Relationship Between Serum Hepatitis B Core-Related Antigen and Hepatitis B Surface Antigen in Hepatitis B Cirrhosis Patients With HBV-DNA Negative Status: A Cross-Sectional Study


 Background: The relationship between serum hepatitis B surface antigen (HBsAg) and HBV core-related antigen (HBcrAg) titers, which is especially clinically important in monitoring virus replication activity and progression of liver disease, is not thoroughly studied in patients with hepatitis B cirrhosis, particularly in HBV-DNA negative patients. This retrospective study explored the relationship between serum HBcrAg and HBsAg titers in hepatitis B cirrhosis patients with HBV-DNA negative status. Methods: We analyzed data and samples from 180 patients diagnosed with HBV liver cirrhosis. Statistical analyses included baseline characteristics, univariate analysis, stratification analysis, three different analytical models, Spearman’s correlation analysis, generalized additive model, and two-piecewise linear regression model. Results: After adjusting for confounders (age, sex, primary hepatic carcinoma, HBeAg, Child–Pugh class, alcoholic liver disease, Diabetes mellitus , and type and duration of use of anti-HBV agents), a non-linear relationship was found between HBsAg and HBcrAg. The inflection point was 0.58. The effect sizes, confidence intervals, and P value on the left and right sides of the inflection point were 0.10, (–0.23–0.42), 0.555 and 0.62, (0.46–0.78), ＜0.0001, respectively. There was statistical significance (all p<0.01) between HBcrAg and HBsAg among all the stratification variables except for the "others" group in type of anti-HBV agents. Spearman’s correlation coefficient between HBsAg and HBcrAg was 0.613 (P<0.0001). Conclusions: There was a clear non-linear relationship between HBcrAg and HBsAg in patients with hepatitis B cirrhosis who were HBV-DNA negative, and the correlation between serum HBcrAg and HBsAg was moderate.


Background
The increasing popularity of the hepatitis-B virus (HBV) vaccine combined with hepatitis B immunoglobulin and the application of powerful anti-HBV agents has led to considerable progress in the prevention and treatment of HBV infection [1,2]. However, even after patients test negative for serum HBV-DNA, some may still develop cirrhosis and liver cancer. Most patients treated with nucleoside/tide analogues (NA) continue to have detectable covalently closed circular DNA (cccDNA), although most of them have undetectable serum HBV DNA [3]. Therefore, continuous surveillance of HBV replication activity is one of the most important means of monitoring disease progression. Because detection of intrahepatic cccDNA requires a liver biopsy, identifying novel and convenient surrogate markers of intrahepatic viral replication activity has been a research hotspot in the past decades [4], especially in HBV-DNA negative patients. Both serum hepatitis B surface antigen (HBsAg) and HBV core-related antigen (HBcrAg) titers can be used as virological markers in HBV-DNA negative patients. The relationship between HBsAg and HBcrAg has already been well studied in chronic hepatitis B (CHB) [5][6], but the correlation has not yet been characterized in patients with hepatitis B cirrhosis who are HBV-DNA negative. Therefore, we explored the relationship between serum HBcrAg and HBsAg in this population. Understanding this relationship can help in further exploration which is more accurate and its applicability in HBV DNA surveillance in this population. Moreover, this can be used as a foundation for future studies to establish prediction models of hepatic adverse events, especially in the case of liver cirrhosis.

Methods
Patients This was a cross-sectional study of HBV-DNA-negative patients diagnosed with hepatitis B cirrhosis who were hospitalized at the Third Central Clinical College of Tianjin Medical University between October 2018 and October 2019. All patients were hepatitis B cirrhosis with HBV-DNA negative status (HBV NDA 20 IU/mL), regardless of the type and duration of use of anti-HBV agents. Participants with hepatitis C or human immunode ciency virus (HIV) co-infection and those missing virological data were excluded from our analyses, as were those without surplus serum samples for HBcrAg measurement. Patients with serum samples not appropriately stored at -20°C were also excluded. Patients with alcoholic liver disease (ALD), primary hepatic carcinoma, and/or Diabetes mellitus were not excluded in our study.

HBV serological markers
Serum HBsAg quanti cation HBsAg levels were quanti ed using the Abbott ARCHITECT i4000SR chemiluminescent microparticle immunoassay (Abbott Diagnostics, Abbott Park, IL, USA), which has a detection range of 0.05-250 IU/mL. The samples with concentrations >250 IU/mL were diluted and retested.
HBeAg and HBV-DNA quanti cation HBeAg levels was tested using commercially available enzyme immunoassay kits (Abbott Diagnostics, Abbott Park, IL, USA). HBV-DNA in serum samples was detected and quanti ed using the Amplly real-time PCR HBV assay performed on Anadas9850 (Amplly Engineering co. Ltd., Xiamen, CN), a fully automated nucleic acid extraction and real-time PCR system. With the Amplly assay, HBV DNA was extracted from 200 μL serum and ampli ed using the Anadas9850. The limit of detection (LOD) was ~20 IU/mL. Quanti ed samples (>5.0×10 8 IU/mL) were diluted and retested, while samples with <20 IU/mL were regarded as undetectable.
HBcrAg quanti cation Surplus serum samples from patients were stored at -20°C and measured using a fully automatic chemiluminescent enzyme immunoassay system (CLIA; Lumipulse G1200, Fujirebio, Tokyo, Japan), according to the manufacturer's instructions. Brie y, the assay provided a linear range of 3-7 log 10 U/mL. However, the machine indicates levels <3 log U/mL down to 2 log U/mL in HBcrAg-positive samples.
HBcrAg levels <2 log U/mL were treated as 2 log U/mL for statistical analysis. Samples with HBcrAg levels >7 log 10 U/mL were diluted with the sera of healthy controls and retested to quantify HBcrAg levels.

Diagnosis of cirrhosis
Cirrhosis was diagnosed according to the Chinese Guidelines on the Management of Liver Cirrhosis (2019 version) released by the Chinese Society of Hepatology, Chinese Medical Association.

Statistical analysis
Measured levels of serum HBcrAg and HBsAg were log 10 transformed. Continuous variables , when their distributions were normal, were expressed as the mean ± standard deviation. While the distributions were skewed, the forms of expression were median (interquartile range). They were expressed as percentage or frequency when categorical variables appeared. Kruskal-Wallis H (when the data were skewed distributions), One-way ANOVA (when the data were normal distributions), and chi-square test (when the data were categorical variables) were used to determine statistical differences between the proportions and means of the groups. Univariate linear regression model was used to evaluate the associations between HBcrAg and HBsAg. Both non-adjusted and multivariate adjusted models are presented in our study. We simultaneously showed the results of unadjusted, minimally adjusted, and fully adjusted analyses. Whether the covariances were adjusted determined by the following principle: if the covariance changed the matched odds ratio at least 10% when added to the model, the covariance was adjusted [7]. Besides, we also used the generalized additive model (GAM) to identify the existence of non-linear relationship between HBcrAg and HBsAg. When the ratio between HBcrAg and HBsAg appeared in the form of smoothed curve, the in ection point was calculated by the recursive method, where the method of the maximum model likelihood was used [8]. The analyses of subgroup were performed by strati ed linear regression models. The interaction and modi cation of subgroups were tested by the method of the likelihood ratio test. Spearman's rank correlation analysis was also performed. All data were analyzed using the Statistical packages R (The R Foundation; http://www.r-project.org; version 3.4.3) and Empower (R) (www.empowersATTs.com, X&Y solutions Inc., MA, USA). For all analyses, P<0.05 (two-sided) was considered to indicate statistical signi cance [9] Results Baseline characteristics In this study, we analyzed 180 eligible HBV-DNA negative patients diagnosed with hepatitis B liver cirrhosis. Of these, 82 were diagnosed with primary hepatic carcinoma with hepatitis B liver cirrhosis. The mean age of the patients was 58.62±9.81 years and 82.78% were male. The mean HBsAg and HBcrAg levels were 2.21±1.33 and 3.85±1.33 log 10 U/mL, respectively. To clarify the relationship between serum HBsAg and HBcrAg, patients were divided into two groups according to HBeAg levels. Baseline characteristics of the patients are summarized in Table 1. Because type of anti-HBV agents vary widely other than entecavir ETV -treated group, the study patients were divided into three groups, based on variability of anti-HBV agents: Treatment-naïve group, ETV-treated group, and other anti-HBV agents group.

Univariate analysis
The results of univariate analysis are presented in Table 2. HBsAg, HBeAg, and anti-HBV agents were strongly correlated with HBcrAg. The data show that continuous variables including age, Platelets PLT , aspartate aminotransferase AST , total bilirubin TBIL , International Normalized Ratio INR , and albumin ALB were negatively correlated with HBcrAg levels, and among categorical variables, sex (female), Child-Pugh class (B vs. A), and Diabetes mellitus (with vs. without) were negatively correlated with HBcrAg levels. All β values were ≤0.001.

Subgroup analyses among strati cation variables
Because 13 patients who were HBsAg showed seroclearance in our study, the total number of participants was 167 for each strati cation variable. The variation trend of β was consistent in all the strati ed variables including sex, primary hepatic carcinoma, HBeAg, family history of hepatitis B, ALD, Diabetes mellitus, Child-Pugh class, Barcelona Clinic Liver Cancer (BCLC), and type of use of anti-HBV agents. There was statistical signi cance (p<0.05) between HBsAg and HBcrAg among all the strati cation variables excluding the others group in type of anti-HBV agents after adjusting AST, HBeAg, family history of hepatitis B, Child-Pugh class, ALD, anti-HBV agents, and duration of use of antiviral agents in each strati cation variable except for when itself was regarded as strati cation variable.

Correlation between HBsAg and HBcrAg in variable primary hepatic carcinoma
Regardless of whether the population was diagnosed with primary hepatic carcinoma, the correlation between HBsAg and HBcrAg was positively and statistically signi cant (all P<0.01) after adjusting for AST, HBeAg, family history of hepatitis B, Child-Pugh class, ALD, type of use of anti-HBV agents, and duration of use of anti-HBV agents. The β and 95% con dence interval (CI) were 0.41 (0.20-0.62) and 0.57 (0.43-0.71) respectively in the primary hepatic carcinoma and cirrhosis groups.

Correlation between HBsAg and HBcrAg in variable HBeAg
The β, 95% CI, and P values were 0.59, (0.25-0.94), 0.0023 and 0.49, (0.36-0.62), <0.0001, respectively, in the HBeAg positive and negative groups. In this strati cation variable, AST, HBeAg, family history of hepatitis B, Child-Pugh class, ALD, and type and duration of use of anti-HBV agents were also adjusted.

Correlation between HBsAg and HBcrAg in variable Child-Pugh Class
Only nine patients were diagnosed with Child-Pugh class C in our study; thus, the β, 95% CI, and P values were not statistically analyzed in Child-Pugh class C. The correlation between HBsAg and HBcrAg was positively and statistically signi cant (all P<0.01) after adjusting for AST, HBeAg, family history of hepatitis B, ALD, anti-HBV agents, and duration of use of anti-HBV agents.

Correlation between HBsAg and HBcrAg in variable BCLC
The β, 95% CI, and P values were 0.50, (0.10-0.90), 0.0214 and 0.52, (0.22-0.81), 0.0071, respectively, in stage A and B of BCLC; further, AST, HBeAg, family history of hepatitis B, ALD, and type and duration of use of anti-HBV agents were also adjusted. Because of the lack of adequate patients in stages 0, C, and D, statistical analysis could not be performed.

Correlation between HBsAg and HBcrAg in variable type of use of anti-HBV agents
Because type of use of anti-HBV agents vary widely other than type of use of ETV, the study patients were divided into three groups according to categories of anti-HBV agents. The correlation between HBsAg and HBcrAg was positively and statistically signi cant (all P<0.01) except for the others group in type of anti-HBV agents after adjusting for AST, HBeAg, family history of hepatitis B, ALD, and type and duration of use of anti-HBV agents.
There also showed the results of subgroups of other categorical variables including sex, ALD, Diabetes mellitus, and family history of hepatitis B. The detailed results of analysis are presented in Table 3.

Analysis of the relationship between HBcrAg and HBsAg levels
To clarify the relationship between serum HBcrAg and HBsAg levels and its trend, patients were divided into three groups based on the value of HBsAg: T1 (-2.00 to -2.01 log 10 U/mL), T2 (2.04-3.04 log 10 U/mL), and T3 (3.06-3.95 log 10 U/mL). A univariate linear regression model was used to evaluate the correlation between HBcrAg and HBsAg. The non-adjusted and adjusted models are shown in Table 4. In the crude model, HBcrAg was correlated with HBsAg (β=0.59, 95% CI: [0.47-0.70], P<0.0001), and the results using the preliminary adjusted model (adjusted only for age and sex) were similar with the crude model (β=0.59, 95% CI: [0.47-0.71], P<0.0001). We also detected an association using adjusted model (β=0.48, 95% CI: [0.37-0.60], P<0.0001), which was adjusted for AST, HBeAg, type and duration of use of anti-HBV agents, primary hepatic carcinoma, Child-Pugh class, ALD, and Diabetes mellitus. For sensitivity analysis, we considered HBsAg to be a categorical variable (Tripartite), and a statistically signi cant trend was observed (all P values for trend<0.0001).

Analyses of non-linear relationships
In this study, we analyzed the non-linear relationship between HBsAg and HBcrAg( Fig. 1), as HBsAg is a continuous variable. Our data showed that the relationship between HBcrAg and HBsAg was non-linear in a generalized additive model (after adjusting for sex, age, primary hepatic carcinoma, HBeAg, Child-Pugh class, ALD, Diabetes mellitus, and type and duration of use of anti-HBV agents). By using a two-piecewise linear regression model, we calculated that the in ection point was 0.58. On the left of the in ection point, the effect size, 95% CI, and P value were 0.10 (-0.23, 0.42), and 0.555. We also observed a positive relationship between HBsAg and HBcrAg on the right side of the in ection point (0.62, 0.46-0.78, and <0.0001) ( Table 5).
Analyses of Spearman's correlation coe cient Because HBcrAg and HBsAg showed non-normal distribution, The Spearman's rank correlation analysis was performed rather than pearson, which was 0.613 (P<0.0001) in the whole population.
HBsAg is already widely applied as a virological monitoring marker in clinical practice. In recent times, HBcrAg has also been considered as a tool to monitor disease progression [13] and proved a novel serological marker of HBV replication [18]. Therefore, it is important to explore the relationship between the two serological markers in order to understand their respective clinical values, particularly in patients with hepatitis B cirrhosis who have HBV-DNA negative status.
The correlation between HBsAg and HBcrAg has already been well studied in CHB [5][6]. Our study showed that HBsAg was positively associated with HBcrAg after adjusting for other covariates in hepatitis B cirrhosis patients with HBV-DNA negative status. The β value remained stable in the crude and adjusted models after regression analysis; furthermore, we found that the relationship between them was nonlinear. To our knowledge, this is the rst study on the correlation between HBsAg and HBcrAg in hepatitis B cirrhosis patients with HBV-DNA negative status.
Single factor analysis showed that HBsAg was signi cantly correlated with HBcrAg (β=0.59, 95% CI: [0.47-0.71], P<0.0001) in hepatitis B cirrhosis patients with HBV-DNA negative status. Although age, HBeAg, Diabetes mellitus, and type of anti-HBV agents also showed positive correlation with HBcrAg (all P<0.05), the β value did not uctuate obviously in the multivariate linear regression models after adjusting other covariates. The Spearman's correlation coe cient was 0.613 (P<0.0001), which means there was moderate correlation between HBsAg and HBcrAg in this population, but weaker than that found in treatment-naive CHB patients [5]. Seto et al. found that the serum HBcrAg correlated strongly with serum HBsAg levels (r=0.703, P<0.001) in Asian treatment-naive CHB patients [5]. Previous studies also showed that the correlation between HBsAg and HBcrAg can be weakened by anti-HBV agents. The serum HBcrAg level was positively associated with serum HBsAg level (r=0.713) at baseline in all CHB patients, but this correlation weakened remarkably during treatment [19]. Another study, mainly focused on HBeAg-positive CHB patients, showed that serum HBcrAg was signi cantly correlated with HBsAg (r=0.696, P<0.001) at baseline, but this correlation weakened at 96 weeks after nucleotide analog (NA therapy (r=0.452, P<0.001) [20].
The conclusions of the previous studies were not always consistent. Another study, which mainly focused on receiving pegylated interferon-based therapy for CHB patients, showed that HBcrAg did not correlate with HBsAg even at baseline [21]. Yet another study suggested that HBcrAg weakly correlated with qHBsAg levels (Spearman's correlation coe cient (r=0.471), P<0.0001) in treatment-naive HBeAgnegative patients [22]. There is a weaker correlation between HBcrAg and HBsAg in the natural history of patients with hepatitis B virus infection [6].
Previous studies have shown that the correlation between HBsAg and HBcrAg was also affected by HBeAg in CHB[22-23]. Chen En-Qiang et al. showed that HBcrAg correlated positively with HBsAg (r=0.564, P<0.001) among HBeAg-positive patients rather than HBeAg-negative patients in treatmentnaive CHB[23]. Our study showed a moderate correlation between HBcrAg and HBsAg in hepatitis B cirrhosis patients with HBV-DNA negative status after adjusting other covariates (including HBeAg). This indicates the correlation between HBsAg and HBcrAg was not affected by HBeAg in hepatitis B cirrhosis patients with HBV-DNA negative status. In addition, a non-linear association was found between HBcrAg and HBsAg by using a generalized additive model (GAM) after adjusting for other covariates (including HBeAg) in our study, which is different from the previous studies on CHB [5,20].
Subgroup analyses, which can promote a better understanding of the true relationship between HBcrAg and HBsAg, are crucial for scienti c research. In this study, we found a stable correlation in nearly all covariates. The variation trend of β value was consistent in all the strati ed variables (including sex, primary hepatic carcinoma, HBeAg, ALD, Diabetes mellitus, Child-Pugh Class, BCLC, and type and duration of use of anti-HBV agents). Previous studies showed that anti-HBV agents can weaken the correlation between HBcrAg and HBsAg [12,19]. Consistent with those results, we also showed a moderate correlation between HBcrAg and HBsAg in our study, wherein most patients experienced a long-term history of HBV infection and/or use of anti-HBV agents. However, our results are not entirely consistent with most studies targeting treatment-naive CHB patients [5,20]. The likely reason may be that HBsAg is translated from mRNAs transcribed from cccDNA and/or from HBV sequences integrated in the host genome. HBsAg levels can be independent from transcriptional activity of cccDNA[24-25]. Thus, HBsAg levels might be in uenced by age, HBV genotype, and HBV DNA levels [26]. HBcrAg can exist stably and re ect the activity of cccDNA and is not affected by preS/S variants. The production of HBcrAg depends on the transcription of mRNA from cccDNA and is barely affected by NA transcriptase inhibition[19.27]. In our study, most patients experienced a prolonged history of HBV infection and/or anti-HBV agent use, and hence, the relationship between HBsAg and HBcrAg may be weakened. In this study, we made it clear that the correlation between serum HBcrAg and HBsAg in hepatitis B liver cirrhosis.
The moderate correlations of serum HBcrAg and HBsAg suggested that the two indicators may not be replaced by each other in their clinical values in this population. This was veri ed by a previous study that evaluated the antiviral effects of long-term antiviral therapy [25]. Another reason HBcrAg and HBsAg cannot replace each other is because of the unique patterns of their distribution at different stages of HBV infection, which results in different possibilities for their applicability in clinical practice [5]. Further, different sources of generation also determine different clinical values. HBcrAg is the coding product of the pre-C/C region. As the second-most important and valuable viral marker[28], HBsAg can not only originate from the presence of abundant circulating empty dane, but also synthesizes from integrated HBV DNA. Thus, serum HBsAg level re ects not only the cccDNA transcription or mRNA translation but also the host immune control over HBV infection [29][30].
Our study has many strengths. First, we both used a generalized linear model and a generalized additive model, to evaluate the linear or non-linear relationship between HBcrAg and HBsAg. Second, we used strict statistical adjustments to minimize residual confounding, which were not adjusted in previous studies. Third, to our knowledge, this is the rst study on the relationship between HBsAg and HBcrAg in hepatitis B cirrhosis patients with HBV-DNA negative status.
Our study also has some limitations. First, it was an analytical cross-sectional, single-center investigation; hence, the applicability of our ndings to other ethnic groups requires further veri cation. Second, the dynamic correlation was not investigated between HBcrAg and HBsAg, which could further con rm the correlation between them. Third, due to the limited availability of tissue from patients with cirrhosis, we were unable to investigate the relationship between HBsAg, HBcrAg, and intrahepatic cccDNA in this population. We had already obtained some liver biopsy samples from this population and are continuing to collect samples; these will be used for future research to obtain more evidence.

Conclusions
In conclusion, we carried out an in-depth exploration of the correlation between HBcrAg and HBsAg in hepatitis B cirrhosis patients with HBV-DNA negative status. Our data enhanced our understanding of the correlation between HBcrAg and HBsAg in this population, who had a prolonged history of HBV infection. The moderate correlations of serum HBcrAg and HBsAg suggested that the two indicators may not be replaced by each other in this population. Future research should be performed to explore and validate their combined application in predicting adverse hepatic events in those with HBV-DNA negative status.

Availability of data and materials
Data is available from the corresponding author on reasonable request.

Ethics approval and consent to participate
This project was performed in accordance with the Declaration of Helsinki. Retrospective testing of stored surplus clinical samples was approved by the medical ethics committee of the Third Central Hospital of Tianjin, China. Informed consent was obtained from each patient. All data were analyzed anonymously.

Consent for publication
Not Applicable.

Competing interests
The authors of this manuscript have no con icts of nancial and non-nancial interest to disclose.  Abbreviations: CI, con dence interval; HBsAg, hepatitis B surface antigen; HBcrAg, hepatitis B virus core antigen-associated antigen; HBeAg, Hepatitis B e Antigen, AST, aspartate aminotransferase, ETV, entecavir, ALD, alcohol-related liver disease. Abbreviations: CI, con dence interval; HBsAg, hepatitis B surface antigen; HBCrAg, hepatitis B virus core antigen-associated antigen; HBeAg, Hepatitis B e antigen.