In spite of an efficient hepatitis B virus (HBV) vaccine and successful vaccination campaigns, infection with HBV remains a significant cause of liver disease worldwide. The natural history of chronic hepatitis B is characterized by an initial, HBeAg-positive tolerant phase with the highest viremia levels, followed by an immunoreactive phase with declining viremia and active liver damage. HBeAg to anti-HBe seroconversion represents the transition from the replicative to the low/nonreplicative phase (1,2), usually associated with remission of liver damage. Recently, HBV genotypes have been documented to be strongly predictive of clinical outcomes. Based on the high nucleotide variability of the genome (>8%), HBV has been classified into 10 genotypes A–J (3), which show a specific geographic distribution: genotypes A and D are predominant in Europe, whereas genotypes B and C are mainly found in Asia. Co-infection with a genotype mixture representing quasi-species with minor and major dominant HBV strains has also been detected (4). Compared with genotypes A and B, genotypes C and D induce delayed anti-HBe seroconversion and a greater propensity for cirrhosis and HCC. There is little information regarding genotype changes during the long-term course of the disease. Jardi et al (4) report on genotype changes under antiviral treatment with lamivudine or adefovir in adults. In addition, genotype changes have also been detected in patients under interferon therapy (5,6). To date, only 2 studies investigated the occurrence and features of these events in children (5,7). The Asian study (7) identified a mean genotype shift of 2.8% in 460 children with chronic hepatitis B harboring predominantly genotype B. There were no genotype changes in the HBeAg-positive cohort, in line with the study of Jardi et al (4). Because genotype changes in this study were only observed in treated patients, the authors concluded that this phenomenon may reflect the selection of certain HBV quasi-species strains with differing sensitivity to treatment; however, the role and effect of genotype changes on the course of the disease remain to be explored. The aim of this study was to assess the frequency of HBV genotype changes in untreated and treated European children with chronic hepatitis B and spontaneous or therapy-associated anti-HBe seroconversion.
Patients belonged to 2 cohorts of European children with chronic hepatitis B consecutively collected in 2 different hospitals—Department of Pediatrics, HELIOS Medical Center Wuppertal, Witten/Herdecke University, Germany, and Clinica Medica 5, University of Padua, Italy. Epidemiological parameters were recorded on a regular basis, and serum samples were prospectively obtained at regular intervals of approximately 6 months during follow-up visits. The analysis of all stored sera was then performed in 1 charge in the pediatric research laboratory of HELIOS Medical Center Wuppertal.
The German cohort included 41 subjects. The epidemiological and serological features of the patients at baseline and the duration of follow-up are summarized in Table 1. Quantitative HBV DNA could be determined in 33 cases. The mean HBV copy number at presentation was 5.6 × 108/mL; 24 subjects (58.5%) received therapy during the observation period either with a nucleos(t)ide analog (lamivudine or adefovir), α-interferon, or a combination of both. All patients were included in ethically approved treatment trials. The Italian cohort consisted of 44 patients (Table 1). All of them were positive for HBeAg and qualitative HBV DNA test at baseline, and all of them seroconverted spontaneously to anti-HBe during follow-up.
HBsAg, HBeAg, anti-HBe, anti-HBs, and anti-HBc were assayed using commercial radioimmunoassays (Abbott Laboratories, North Chicago, IL). All patients were negative for antibodies to hepatitis C virus, hepatitis D virus, and human immunodeficiency virus. HBV DNA was determined quantitatively by a liquid hybridization assay (Hybride Capture Systems, Digene Diagnostics, Beltsville, MD).
An aliquot of 200 μL of each patient's serum was obtained to isolate viral DNA using the QIAamp Blood Kit (Qiagen, Hilden, Germany). The DNA was resolved in 50-μL distilled water. The pre-S1/S2 open reading frame was amplified by polymerase chain reaction (PCR) as previously described (5). PCR products were separated by agarose gel electrophoresis and visualized. The HBV strains were classified by restriction fragment length polymorphism to the genotypes A to D using the restriction endonucleases (Ava II, Dpn II; New England Biolabs, Frankfurt a.M., Germany). In some cases in which genotypes could not be determined by restriction fragment length polymorphism, PCR fragments were cloned into the pGEM-T easy vector (Promega, Baltimore, MD) and subsequently sequenced using the T7 or M13-RP sequencing primers, respectively. Sequence alignments to RefSeq records of HBV genomic sequences representing all genotypes were performed using Clustal W implemented into Mega 4.1 software (8) and manually inspected. Serum samples with 3 identical sequencing results (Phred quality score 20) were classified as single genotype, whereas in case of sequence variations, the genotype of the serum sample was annotated as mixed. Sequences from sampled DNA were assigned to specific genotypes by means of phylogenetic analyses. Tree calculations were conducted using the neighbor-joining method in MEGA 4.1. The bootstrap consensus tree inferred from 1000 replicates.
Testing for significant differences between groups was performed using a 1-way analysis of variance test for values with Gaussian distribution and the Mann-Whitney test for values without Gaussian distribution. The Kolmogorov-Smirnov test was used to rule out non-Gaussian distribution. The χ 2 test was applied to compare qualitative data, where appropriate. Additionally Pearson regression analysis was performed. Values for P < 0.05 were considered statistically significant. All analyses were performed using GraphPad version 5.01 (GraphPad Software Inc, La Jolla, CA).
The distribution and the effect of HBV genotypes in the 85 patients at entry and at last visit are shown in Table 2.
Factors Associated With Genotype Changes
In the first cohort of 41 patients, 21 (51.2%) seroconverted to anti-HBe. A genotype change was observed in 5 cases (12.2%), including 2 shifting from D to A, 2 from D to B, and 1 from A to D. In 4 (19%) of the 21 cases, the change was associated with anti-HBe seroconversion. Three of them with genotype change had antiviral treatment and seroconverted. Seroconversion tended to occur earlier in patients with genotype change when compared with patients with seroconversion without genotype change (3.2 vs 4.4 years, P = 0.24, n.s.). One subject with genotype change remained HBeAg positive. There was no association with antiviral treatment. Figure 1 illustrates the distribution of cases in relation to the cohorts, to HBeAg/anti-HBe status, to treatment, and to the proportion of patients with genotype changes. In the second cohort of 44 anti-HBe seroconverted patients, 11 (25%) genotype changes were documented during follow-up. Four (36.35%) subjects switched from D to A, 1 (9.1%) from D to A/D mix, 1 (9.1%) from D to B, 4 (36.35%) from A to D, and 1 (9.1%) from A to B. Table 2 illustrates the numbers and directions of detected genotype changes.
The proportion of genotype changes during long-term follow-up of patients who experienced anti-HBe seroconversion was comparable in both cohorts (19% vs 25%, n.s.). In total, 15 subjects (23%) had a genotype change. Six individuals (42.9%) changed from D to A, 3 (21.4%) from D to B, 5 (33.3%) from A to D, and 1 (7.1%) from A to B. The genotype distribution, denoting the proportion of the presence of genotypes A, B, C, and D, was different in anti-HBe–positive patients who did or did not experience genotype change (Fig. 2A). Individuals with genotype changes tended to have lower HBV copy numbers at diagnosis (1.8 × 108 vs 3.4 × 108 copies per milliliter) and seroconverted earlier when compared with patients without genotype changes (4.9 vs 5.9 years), but none of these differences were statistically significant. Genotype change occurred in close relation to seroconversion to anti-HBe and appeared in most cases (n = 14; 93.3%) shortly thereafter. Only in 1 case was the genotype change observed before seroconversion (Fig. 2B). Pearson correlation coefficient was 0.88, suggesting a close linear correlation between time of genotype change and seroconversion (Fig. 2C). No other factors including therapy were found to be associated with genotype change.
To our knowledge, this is the largest longitudinal study of chronic hepatitis B in children and adolescents providing information on the frequency and characteristics of HBV genotype changes during the course of the disease. As expected, the rate of genotype changes was variable in the different phases of chronic infection. During the immune-tolerant or immunoreactive phase preceding HBeAg to anti-HBe seroconversion, the rate of changes was low in our patients (4.8%) in agreement with data from previous studies in adults and children (4,6,7). Conversely, different rates of genotype changes, from 4% to 53% (4–7), were obtained in cohorts including variable proportions of anti-HBe–positive patients. It can be hypothesized that part of these patients had already seroconverted to anti-HBe and changed genotype before entry. In our study, all 85 children were HBeAg positive at entry; 65 (76.5%) cases later seroconverted to anti-HBe and 15 of them (23%) showed genotype change in timely correlation with anti-HBe seroconversion. Of the 20 patients who did not clear HBeAg, only 5% changed genotype during the survey.
The influence of antiviral therapy on the prevalence and distribution of genotype changes has been investigated in adults by Jardi et al (4), who found that A to D was the prevalent shift possibly because of a lower sensitivity of genotype A to nucleos(t)ide analogues. Hannoun et al (6) also reported genotype selection in adult patients receiving interferon therapy. In this study the change from genotype A to D was explained by a possible higher sensitivity of genotype A to interferon. Both of these studies experienced a selection bias because Hannoun et al did not include untreated patients as a control group, whereas Jardi et al (4) had included only patients receiving antiviral therapy. In children and adolescents, it has been clearly shown that genotype change is not significantly associated with interferon treatment (5). Our study revealed that 12.5% of treated and 21.3% of untreated subjects displayed genotype changes, indicating that such changes may occur independent of antiviral therapy. Based on the limited experience, several questions on the significance of HBV genotype changes remain unanswered. In Asian countries, genotypes B and C are predominant, and anti-HBe seroconversion rate is higher in patients infected with genotype B (9,10). Patients with genotype C seem to be at a higher risk for developing hepatocellular carcinoma (11). In European countries, the most prevalent genotypes are A and D. There is also evidence that patients with genotype D have a higher anti-HBe seroconversion rate than patients with genotype A (12). The course of the disease in subjects with genotype B and D may be associated with a higher inflammatory activity, which, in the long run, can promote seroconversion to anti-HBe (13).
HBV infection is a dynamic process, which requires the continuous selection of molecular variants to conform to changing environmental conditions. This selection may occur either during antiviral treatment (14) or under natural pressure of the host immune system in the absence of modulating therapy (15). During this process, viral quasi-species could play an important role. It seems rather likely that some patients, if not all, become infected with a mixture of variants (quasi-species), which harbor 2 different genotypes, 1 of which will eventually emerge. This hypothesis is underscored by the detection of mixed genotypes A and D in one of our patients before and another one after anti-HBe seroconversion. Under increased immunological pressure, in particular during HBeAg to anti-HBe seroconversion, the predominant genotype could be suppressed, whereas the initially suppressed could eventually emerge. This is underscored by the observation in our study and others that the probability of spontaneous anti-HBe seroconversion in children increases with higher aminotransferase levels, reflecting an increased immunological reactivity (16). In conclusion, about one-fourth of European children with chronic hepatitis B will change HBV genotype preferably after recent HBeAg to anti-HBe seroconversion. The mechanisms implicated in this event are largely unknown, but deserve further investigation to tailor treatment of hepatitis B in children.
1. Jonas MM, Block JM, Haber BA, et al. Treatment of children
with chronic hepatitis B
virus infection in the United States: patient selection and therapeutic options. Hepatology
2. Hadziyannis SJ. Natural history of chronic hepatitis B
in Euro-Mediterranean and African countries. J Hepatol
3. Norder H, Courouce AM, Magnius LO. Complete genomes, phylogenetic relatedness, and structural proteins of six strains of the hepatitis B virus, four of which represent two new genotypes. Virology
4. Jardi R, Rodriguez-Frias F, Schaper M, et al. Analysis of hepatitis B genotype changes
in chronic hepatitis B
infection: influence of antiviral therapy. J Hepatol
5. Gerner PR, Friedt M, Oettinger R, et al. The hepatitis B virus seroconversion to anti-HBe is frequently associated with HBV genotype changes
and selection of preS2-defective particles in chronically infected children
6. Hannoun C, Krogsgaard K, Horal P, et al. Genotype mixtures of hepatitis B virus in patients treated with interferon. J Infect Dis
7. Ni YH, Chang MH, Wang KJ, et al. Clinical relevance of hepatitis B virus genotype in children
with chronic infection and hepatocellular carcinoma. Gastroenterology
8. Tamura K, Dudley J, Nei M, et al. MEGA4: Molecular Evolutionary Genetics Analysis (MEGA) software version 4.0. Mol Biol Evol
9. Ishikawa K, Koyama T, Masuda T. Prevalence of HBV genotypes in asymptomatic carrier residents and their clinical characteristics during long-term follow-up: the relevance to changes in the HBeAg/anti-HBe system. Hepatol Res
10. Livingston SE, Simonetti JP, Bulkow LR, et al. Clearance of hepatitis Be antigen in patients with chronic hepatitis B
and genotypes A, B, C, D, and F. Gastroenterology
11. Iloeje UH, Yang HI, Chen CJ. Natural history of chronic hepatitis B
: what exactly has REVEAL revealed? Liver Int
12. Verschuere V, Yap PS, Fevery J. Is HBV genotyping of clinical relevance? Acta Gastroenterol Belg
13. Kao JH. Hepatitis B viral genotypes: clinical relevance and molecular characteristics. J Gastroenterol Hepatol
14. Ji F, Zhou L, Ma S, et al. Dynamic changes of HBV quasispecies and deletion patterns in a chronic hepatitis B
patient. J Med Virol
15. Cui XJ, Cho YK, Song HJ, et al. Molecular characteristics and functional analysis of full-length hepatitis B virus quasispecies from a patient with chronic hepatitis B
virus infection. Virus Res
16. Wu JF, Su YR, Chen CH, et al. Predictive effect of serial serum alanine aminotransferase levels on spontaneous HBeAg seroconversion in chronic genotype B and C HBV-infected children
. J Pediatr Gastroenterol Nutr