Article Text
Abstract
Background and objective Precore (PC) variant (G1896A) and basal core promoter (BCP) variant (A1762T/G1764A) of HBV are associated with risk of hepatocellular carcinoma in HBV carriers. However, little is known about their impact on the adverse outcomes of hepatitis B e antigen (HBeAg)-negative hepatitis and liver cirrhosis.
Methods 251 spontaneous HBeAg seroconverters who had genotype B or C infection and received a long-term follow-up were enrolled. PC and BCP mutants were determined qualitatively and quantitatively to correlate with these adverse outcomes. The findings were validated by an independent case–control study, which included 184 patients with biopsy-proven liver fibrosis stages.
Results In the longitudinal cohort study, BCP mutant and possibly PC wild type were associated with cirrhosis development, but not HBeAg-negative hepatitis. Multivariable analysis showed that only BCP mutant was an independent risk factor for cirrhosis development. Using quantitative analysis of BCP mutant, a higher proportion of BCP mutant, defined as a continuous variable, a dichotomous variable or an ordinal variable, was associated with a higher risk of cirrhosis. If we chose 45% of BCP mutant as the cut-off, the risk of cirrhosis was higher in patients with BCP mutant ≥45% compared to <45% in the longitudinal cohort; this finding was validated by the case–control study (adjusted OR: 2.81, 95% CI 1.40 to 5.67).
Conclusions A higher proportion of BCP mutant increases the risk of liver cirrhosis development in HBV carriers with genotype B or C infection.
Statistics from Altmetric.com
Signification of this study
What is already known on this subject?
-
Chronic hepatitis B infection is an important risk factor for cirrhosis and hepatocellular carcinoma (HCC) development.
-
Precore (PC) wild type (G1896) and basal core promoter (BCP) mutant (A1762T/G1764A) of HBV, which were determined mostly by quantitative assay, are shown to be associated with HCC development. However, the relationships between the viral mutants and HBeAg-negative hepatitis and cirrhosis are unclear.
-
Novel technology, such as pyrosequencing, can determine the proportion of viral mutant precisely.
What are the new findings?
-
Neither PC wild type nor BCP mutant of HBV was associated with HBeAg-negative hepatitis.
-
BCP mutant, but not PC variant, was shown to be an independent risk factor for cirrhosis development.
-
The longitudinal cohort study showed a dose–response relationship between the proportion of BCP mutant and cirrhosis risk and the relationship was validated by a cross-sectional case–control study.
-
Cirrhosis risk can be stratified by different proportions of BCP mutant.
How might it impact on clinical practice in the foreseeable future?
-
In this study, we have shown that a higher proportion of BCP mutant is an independent risk factor for cirrhosis.
-
All the treatment guidelines recommend initiating antiviral therapy in HBeAg-negative, non-cirrhotic patients on the basis of high viral loads and elevated alanine aminotransferase levels.
-
Our findings suggest that a high percentage of BCP mutant should be considered as a new criterion to commence antiviral therapy to decrease the risk of cirrhosis.
Introduction
HBV infection is a global health problem, resulting in over 1 million deaths annually worldwide.1 ,2 Patients with chronic HBV infection are at increased risk of developing cirrhosis and hepatocellular carcinoma (HCC) over time.2–5 It is estimated that 25–40% of HBV carriers who contract the virus early in life may eventually develop these disastrous complications.4
Previous studies have shown that serum HBV DNA and hepatitis B surface antigen (HBsAg) levels are positively associated with disease progression in patients with chronic HBV infection.6–10 In particular, patients with serum HBV DNA levels ≥2000 IU/mL at study entry have increased likelihood of developing hepatitis B e antigen (HBeAg)-negative hepatitis, cirrhosis and HCC during long-term follow-up.6 Apart from these quantifiable viral factors, both HBV genomic variants, precore (PC) wild type (G1896) and basal core promoter (BCP) mutant (A1762T/G1764A), are documented to be associated with HCC development.11–15 However, whether the quantified level of these HBV variants could predict HBV-related adverse outcomes other than HCC, such as HBeAg-negative hepatitis and cirrhosis, remains largely unknown and deserves further study.
Qualitative analysis, such as Sanger sequencing, is usually used to determine the dominant viral strain, which is defined as >50% of the virus. Therefore, the qualitative analysis is limited as it cannot describe exact proportions of viral mutants. Furthermore, it is difficult to interpret the data when the proportions of wild type and mutant viruses are comparable. With recent advances in biotechnology, quantification of these viral variants has become possible and more accurate,16 which allows us to address these issues properly. We thus quantified the HBV variants using pyrosequencing,17 and investigated whether different proportions of viral variants could predict HBV-related adverse outcomes.
We used a longitudinal cohort of spontaneous HBeAg seroconverters18–21 and determined their PC and BCP variant status qualitatively and quantitatively. The proportions of these variants were used to correlate with the clinical endpoints of HBeAg-negative hepatitis and cirrhosis. We then validated the findings using a case–control study on patients with biopsy-proven liver fibrosis stages.
Materials and methods
Longitudinal cohort study
This study utilised an established cohort of patients from the Study of E Antigen seRoClearance of Hepatitis B patients (SERACH-B) as previously described; the flow of enrolment is summarised in figure 1A.18–21 Initially, the SEARCH-B study included 390 spontaneous HBeAg seroconverters. Among them, we enrolled 251 spontaneous HBeAg seroconverters who had HBV-DNA levels ≥200 IU/mL at 1 year post HBeAg seroconversion and who were not indicated for antiviral treatment at that time.22 ,23 One year post HBeAg seroconversion was chosen as serum HBV-DNA levels at this time point correlate better with subsequent adverse outcomes.18 ,19
Cross-sectional case–control study
We recognised that the major flaw of this longitudinal cohort study in investigating the adverse outcome of liver cirrhosis was that most cases were diagnosed via abdominal ultrasound examination, not liver biopsy. Hence, to validate the findings derived from the longitudinal cohort study, a cross-sectional case–control study was conducted using frequency match of sex and age (at a 5-year interval); the flow of enrolment is summarised in figure 1B. A total of 92 patients with biopsy-proven cirrhosis and 92 sex- and age-matched controls without cirrhosis on liver histology were enrolled.
The patients of the case–control study were enrolled from 2000 to 2009. They were all HBeAg-negative and most received liver biopsy according to the Taiwanese reimbursement guidelines.24 ,25 They had elevated alanine aminotransferase (ALT) levels and none had HCC at the time of liver biopsy. Serum and liver samples were collected simultaneously for the correlation between HBV variants and hepatic fibrosis stages.
Informed consent was obtained from all participants of the cohort study and case–control study as per the ethical committee of the National Taiwan University Hospital.
Data collection
Patients were initially tested for serological markers (HBsAg, HBeAg, anti-HBe, antibody against HCV (anti-HCV) and antibody against hepatitis D virus (anti-HDV)). During the follow-up, ALT and α-fetoprotein levels were assayed every 6 months or more frequently if clinically indicated. Abdominal ultrasound examination using a high-resolution, real-time scanner was performed at least every 6 months after enrolment to detect cirrhosis or HCC. At each clinical visit, a serum sample was collected and stored at −20°C until used.
Definitions of sustained HBeAg seroconversion and study endpoints
HBeAg seroconversion was defined as persistent HBeAg negativity and anti-HBe positivity for at least 1 year. We evaluated the longitudinal cohort with two separate endpoints, HBeAg-negative hepatitis and cirrhosis. HBeAg-negative hepatitis was defined as ALT elevation more than twice the upper limit of normal (ULN) with concomitant serum HBV-DNA levels ≥2000 IU/mL.8 ,18 ,23 The ULN for serum ALT was 40 U/L. Apart from the aforementioned criteria, patients with drug or alcohol usage and those with serological evidence suggestive of other viral hepatitis infection or autoimmune liver disease were excluded.
Cirrhosis was diagnosed using histology or ultrasonographic findings together with clinical features such as thrombocytopenia, gastro-oesophageal varices or ascites.26–28 The detection of liver cirrhosis by abdominal ultrasonography was according to a scoring system based on the features of liver parenchyma, liver surface, hepatic vessel and spleen size. For ultrasonographic diagnosis of cirrhosis to be made, the score had to be ≥8 on at least two ultrasonographic studies more than 6 months apart, which could yield specificities of more than 90%.26 In the cross-sectional study, hepatic fibrosis stage was evaluated by the Metavir score.
Serological assays
Serum HBsAg, HBeAg, anti-HBe, anti-HCV and anti-HDV were tested by commercial kits (Abbott Laboratories, North Chicago, Illinois, USA).
Extraction of serum HBV DNA
Serum viral DNA was extracted using commercial kits (QIAamp DNA Blood and Tissue Mini Kit; QIAGEN, Valencia, California, USA). The extracted DNA was used to quantify HBV DNA and determine HBV genotype, PC (1896) and BCP (1762/1764) sequences of the HBV genome.
Quantification and genotyping of HBV
Serum HBV-DNA level and genotype were determined using real-time PCR-based single-tube assay as previously described.29 This method consists of two consecutive steps. The first step uses PCR to amplify the region (nt 1261–1600), and the second step uses melting curve analysis to genotype HBV. The detection limit of HBV-DNA levels is 20 IU/mL (100 copies/mL).
Qualitative analysis of dominant sequences PC (1896) and BCP (1762/1764) by Taqman assay
The dominant sequences of PC 1896 and BCP 1762/1764 were determined using the TaqMan assay, as previously described.30 Primer and probe sets were designed to prime highly conserved genomic sequences and are summarised in online supplementary table S1. We used TaqMan Genotyping Master Mix (Applied Biosystems, Foster City, California, USA) to amplify the PC and BCP regions. The total volume of reactants was 50 µL, which consisted of 20 µL of extracted DNA. The final concentration of each primer was 0.9 μmol/L, and the final probe concentration was 0.2 μmol/L. The thermocycling conditions used were: 60°C for 30 s, 95°C for 10 min, and 45 cycles of 95°C for 12 s and 60°C for 1 min. Post-PCR read was performed at 60°C for 30 s. All reactions were performed under the same conditions using the ABI StepOnePlus Real-Time PCR System (Applied Biosystems). In addition, three control samples were used in every reaction. This included a wild type, a mutant and a 1:1 mixture of wild type plus mutant. After completion of the assay, the signals generated from the software (StepOne software V.2.1) enabled us to obtain the major sequences of PC 1896 and BCP 1762/1764. This assay could determine the major sequence if the viral load is higher than 1000 copies/mL (about 200 IU/mL).
Quantitative analysis of BCP mutant (A1762T) using pyrosequencing
We used pyrosequencing to quantify BCP mutant as described previously.17 In short, the region containing 1762 and 1764 (nt 1783–1875) was first amplified by PCR with the biotinylated primer. The biotin-labelled single stranded DNA was then captured by beads with streptavidin and was further quantified by pyrosequencing on the Qiagen platform (Qiagen, Germany). We added 10 µL of HBV DNA in each test to ensure at least 100 copies of HBV DNA are present in PCR. The detection limit of this assay was 10 000 copies copies/mL (about 2000 IU/mL). Since there is a high correlation between A1762T and G1764A, only the data and percentage of A1762T were used for subsequent analysis.17
Quantification of HBsAg
The serum HBsAg level was quantified using the Architect HBsAg QT (Abbott Laboratories) according to the manufacturer's instructions.31 ,32 The dynamic range of HBsAg levels is from 0.05 to 250 IU/mL. If the HBsAg level was found to be higher than 250 IU/mL, the samples were diluted to 1:100 to 1:1000 to obtain a reading within the range of the calibration curve.
Statistical analysis
Mean and SD were calculated for continuous variables. Percentages were used for categorical variables. In the longitudinal study, follow-up started from 1 year after HBeAg seroconversion because the viral factors were determined at the end of the first year after HBeAg seroconversion. The person-years were censored on the date of reaching the aforementioned endpoints, the date of initiating antiviral therapy, the last date of follow-up, or 31 October 2010, whichever came first. The date was chosen because Taiwan health insurance started to reimburse antiviral treatment in HBeAg-negative hepatitis patients with elevated ALT levels and high viral load. For patients with multiple episodes of HBeAg-negative hepatitis, only the earliest episode of hepatitis was counted. The cumulative incidence of the designated endpoints was derived using the Kaplan–Meier method and the log-rank test was used to test for statistical difference. Cox proportional hazards regression model was used to calculate the crude and multivariable-adjusted HRs and 95% CI of each endpoint. The trend test for the proportion of viral mutants and cirrhosis risk was derived from the univariable analysis of continuous proportion of viral mutants in a Cox regression of time to cirrhosis.
We adopted receiver operating characteristic (ROC) curve analysis to determine the optimal cut-off of BCP mutant proportion that predicts cirrhosis development within 9 years. The sensitivities and specificities of different cut-offs were calculated. The optimal cut-off percentage was derived using the Youden index, which is defined as sensitivity plus specificity minus one.8
In the case–control study, the percentages of BCP mutant were compared between the cases with cirrhosis and matched controls as a continuous variable and a categorical variable using the cut-off derived from the longitudinal study. Data were analysed by the χ2 test, Fisher's exact test and the Student t test, where appropriate. The logistic regression model was used to calculate the crude and multivariable-adjusted OR and 95% CI.
Results
Baseline characteristics of the longitudinal cohort
Table 1 shows the demographic data of 251 spontaneous HBeAg seroconverters with HBV DNA ≥200 IU/mL. The prevalence of genotype B infection, PC wild type and BCP mutant were 79.7%, 41.8% and 31.9%, respectively. The proportion of double wild type of PC and BCP was 22.2%, and 12.9% for double mutation. The prevalence of PC mutant and BCP mutant was analysed in patients with different viral and host factors (figure 2A,B). We found that genotype B infection was associated with a higher frequency of PC mutant but a lower frequency of BCP mutant compared to genotype C infection. Also, patients with BCP mutant had a lower frequency of PC mutant, compared to those with BCP wild type (figure 2A).
Cumulative incidence rates of HBeAg-negative hepatitis and cirrhosis
In 251 spontaneous HBeAg seroconverters with HBV-DNA levels higher than 200 IU/mL at 1 year post HBeAg seroconversion, 7 developed HCC, 26 developed cirrhosis and 92 developed HBeAg-negative hepatitis. Twenty-four (9.6%) of the 251 patients were censored by the initiation of antiviral therapy when using cirrhosis as the endpoint. All of them received antiviral treatment due to HBeAg-negative hepatitis. It should be noted that the prevalence of PC and BCP mutant was comparable between patients with and without antiviral therapy.
Of the 26 patients who developed cirrhosis during follow-up, none had evidence of cirrhosis within the first year after HBeAg seroconversion. Six (23.1%) of the patients had cirrhosis diagnosed histologically. Among the remaining 20 patients who had cirrhosis diagnosed by ultrasonographic findings, 12 (46.2%) had thrombocytopenia and 2 (7.7%) had evidence of oesophageal varices.
Incident rates of HBeAg-negative hepatitis and cirrhosis categorised by PC and BCP status
When using HBeAg-negative hepatitis as the endpoint, neither PC wild type nor BCP mutant was associated with the endpoint (table 2). In contrast, patients with BCP mutant had a higher incidence of cirrhosis than those with BCP wild type (2.75 per 100 person-years (95% CI 1.78 to 4.27) vs 0.39 per 100 person-years (95% CI 0.18 to 0.87)).
Risk factors associated with HBeAg-negative hepatitis
We correlated baseline factors with risk of HBeAg-negative hepatitis; the results are summarised in online supplementary table S2. Male, older age at HBeAg seroconversion, higher viral load and genotype C infection were shown to be associated with higher risks of HBeAg-negative hepatitis using multivariable analysis. However, neither PC wild type nor BCP mutant was associated with emergence of HBeAg-negative hepatitis in multivariable analysis.
Risk factors associated with cirrhosis
We then correlated PC and BCP status with cumulative incidence of cirrhosis (figure 2C,D). Interestingly, BCP mutant and possible PC wild type were associated with a higher risk of cirrhosis (p<0.001 and p=0.055, respectively). If patients were stratified by both PC and BCP status, those with double wild type and those with PC mutant and BCP wild type had similar and the lowest risk of cirrhosis, followed by those with double mutants. Patients with PC wild type and BCP mutant had the highest risk of cirrhosis (figure 2E). However, we did not adopt this stratification as a factor in multivariable analysis due to inadequate statistic power, which was owing to a small case number in each category.
Using multivariable analysis with PC and BCP as two variables and adjusted for age, sex, genotype and levels of ALT, HBV DNA and HBsAg, we found that BCP mutant served as an independent risk factor for cirrhosis development (HR 4.26, 95% CI 1.32 to 13.77) in addition to older age and a higher HBV DNA level (table 3).
We then determined BCP status at the time when cirrhosis was detected. Of 20 patients with BCP mutant at baseline, 19 had detectable serum HBV DNA level when cirrhosis was detected and 16 remained BCP mutant as the dominant strain (84.2%), suggesting that BCP mutant persisted to be dominant through the process of cirrhosis development.
We further investigated whether BCP mutant remained a risk factor for cirrhosis after adjustment for emergence of HBeAg-negative hepatitis during follow-up. Our data showed that the HR of BCP mutant for cirrhosis was 6.46 (95% CI 2.38 to 17.56, p<0.001) in patients who developed HBeAg-negative hepatitis during follow-up, whereas in those without HBeAg-negative hepatitis, the HR of BCP mutant was 10.51 (95% CI 1.17 to 94.10, p=0.035).
Quantification of BCP mutant in patients with HBV DNA ≥2000 IU/mL and its correlation with cirrhosis development
Since BCP mutant was the only viral variant associated with cirrhosis development, we determined the percentage of virus containing BCP mutant in patients with HBV DNA ≥2000 IU/mL and evaluated its role in predicting cirrhosis risk. The mean percentage of BCP mutant was 37.9±40.0 (table 1). When correlated with the cumulative incidence of cirrhosis, a higher proportion of BCP mutant was positively associated with a higher risk of cirrhosis (p for trend <0.001). After adjustment for all the factors mentioned in table 3, it was found that the HR of cirrhosis was 1.02 for every 1% increase in BCP mutant percentage (95% CI 1.00 to 1.03, p=0.013).
Next, we used ROC curve analysis to determine the optimal cut-off for predicting cirrhosis risk. We noted each patient in this cohort had a different follow-up period, and thus included only patients with at least 9 years of follow-up for analysis. This timeframe was chosen because the median follow-up period was 8.9 years. Using ROC curve analysis, the area under ROC of quantitative BCP mutant in predicting 9-year cirrhosis risk was 0.73 (95% CI 0.62 to 0.84) (figure 3A). The predictive performance of different percentages of BCP mutant is summarised in online supplementary table S3; the optimal cut-off is 45%.
Cirrhosis risk in categorised percentages of BCP mutant
In the first model, we used 45% of BCP mutants as the cut-off and found a higher risk of cirrhosis in patients with BCP mutant ≥45% compared with <45% (p<0.001, figure 3B). After adjustment for age, sex, ALT level, HBV DNA level, HBsAg level, HBV genotype and PC status, BCP mutant ≥45% still remained a risk factor for cirrhosis with adjusted HR of 5.94 (95% CI 1.78 to 19.80) (table 3).
In the second model, we initially divided patients by quartiles of BCP mutants and analysed the cumulative incidence of cirrhosis in these four groups of patients (figure 3C). The groups with BCP mutant <4.5% and 4.5–12.0% had a similar and lowest cumulative incidence rate, followed by the group of BCP mutant of 12.0–89.1%; BCP mutant >89.1% had the highest rate. We thus decided to merge the two groups with lower proportions of BCP mutant and recategorised the patients using cut-offs of 10% and 90% as they are the two nearest integers to 12.0% and 89.1%. It was evident that cirrhosis risk could be categorised into three groups by BCP mutant (figure 3D); <10% had the lowest risk (as reference), followed by BCP mutant between 10% and 90% with HR of 3.55 (95% CI 1.06 to 11.85), and highest in BCP mutant >90% with HR of 7.68 (95% CI 2.53 to 23.37). With adjustment for age, sex and levels of ALT, HBV DNA and HBsAg, BCP mutant between 10% and 90% and >90% still remained as risk factors for cirrhosis. It should be noted that the viral loads and HBsAg levels were comparable among the categories stratified by BCP mutant of 45% or by BCP mutant of between 10% and 90% (data not shown).
Validation of the relationship between BCP mutant percentage and cirrhosis in a case–control study
To validate the relationship between BCP mutant and the risk of cirrhosis, we conducted an independent cross-sectional case–control study. Both the case and control groups enrolled patients who were HBeAg-negative and had received liver biopsy. In total, 92 cases with cirrhosis and 92 age- and sex-matched controls by frequency match were enrolled; their characteristics are summarised in table 4.
In the case–control study, patients with cirrhosis had a higher percentage of BCP mutant than those without (mean±SD, 70.0±42.2 vs 51.4±44.3, p=0.004). When we grouped the patients either by BCP mutant of 45% (see online supplementary figure S1A) or by BCP mutant of between 10% and 90% (see online supplementary figure S1B), the cirrhosis group had a higher proportion of BCP mutants compared to the non-cirrhosis group (both p<0.05). When patients were categorised according to BCP mutant percentage ≥45% or <45%, it was evident that a high percentage of BCP mutant was associated with a higher risk of cirrhosis (OR 2.63, 95% CI 1.43 to 4.82, p=0.002). After adjustment for all covariates listed in table 4, BCP mutant ≥45% remained to be associated with a higher risk of cirrhosis with adjusted OR of 2.81 (95% CI 1.40 to 5.67, p=0.004).
Discussion
Liver cirrhosis is the strongest risk factor for HCC development.4 ,33 HBV viral variations are proven to shape clinical outcomes, as evidenced in the association between BCP mutant and increased HCC risk.11 ,12 However, whether BCP mutant is associated with higher risk of disease progression, such as hepatitis activity and cirrhosis, has not been clearly addressed. Our longitudinal cohort study was the first study to show that BCP mutant was associated with liver cirrhosis development, but not HBeAg-negative hepatitis. The correlation was further supported by a dose–response relationship when correlating BCP mutant percentage to risk of cirrhosis. Irrespective of whether BCP mutant was defined as a continuous variable, a dichotomous variable or an ordinal variable, a higher percentage of BCP mutant was persistently shown to be associated with increased risk of cirrhosis. This correlation was further validated in an independent case–control study with different stages of biopsy-proven hepatic fibrosis.
Taking together these lines of evidence, it could be proposed that the percentage of BCP mutant might serve as a predictor for cirrhosis development in HBV carriers with genotype B or C infection. This factor could be integrated into the current risk calculator for cirrhosis development, especially in those with high viral loads.34 All the treatment guidelines recommend initiating antiviral therapy in HBeAg-negative, non-cirrhotic patients on the basis of high viral loads and elevated ALT levels.22 ,23 ,35 Our findings suggested that patients with a high percentage of BCP mutant should be considered to receive antiviral treatment in order to prevent cirrhosis. Nonetheless, further validation with other populations with other HBV genotype infections is required.
In patients with chronic HBV infection, advanced fibrosis or cirrhosis is usually caused by accumulation of fibrous tissue in liver, which corresponds to the activity of HBV-related hepatitis.4 A previous cross-sectional study has shown that BCP mutant may be associated with cirrhosis development.36 Our results showed that BCP mutation was not associated HBeAg-negative hepatitis but directly linked to cirrhosis. These findings suggest that hepatic fibrogenesis may not be solely induced by hepatitis activity but also through the virus itself. A previous study has indicated that A1762T/G1764A double nucleotide mutations in the core promoter region affect codons 130 and 131 of the X gene (K130M and V131I), which alters the activity of HBx protein and contributes to hepatocarcinogenesis.37 Another in vitro study suggested HBx protein activates hepatic stellate cells through paracrine secretion of TGF-β, which is responsible for fibrogenesis.38 Therefore, the alterations of HBx protein may increase its fibrogenic activity, which causes a higher cirrhosis risk. To prove this hypothesis, further in vitro and in vivo studies are required to clarify the functional differences between wild type and mutant HBx protein.
Our study design had several unique characteristics. First, we found that most BCP mutants (84.2%) persisted throughout the course of cirrhosis development, suggesting that BCP mutant may retain the fibrogenic activity throughout the process of hepatic fibrogenesis. Second, Nie et al16 first proposed the concept that quantifying HBV mutants could be clinically useful in 2012. We then showed the roles of PC and BCP quantification in predicting responses to interferon-based treatment.17 In this study, we further showed that BCP quantification could be useful in predicting liver cirrhosis development in HBV carriers. Qualitative assays can only describe the viral characteristic with a certain viral strain higher or less than 50% of the total viral strains. The quantitative assay breaks the limitation and indicates that categorising patients by either 45% or between 10% and 90% can better stratify the cirrhosis risk. We believe quantification of BCP mutant provides additional information for personalised management of chronic hepatitis B patients.
Our study had a few limitations. First, we did not analyse the relationship between viral mutants and HCC because only seven cases developed HCC during follow-up. Even using liver cirrhosis as the endpoint, the case number was also relatively small (N=26), thus the 95% CI of the analysis was wide. We thus adopted another independent case–control study with biopsy-proved hepatic fibrosis stages to confirm the relationship between BCP mutant and cirrhosis development. Second, we could not quantify BCP mutant percentages in patients with HBV DNA level <2000 IU/mL due to the detection limit of the assay. However, it should be noted that the risk of cirrhosis was low in patients with low viral loads and HBsAg level alone has been shown to stratify cirrhosis risk effectively.8 ,34 Taken together, in patients with higher viral loads, quantification of BCP mutant may complement HBV DNA level to predict cirrhosis risk.
In summary, a higher percentage of BCP mutant is closely associated with an increased risk of cirrhosis development. Quantification of BCP mutant may be integrated in the prediction model of HBV-related liver cirrhosis, at least in patients with genotype B or C infection.
Acknowledgments
We thank Abbott Company for providing the quantitative HBsAg kits, and colleagues at the National Taiwan University Hospital, Taipei, Taiwan, who enrolled and followed the patients. We also thank research assistants who assisted in laboratory analyses and collection of clinical information.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Files in this Data Supplement:
- Data supplement 1 - Online supplement
- Data supplement 2 - Online table
Footnotes
-
Contributors Study concept and design: T-CT and J-HK; acquisition of data: C-JL, H-CY, T-HS, C-HL, P-JC, D-SC, and J-HK; lab work: C-ST, and FCV; analysis and interpretation of data: T-CT, W-TY, C-LC, and J-HK; drafting of the manuscript: T-CT and SF-TK; critical review of the manuscript for important intellectual content: P-JC, D-SC, and J-HK; statistical analysis: C-LC, W-TY, and T-CT; obtained funding: C-CW, T-CT, C-JL, and J-HK; technical or material support: T-CT and H-CY; study supervision: D-SC and J-HK.
-
Funding This work was supported by grants from the Taipei Tzuchi Hospital (TCRD-TPE-101-12, TCRD-TPE-103-32 and TCRD-TPE-NSC-102-02), the Department of Heath (DOH99-DC-1001 and DOH100-DC-1019) and the National Science Council, Executive Yuan, Taiwan (DOH102-DC-1101 and MOHW103-CDC-C-114-123105).
-
Competing interests None.
-
Patient consent Obtained.
-
Ethics approval Ethical Committee of National Taiwan University Hospital.
-
Provenance and peer review Not commissioned; externally peer reviewed.