’s design, which is crucial for improving detection accuracy and ultimately enabling meaningful comparisons across studies.

The prevalence and clinical implications of OBI in individuals positive for both HBcAb and HBsAb requires careful consideration. While isolated HBcAb has been increasingly linked to OBI67, the protective role of HBsAb in dually seropositive individuals remains unclear5,6. In our study, where HBsAb-positivity was defined by the standard protective threshold (≥ 10 mIU/mL), OBI prevalence did not significantly differ between HBsAb-positive and -negative populations. This finding, however, likely reflects the limitations of a binary stratification, which overlooks the quantitative nature of the antibody response. HBsAb titers and OBI prevalence may have a dose-dependent relationship; even low, non-protective titers may offer some viral control, while higher levels may confer more robust protection68,69. Therefore, while our analysis based on the standard cutoff provides a broad overview, future studies utilizing quantitative HBsAb data are crucial to delineate this relationship and refine clinical guidelines for populations such as blood donors.

Limitations

This meta-analysis offers valuable insights into OBI prevalence in HBcAb-positive individuals but has limitations. The scarcity of studies using liver tissue, the gold standard for OBI diagnosis, hindered detailed subgroup analyses due to ethical and practical issues with liver biopsies. A primary limitation of this meta-analysis is the exceptionally high heterogeneity observed in the pooled prevalence of OBI, with I² values often exceeding 90%. This indicates substantial variability among the included studies that could not be fully explained by our subgroup analyses. We performed a meta-regression to formally test for sources of this variation, but no single factor was identified as a statistically significant contributor. This suggests that the heterogeneity is likely multifactorial and stems from a complex interplay of sources that are difficult to quantify. Importantly, this high level of heterogeneity is, in itself, a significant finding, highlighting the urgent need for standardized methodologies and more detailed reporting in future OBI research.

Conclusion

This study demonstrates a significant disparity in the prevalence of OBI based on HBcAb status. The prevalence of OBI is substantially higher in the HBsAg-negative/HBcAb-positive population compared to their HBcAb-negative counterparts, a finding that is most pronounced in liver tissue, where the pooled prevalence in HBcAb-positive individuals exceeded 50%. Furthermore, the prevalence of OBI was higher in regions with greater HBV endemicity. Notably, OBI was detected in HBcAb-positive individuals regardless of their HBsAb status, with no significant difference in prevalence found between HBsAb-positive and HBsAb-negative groups. Collectively, these findings underscore that the HBcAb-positive population carries a considerable burden of OBI and highlight the urgent need for enhanced diagnostic strategies and further research into the mechanisms of viral persistence to inform clinical and public health policies.