Quantification Of Hepatitis B Virus Markers And Their Relationship To Hepatitis B E Antigen Seroconversion

Background:Hepatitis B virus (HBV) was firstly recognized by the discovery of "Australia antigen" in serum by Blumberg in1965. Since then, it has been used as the hallmark for the diagnosis of HBV infection and named as hepatitis B surface antigen (HBsAg). It is the first marker appeared in serum of patients infected with HBV. Its persistence in serum for more than6months is the key diagnostic criterion for chronic Hepatitis B (CHB). According to World Health Organization (WHO), there are more than350million people infected with HBV chronically. HBV infection accounts for annually1million deaths worldwide from cirrhosis, liver failure, and hepatocellular carcinoma (HCC). So it is an important and difficult task to manage individuals with CHB.Hepatitis B e antigen (HBeAg) is another HBV marker appeared in serum together or later than HBsAg. Its existence in serum means that virus is proliferating actively. If HBsAg or HBeAg disappear, and their corresponding antibody can be detected, serological conversion (Seroconversion) happens. It usually indicates a decline in viral replication and the subsequent remission of liver disease, which is a favorable clinical outcome. So they are always used as markers to evaluate antiviral therapy. Several different treatment cohorts have shown that a lower pretreatment HBsAg and HBeAg level is associated with a higher rate of their seroconversion. With the development of technology, it is possible to quantify HBsAg and HBeAg level. It is a new way to explore the HBsAg and HBeAg level and their change pattern to evaluate antiviral therapy. As a result, there is a growing need for automated sample processing systems to provide accurate quantification of HBsAg and HBeAg in clinical practice.Currently, there are two commercialized assays that can measure the HBsAg quantification, the Architect HBsAg assay (Abbott Diagnostics) and the Elecsys HBsAg Ⅱ assay (Roche Diagnostics). The Architect HBsAg assay is the most widely used assay for quantitative measurement of HBsAg. The Elecsys HBsAg assay can be utilized to quantify HBsAg through a simple dilution algorithm. Karsten Wursthorn and Milan J. Sonneveld have made comparisons between Elecsys and Architect Assays. They found that the Elecsys assay present a high correlation with the Architect assay in quantifying HBsAg. But the number of serum samples infected with genotypes B and C was too small in their study. So the first part of my study was to compare the two assays in quantifying HBsAg in serum infected with genotype B and C.The two assays were also used for quantifying HBeAg levels, although there is no commercial assay marketed as HBeAg quantification assay. The Architect HBeAg assay need to use reference standards obtained from the Paul Ehrlich Institute (PEI U/ml). While the Elecsys HBeAg assay get quantification through a formula provided by the manufacturer. The formula was not affected by reagent lot. We secondly aimed to compare the performance of the Elecsys and the Architect assays in quantifying HBeAg in patients with CHB.HBV core gene is a highly variable region. There are two common mutation sites: G1896A mutation and the A1762T/G1764A double mutation. The G1896A mutation can terminate HBeAg translation through forming stop codon, while A1762T/G1764A mutation can reduce serum HBeAg level obviously through down-regulating transcription. Many research found that a lower HBeAg level is associated with a higher rate of HBeAg seroconversion. But we also observed some CHB patients who maintained HBeAg positive with a lower HBeAg level. There are many mechanisms explaining a lower HBeAg level. So the third part of study is to explore the relationship among BCP/PC mutation, low HBeAg level and HBeAg seroconversion.Objective:1. To compare the performance of the Elecsys and Architect assays for HBsAg and HBeAg quantification according to genotype and treatment status.2. To explore the relationship among BCP/PC mutation, low HBeAg level and HBeAg seroconversion.Methods:1. Quantification of HBsAg and HBeAg: Correlation between Elecsys and Architect Assays1.1Source of samples included in present study:We obtained1308sera of patients with CHB from Hepatology Unit of Nanfang hospital, Guangzhou, China. A total of1292sera with both HBeAg and HBsAg positive detected by both the Architect and the Elecsys assays were included in this study. The remaining16samples were excluded due to negative HBeAg detected by one or two methods. The serum samples were stored at-30℃until use.1.2The Architect and the Elecsys assays1.2.1HBsAg assayThe Architect HBsAg assay is a chemiluminescent microparticle immunoassay for quantifying HBsAg in human serum, while the Elecsys HBsAg Ⅱ assay is a electrochemiluminescence immunoassay. The Elecsys HBsAg was used to quantify the HBsAg levels through a dilution algorithm.1.2.2HBeAg assayBoth the Architect and Elecsys HBeAg assays are qualitative tests for detecting HBeAg in human serum and plasma. However, HBeAg quantitative evaluation can be performed using HBeAg Paul-Ehrlich international (PEI) reference standard.1.3Genotype determination The HBV genotype was determined by direct sequencing of a portion of the overlapping genes encoding HBsAg and the B and C sub-domains of the reverse transcriptase of HBV polymerase, followed by phylogenetic analysis.1.4Statistical analysisComparisons between the Architect and the Elecsys assays were achieved by plotting the paired observed values expressed in log10units for each sample. Both HBsAg and HBeAg levels were logarithmically transformed for Pearson’s correlation analysis. Furthermore, consistency of the differences between the log10quantitative values obtained from the Architect and the Elecsys assays was assessed according to the analysis proposed by Bland-Altman plot. Continuous variables were compared between groups using the Mann-Whitney test. Statistical significance of all tests was defined as P<0.05by2-tailed tests. Analyses were performed using SPSS18.0statistical software and GraphPad prism5.0.2. The relationship among PC/C mutation, low HBeAg level and HBeAg seroconversion.2.1Source of samples:All patients included in this part were from the clinical trail of EFFORT project supported by grants from National eleven-five project of China. A total of83patients from Nangfang hospital were included.2.2Mutation analysis:HBV DNA was extracted from100μl serum samples by QIAamp DNA Blood Mini Kit. The C region of HBV was amplified by nest PCR and the products of the second round were sent to Yinjun Company for direct sequencing. Maps of sequencing were analyzed by chromas software. The G1896A mutation and the A1762T/G1764A double mutation was analyzed by Mutation surveyor3.0.2.3Statistical analysis:Continuous variables were compared between groups using the Mann-Whitney test and Kruskal-Wallis H test. Fisher’s exact test was used to compare quantitative data. Statistical significance of all tests was defined as P<0.05by2-tailed tests. All analyses were performed using SPSS18.0. Both HBV DNA and HBeAg levels were logarithmically transformed for analysis.Results:1. Quantification of HBsAg and HBeAg:Correlation between Elecsys and Architect Assays1.1The distribution of genotype and treatment statusDistribution of genotypes was514(39.78%) genotype B,776(60.06%) genotype C,2(0.16%) genotype D. Of1120pretreatment samples,442(39.46%) were genotype B,676(60.36%) genotype C,2(0.18%) genotype D. According to treatment status,1120(86.69%) were taken before treatment and172(13.31%) were taken after therapy for24weeks.1.2The distribution of HBsAg and HBeAg quantification according to genotypeThe median HBsAg quantification in sera pretreatment was significantly higher in the genotype B group compared to genotype C group (genotype B:C4.53versus4.08log10IU/mL, P=0.000, Z=-10.845). While the difference in median HBeAg quantification in sera pretreatment was nonsignificant according to genotype (median B:C2.96versus2.88log10PEI U/mL, P=0.203, Z=-1.274)1.3Systems comparisonThe correlation coefficient of HBsAg quantitation was0.939(95%CI from0.932to0.945, P=0.000) for the Architect assay versus the Elecsys assay on1292routine clinical samples. The difference in quantification of HBsAg was0.075±0.241log10IU/mL (log10IU/mL[Elecsys]-the log10IU/mL [Architect]). The correlation between the two assays in HBeAg quantification was excellent (r=0.987,95%CI from0.985to0.988). The difference was-0.149±0.165log10PE IU/ml (log10PEIU/ml [Elecsys]-the log10PE IU/ml [Architect]).1.4Quantification correlation according to HBV genotypeThere was a significant relationship between measurements by the Architect and Elecsys assays of HBsAg quantification, irrespective of the HBV genotype.(genotype B:r=0.942,95%CI from0.932to0.951; genotype C:r=0.948,95%CI from0.940to0.955. P=0.000). The correlation between measurements by the Architect and Elecsys assays of HBeAg quantification were excellent, independent of the HBV genotype.(genotype B:r=0.988,95%CI from0.985to0.990; genotype C:r=0.986,95%CI from0.984to0.988. P=0.000). Bland-Altman analyses of the discrepancies in HBsAg levels between the Elecsys and the Architect assay showed a bias of0.993to1.037. While the discrepancies between measurements by the Architect and Elecsys assays of HBeAg levels were shown as a bias of0.815to0.957.1.5Quantification correlation according to HBV genotype and treatment statusAll study samples were divided into four groups according to HBV genotype and treatment status. For HBsAg quantification, the correlations between measurements detected by the two assays in all groups were excellent.(Naive:genotype B:r=0.919,95%CI from0.903to0.932; genotype C:r=0.945,95%CI from0.936to0.952; After therapy:genotype B:r=0.979,95%CI from0.967to0.987; genotype C:r=0.956,95%CI from0.935to0.970. P=0.000). As to HBeAg quantification, both the genotype and treatment status did not influence on the high correlation and agreement between the two HBeAg assays.(Naive:genotype B:r=0.985,95%CI from0.981to0.987; genotype C:r=0.984,95%CI from0.981to0.986; After therapy:genotype B: r=0.980,95%CI from0.968to0.988; genotype C:r=0.982,95%CI from0.973to0.988. P=0.000).2. The relationship among PC/BCP mutation, low HBeAg level and HBeAg seroconversion.2.1The HBeAg level and its relationship to HBeAg seroconversion rateAccording to HBeAg level, all samples were divided into low or high HBeAg group. Their HBeAg seroconversion rates at12,24,36, and52weeks after therapy were23.53%VS0%(P=0.003),35.29%VS1.92%(P=0.001),41.18%VS5.77%(P=0.001),52.94%VS17.31%(P=0.009), respectively.2.2The HBeAg level and its relationship to PC/BCP mutationMedian HBeAg titer was higher in the WT group compared to the patients with either BCP or PC variants (3.12versus2.691og10PEIU/mL, P=0.032). The effect of the BCP/PC variants persisted after adjustment for viral load (P=0.003). Median HBeAg titer was lower in the BCP pure mutation group compared to the patients with mixed BCP variants and wild group (2.11versus3.10versus3.081og10PEIU/mL, P=0.002). The effect of the BCP variants persisted after adjustment for viral load (P=0.006). A nonsignificant trend was also noted for lower HBeAg titer in the setting of wild group compared to mixed BCP variants (P=0.693).2.3The PC/BCP mutation and its relationship to HBeAg seroconversion rateAccording to the PC/BCP mutation status, all samples were divided into mutation or wild group. Their HBeAg seroconversion rates at12,24,36, and52weeks after therapy were9.5%VS0.0%(P=0.150),9.5%VS11.1%(P=1.00),9.5%VS22.2%(P=0.173),23.8%VS29.6%(P=0.589), respectively.Conclusions:1. This study demonstrates a high correlation between the Elecsys and the Architect assays in quantifying HBsAg and HBeAg of CHB patients, regardless of HBV genotype and treatment phase. Both the two assays can be used to monitor the HBsAg and HBeAg levels in patients with CHB.2. The G1896A and/or A1762T/G1764A variants can decrease serum HBeAg level obviously.3. A lower serum HBeAg level is associated with a higher rate of HBeAg seroconversion.4. The G1896A and/or A1762T/G1764A variants are not directly related to HBeAg seroconversion in this study.