Background
Hepatitis B virus (HBV) infection is a global epidemic, with more than 350 million of chronically infected carriers of the virus surface antigen (HBsAg) [
1]. Without treatment, 15 to 40% of those with chronic HBV infection will develop cirrhosis that can potentially lead to hepatocellular carcinoma [
2].
HBV is hyperendemic in Senegal, where the prevalence of chronic HBsAg carriage is about 15-20% in the general population [
3]. Senegalese patients are primarily infected during early childhood, and genotypes E and A predominate in these infections. HBV infection in Senegal shows an 83% rate of precore mutations [
4]. Reliable, easy-to-perform markers are needed to assess the impact of chronic HBV infection on liver diseases. Quantification of plasma HBV DNA by real-time PCR, combined with liver biopsy histo-pathological stage and biological markers of hepatocyte cytolysis level, can distinguish inactive HBsAg carriers from patients with active disease who require treatment. Patients should be considered for treatment when they have HBV DNA levels above 2000 IU/mL (3.2 log
10 IU/mL), or serum alanine aminotransferase (ALT) levels above the upper limit of normal (ULN) and liver biopsy suggestive of significant fibrosis (≥F2) [
5]. In developing countries, viral load measurement is rarely accessible. If only one measure of viral load is available, it may not reflect real viral activity. Possible fluctuations may affect the indication of treatment. However, several studies have shown long-term HBV DNA spontaneous fluctuations [
6,
7]. The magnitude of these changes is likely to change a treatment decision [
7]. Short-term fluctuations and factors affecting them remain unknown [
8].
Recently, HBsAg levels have been proposed as a marker for monitoring HBV infected patients [
9,
10]. HBsAg levels change over the natural course of chronic HBV infection and during antiviral therapy. Moreover, HBsAg quantification can be used to differentiate true inactive carriers with HBsAg level <1000 IU/mL from patients in remission who are likely to progress to cirrhosis [
11]. The aim of this work is to describe spontaneous HBV DNA and HBsAg level fluctuations over a four month- follow-up period among Senegalese patients positive for HBsAg with normal ALT, so as to identify factors associated with these fluctuations.
Methods
Patient population
Patients were consecutively enrolled by private practitioners and by four public hospitals in Dakar, Senegal’s capital city, from September 2005 to April 2006. Treatment-naive patients above 18 years old, with positive HBsAg over six months, symptom-free, HIV, HCV and HDV negative, were eligible for enrolment.
Eighty-seven patients with persistent normal ALT values were prospectively monitored for determination of ALT and aspartate aminotransferase (AST), quantification of HBV DNA three times at two-month intervals (M0, M2, and M4). For those with positive HBV DNA, genotyping and mutations affecting the expression of HBeAg were performed.
Ethical approvals
The protocol was in accordance with Declaration of Helsinki ethical guidelines and was approved by the Senegalese Health Research National Council. Patients fulfilling the inclusion criteria were enrolled after providing written and informed consent. Lamivudine was proposed to patients eligible for treatment, according to the above mentioned criteria.
Material and methods
Data collection
The following data were collected [
1]: general characteristics (age, sex, weight, height, known HBV infection duration, body mass index (BMI)) [
2]; biological markers at the inclusion (genotype, HBeAg, HBsAg and platelets). HBV DNA, HBsAg quantification, ALT, AST, were measured three times with intervals of two months (M0, M2 and M4). ALT and AST results were expressed relative to the normal values according to the technique used (Ortho Clinical Diagnostics, Issy-les-Moulineaux, France) i.e. 52 IU/mL (females) and 72 IU/mL (males) for ALT, and 36 IU/mL (females) and 59 IU/mL (males) for AST. Variations of ALT, AST, HBsAg and HBV DNA between two visits were determined and expressed as ΔALT, ΔAST, ΔHBsAg and ΔHBV DNA, respectively.
Biological markers
HBeAg and qualitative HBsAg were performed by an automated EIA method (Axsym, Abbott Diagnostics, Rungis, France). Biochemical parameters were determined by Vitros 250 instrument (Ortho Clinical Diagnostics, Issy-les-Moulineaux, France). Platelets were determined by Cell-Dyn 3700 (Abbott).
Virological analyses
The HBV DNA quantification was performed using the Cobas AmpliPrep/Cobas Taqman HBV test, v1.0 assay (Roche Diagnostics, Meylan, France), with a detection threshold of 12 IU/mL (1.1 log10 IU /mL).
The core mutation W28 at nucleotide 1896 (precore mutation C28) and clade genotyping were determined by research DNA microarray (bioMérieux, Marcy l’Etoile, France) [
12].
HBsAg quantification
Serum HBsAg levels were retrospectively quantified using sera stored at −20 °C, which had been used for HBV DNA measurement. Architect HBsAg EIA (Abbott, Rungis, France) [
13] was used with a dynamic range of 0.05-250 IU/mL. Samples with HBsAg >250 IU/mL were diluted to 1/100 to bring the value within the range of calibration [
14].
Statistical analysis
Continuous variables were expressed as median and interquartile ranges (IQR), and categorical variables were expressed as percentages. Univariate analyses were based on Fisher’s exact test for categorical variables and Mann–Whitney test for continuous variables. The HBsAg (log10 IU/mL) to HBV DNA (log10 IU/mL) ratio was assessed for all serum samples according to HBV DNA levels (≤3,]3–4],]4–5], and > 5 log10 IU/mL). Comparison of ratios between levels of HBV replication was analyzed with the Kruskal-Wallis non-parametric test. When a significant difference was detected, Mann–Whitney test was performed, with Bonferroni correction for multiple testing. HBsAg and HBV DNA fluctuations were measured three times at two month intervals (M0, M2 and M4). For each subject, the largest difference in HBV DNA levels between two consecutive visits was recorded and classified according to four classes: ≤ 0.5,]0.5-1][,]1-2] and >2 log10 IU/mL. Friedman’s test was used to compare HBsAg fluctuations between the three values. Correlation between viral load fluctuations was based on the Spearman coefficient.
All variables associated with HBV DNA fluctuations > 0.5 log10 IU/mL in univariate analysis (p < 0.25) were included in a backward stepwise logistic regression model. A P value of ≤0.05 was considered to denote statistical significance. Statistical analyses were performed using STATA software version 12.0 (Stata Corporation, College Station, TX).
Discussion
This study highlights the frequency and magnitude of spontaneous HBV DNA fluctuations over a short period (4 months) among Senegalese patients with normal transaminases, as well as their possible impact on clinical characterization of the disease and its therapeutic management. The lack of association between these fluctuations and demographic parameters (age, gender) or biochemical (transaminases) or virological data (genotype, mutations affecting the expression of HBeAg) could be explained by a lack of power because of the small number of subjects included in the study. Low BMI < 21 kg/m
2 was the only factor associated with HBV DNA fluctuations; patients with a lower BMI values were at higher risk of having fluctuations. One hypothesis is that HBV DNA fluctuations can reflect more efficient immune response which is more present in healthy, low BMI patients. As previously demonstrated, there is strong evidence that excess adiposity, defined by high BMI, negatively impacts immune function and host defenses in obese individuals [
15].
The stability of the biochemical markers contrasts with the volatility of the HBV DNA.
The lower replicativity among HBeAg negative patients does not explain the HBV DNA variability observed in our study.
The frequency and magnitude of DNA fluctuations and their possible impact on clinical characterization of the disease and its therapeutic management must be emphasized. Guidelines recommend that to be eligible for treatment, patients must have a viral load greater than or equal to HBV DNA 3.2 log IU/mL. Only 17% of our patients constantly exceeded this threshold in the three samples tested (M0, M2 and M4). Forty-two percent of patients would have been eligible or ineligible for treatment according to the viral load value at a given time. It therefore seems necessary to measure viral load for multiple occasions before deciding to treat.
More recently, studies have reported that the management of chronic hepatitis B can be optimized with HBsAg quantification, used as a biomarker for stratifying the risk of disease progression [
16] and for predicting treatment response mainly in patients receiving pegylated interferon (PEG-IFN) therapy [
17].
In clinical practice, HBsAg quantification cannot replace viral load measurement. Combining both measures has been shown to be important for monitoring the natural history of the disease and treatment outcome [
11]. We observed a significant reduction in the HBsAg level between the second and third visits. Consistent with other studies, we found no correlation between HBV DNA levels and HBsAg in these HBeAg-negative patients [
18]. It has been shown that the HBsAg / HBV DNA ratio, which reflects the association between HBsAg production and HBV replication, increased after seroconversion without any HBsAg level modification. Confirming immune control over viral replication was the first step of immune clearance. This ratio has been shown to be higher during the low-replicative phase, compared to immune-tolerant, immune-clearance and HBeAg negative hepatitis phase and to be repeatable regardless to ethnicity or genotype [
19-
21]. Nevertheless, in our study, the HBsAg / HBV DNA ratio was higher in samples with low HBV DNA values than in those with high HBV DNA values, as reported by others [
22]. These results suggest that the production of HBsAg is more conserved than the HBV DNA replication indicating that the association between HBsAg production and HBV DNA replication seems disconnected [
23,
24].
Acknowledgements
We thank Tamara Giles-Vernick for reading the manuscript and the reviewers to improve the manuscript.
Finally, the authors are also grateful to the patients and clinical teams for their commitment to the protocol Study.
Competing interests
The authors declare that they have no competing interests.
Authors’ contribution
All authors read and approved the final version of the manuscript. Conceived and designed the experiments: MV, FS, PSM. Recruited and collected clinical data: AS, MLE, FF, JD, AD, JMD. Performed the experiments: JMS, SM. Analyzed the data: LC, SM, JMS, FS, MV. Drafted the manuscript: LC, SM, JMS, FS, MV. Revised the manuscript: LC, SM, JMS, FS, MV, AS, MLE, FF, JD, AD, JMD, PSM.