Background
Despite the implementation of universal vaccination programs, hepatitis B virus (HBV) infection of children remains a significant global public health challenge [
1‐
3]. According to the report from WHO in 2015, the prevalence of hepatitis B surface antigen (HBsAg) positivity among children under the age of 5 was estimated to be approximately 1.3% on a global scale [
3] and it was estimated to be 0.7% in 2022 [
4]. While chronic HBV infection typically follows a benign disease course during childhood, it is noteworthy that a proportion of children (3%-5%) may still experience progression to cirrhosis, and 0.01%-0.03% may develop hepatocellular carcinoma (HCC) in adulthood [
5]. Therefore, children with chronic hepatitis B (CHB) require active antiviral therapy to delay the disease progression and minimize the risk of long-term adverse outcomes [
6]. The optimum result of antiviral therapy for CHB is commonly considered a functional cure, marked by the absence of HBsAg and undetectable HBV-DNA in the serum, with or without the seroconversion of hepatitis B surface antibodies (HBsAb) [
7]. The pursuit of a functional cure should be given high priority for certain select patients.
Serologic patterns, encompassing various serologic markers of HBV, are of utmost importance in the diagnosis and evaluation of CHB. HBsAg is a crucial serologic marker that serves as a significant indicator of ongoing HBV infection. However, HBsAb is an antibody with specific protective properties that can neutralize HBsAg. Consequently, HBsAb positivity often signifies the recovery of HBV infection and the subsequent loss of HBsAg. There has been a growing number of studies on the coexistence of HBsAg and HBsAb in clinical practice since its initial report in 1976 [
8‐
13]. This atypical serological pattern in HBV infection lacks a well-established understanding of its underlying mechanism and clinical significance [
12,
14‐
16]. Several studies have indicated a potential association between this serological profile and significant clinical implications, such as advanced fibrosis, HCC, and liver failure [
17‐
20]. Nevertheless, prior research has primarily concentrated on adult populations, with limited data on children. Furthermore, the impact of the coexistence of HBsAg and HBsAb on antiviral treatment responses remains unknown. Therefore, there is a compelling need to conduct further research in this area.
In the current study, we sought to analyze the characteristics of children with HBsAb-positive CHB. Furthermore, we explored the potential correlations between this serological profile and the clinical outcomes of antiviral treatment, providing an important basis for decision-making in the therapy of CHB.
Methods
Study population
This retrospective cohort study included children and adolescents, aged 1–17 years, with treatment-naïve CHB who received antiviral treatment at Hunan Children’s Hospital between June 2016 and April 2023. The diagnosis of CHB adhered to the guidelines on prevention and treatment for chronic hepatitis B (2022 version) [
6] including HBsAg positive for over 6 months, persistently or repeatedly abnormal ALT levels, or indications of significant inflammation, necrosis, or significant fibrosis in liver histology. Individuals with nonalcoholic hepatic steatosis or other known causes of chronic liver diseases; coinfection with HCV, HDV, HIV, or cytomegalovirus; concomitant presence of other cancerous neoplasms; insufficient available data; or duration of antiviral treatment of less than 6 months were excluded from the study. Approval for this study was granted by the ethics boards of Xiangya School of Public Health Central South University (XYGW-2023–123). Owing to its retrospective design, there was an exemption from informed consent.
Definitions and outcomes
The coexistence of HBsAg and HBsAb was determined through simultaneous positivity for those two markers. The participants of the study were categorized into two distinct groups: the HBsAb-positive group consisting of patients who tested positive results for both HBsAg and HBsAb at baseline, and the HBsAb-negative group comprising patients who were HBsAg single-positive at baseline. A positive threshold was established for the HBsAb titer, defined as greater than 10 IU/L (1 log10 IU/L). During the follow-up period, regular examinations of virologic and biochemical biomarkers were conducted. The primary outcome was the achievement of HBsAg loss, which was defined as having two consecutive measurements under 0.05 IU/mL. Additional outcomes included HBeAg clearance, which was defined as attaining HBeAg level below 1 COI, and HBV DNA undetectability, defined as achieving DNA level below 100 IU/mL. The follow-up duration was determined by measuring the time elapsed from the initiation of enrollment until the occurrence of endpoints, loss to follow-up, or the concluding follow-up date (April 2023).
Laboratory testing
Serologic markers for HBV were assessed by Electrochemiluminescence assay (Roche Diagnostics GmbH, Mannheim, Germany). HBV DNA levels and HBV genotype were assessed using the Quantitative Fluorescence Diagnostic HBV Kit and HBV Genotype Kit (Sansure Biotech, Changsha, China), respectively. The alanine aminotransferase (ALT) and aspartate aminotransferase (AST) were measured by the Bayer-2400 automatic biochemistry analyzer. Experienced pathologists, who were unaware of the corresponding clinical data, performed the assessment of liver biopsy samples. The stages of liver fibrosis and grades of inflammation were determined using the Scheuer scoring system, which categorized them into five stages (0–4).
Statistical analysis
Continuous variables were condensed by presenting mean and standard deviation, or median and interquartile range, based on the data's distribution characteristics. Categorical variables were expressed by frequencies and percentages. For the comparison of different categories of variables, statistical tests such as t-tests, Mann–Whitney U tests for continuous variables and chi-square tests for categorical variables were utilized. The crude rates of clinical outcomes were presented in person-years (PYs). The 95% confidence intervals (CIs) of relative incidence were obtained using Poisson regression. Kaplan–Meier curves were utilized to estimate the cumulative incidence of clinical outcomes, and the log-rank test was employed to compare disparities between the two groups. Cox proportional hazard regression was used to evaluate the associations between the coexistence of HBsAg and HBsAb with clinical treatment outcomes, while subgroup analyses were conducted according to HBV genotype, inflammation grade, and fibrosis stage. Using restricted cubic spline (RCS) plots within Cox proportional hazard models, we examined the associations of HBsAb levels with all concluding events, considering HBsAb as a continuous scale. To assess the predictive effect of HBsAb, receiver operating characteristic curve (ROC) was conducted, with the optimal cutoff value pinpointed via the maximum of the Youden index. The statistical analyses and visualization were conducted utilizing R software (version 4.2.2) and Graphpad Prism (version 9.0), with a significance level of P < 0.05.
Discussion
In this study, 23.39% (87/372) of children with CHB were tested positive for HBsAb. In addition, these children demonstrated higher rates of achieving favorable outcomes, including seroclearance of both HBsAg and HBeAg, as well as the undetectability of HBV DNA, compared to those children who are negative for HBsAb. Importantly, the strength of the association increased with higher levels of HBsAb.
Despite it has been about half a century since this serological pattern was first reported [
8], the molecular mechanisms underlying this phenomenon are still unclear. The current mainstream hypotheses included mutations in the viral genome and hosts immune status [
12,
15]. Previously reported rates of coexistence of HBsAg and HBsAb vary widely, ranging from 0.3% to over 30% [
10,
11,
15,
18,
21,
22]. The coexistence of HBsAg and HBsAb (23.39%, 87/372) in our study was comparable with the incidence (20.67%, 234/1132) reported by Wang et al. [
23]. However, the incidence of coexistent HBsAg and HBsAb among children with chronic HBV infection was only 0.81% (1/124) in North America [
24], which is much lower than our result. A large real-world study from China revealed that the occurrence rate was 4.24% (277/6534) among treatment-naive adults with CHB [
20]. Additionally, another study from China indicated a prevalence of 47.30% (205/433) for this phenomenon in the population with CHB [
18] Possible reasons for this difference in incidence include heterogeneity of study populations, different disease phases and the method of detection [
25‐
27].
The analysis of clinical characteristics in our study revealed that children with HBsAb-positive CHB had lower levels of HBsAg and HBV DNA, which aligns with results from previous studies [
28,
29]. The possible explanation is that HBsAb could neutralize HBsAg, thereby reducing the levels of HBsAg. Moreover, the generation of HBsAb signifies a certain degree of immune response within the host, leading to a reduction in the levels of viral DNA [
28,
30]. Additionally, these children with HBsAb displayed elevated levels of ALT and AST, which may be attributed to a more pronounced immune response against HBV due to the presence of HBsAb. Because the vigorous immune response usually leads to extensive liver inflammation and subsequent hepatocellular damage, thereby contributing to increased ALT and AST levels.
Previous studies demonstrated that HBsAb may promote clearance of HBV and achieve functional cure via multiple mechanisms [
31]. Our results further support these findings, as we observed significantly higher incidence rates of HBsAg loss, HBeAg seroclearnace, and undetectable HBV DNA among children with HBsAb. This observation aligns with a Korean study that showed that individuals with this serological pattern exhibited a higher rate of HBsAg clearance compared to those without HBsAb [
29]. Similarly, some reports documented that patients with coexisting of HBsAg and HBsAb had higher probabilities to achieve clearance of HBsAg, HBeAg, and HBV DNA after treatment with NAs or interferon [
32,
33]. Furthermore, it has been widely reported that baseline higher ALT levels, lower levels of HBV DNA and HBsAg are predictive indicators for the effectiveness of antiviral therapy [
34,
35]. This may partly explains the increased clearance rates of HBsAg and HBeAg in children with coexistence of HBsAg and HBsAb. Moreover, the phenomenon that children with HBsAb-positive CHB were more likely to attain HBsAg loss seemed to hint the coexistence of HBsAg and HBsAb may be an intermediate transition in the clearance of HBsAg.
The AUC of the prediction model was estimated to be 0.63 when we used HBsAb level to predict HBsAg loss. Combining with age, the value of AUC improved to 0.71. A combination of HBsAb ≥ 0.84 log10 IU/L and age ≤ 5 can help clinical physicians identify patients likely to achieve HBsAg loss after antiviral therapy. This suggests that younger patients who had relatively high levels of HBsAb have a greater likelihood of achieving HBsAg loss, which is a key milestone in achieving the goal of hepatitis B elimination. Therefore, timely diagnosis and early initiation of antiviral treatment in pediatric patients with high HBsAb levels can significantly improve treatment outcomes.
Our study deserves attention as this research marks the inaugural investigation into the association between the coexistence of HBsAg and HBsAb and clinical treatment outcomes in children with CHB. Inevitably, the current study has several limitations. Firstly, although this study investigated the longitudinal outcomes of children with HBsAb, the underlying mechanism explaining the increased rates of favorable outcomes remains unclear. Further research is needed to fully elucidate this. Secondly, age may serve as a confounding factor, although it has been appropriately adjusted in multivariate analysis, which may lead to selection bias. Thirdly, this study was conducted at a single hospital. The generalizability of the findings to different clinical settings may be limited. Therefore, to validate the results and enhance the external validity of the findings, it is recommended to conduct multicenter studies with larger sample sizes.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.