Influence of genotype on hepatitis B surface antigen kinetics in hepatitis B e antigen-negative patients treated with pegylated interferon-α2aRami Moucari1,2,3, Michelle Martinot-Peignoux1,2,3, Vincent Mackiewicz4, Nathalie Boyer1,2,3, Marie-Pierre Ripault1,2,3, Corinne Castelnau1,2,3, Laurence Leclere1,2,3, Agnes Dauvergne5, Dominique Valla1,2,3, Michel Vidaud4, Marie-Hélène Nicolas-Chanoine2, Patrick Marcellin1,2,3,*
Date accepted: 07 July 2009
Date published online: 10 December 2009
Copyright (c) 2009 International Medical Press, all rights reserved.
Background: The aim of this study was to assess the influence of hepatitis B virus (HBV) genotypes on serum hepatitis B surface antigen (HBsAg) kinetics in hepatitis B e antigen (HBeAg)-negative patients treated with pegylated interferon-α2a (PEG-IFN-α2a).
Methods: A total of 48 consecutive patients treated with PEG-IFN-α2a (180 μg/week) for 48 weeks were assessed. HBV genotype was determined. Serum HBV DNA and HBsAg were assessed at baseline, during treatment (weeks 12, 24 and 48) and during follow-up (weeks 72 and 96).
Results: The distribution of HBV genotype was A 27%, B 17%, C 12%, D 29% and E 14%. Mean ±sd pretreatment serum HBV DNA (6.9 ±1.5 log10 copies/ml) was not different between genotypes and decreased under treatment in all genotypes without significant difference. Mean ±sd pretreatment serum HBsAg (3.6 ±0.6 log10 IU/ml) was significantly different between genotypes (P<0.001), with high levels in genotypes A and C, intermediate levels in genotypes D and E, and low levels in genotype B (4.0 ±0.3, 4.1 ±0.7, 3.6 ±0.5, 3.6 ±0.4 and 2.7 ±0.6 log10 IU/ml, respectively). Serum HBsAg decreased under treatment in all genotypes with a significant difference. At the end of treatment, mean ±sd decrease was high in genotype A, intermediate in genotypes B and D, and low in genotypes C and E (1.3 ±1.6, 0.7 ±0.7, 0.6 ±0.9, 0.4 ±1.0 and 0.4 ±0.9 log10 IU/ml, respectively; P<0.001). During follow-up, serum HBsAg continued to decrease in genotypes A and D, whereas rebound was observed in genotypes B, C and E.
Conclusions: HBV genotype has a strong influence on serum HBsAg kinetics during PEG-IFN-α2a therapy in HBeAg-negative patients.
Chronic hepatitis B (CHB) infection is a dynamic process resulting from a continuous interplay between the virus (replication level and emergence of mutations), the host (age at the time of infection and immune competence status) and the environment (infection with other viruses and antiviral therapy). The analysis of the entire hepatitis B virus (HBV) genome has revealed the existence of eight different genotypes (A–H) . In contrast to chronic hepatitis C virus infection (where the genotype has a major influence on the response to treatment, and the duration and dose of therapy), the therapeutic relevance of HBV genotype is not well determined. A recent analysis of published literature showed that HBV genotypes do not influence the outcome after treatment with nucleoside/nucleotide analogues. However, interferon (IFN)-based therapies showed higher efficacy (hepatitis B e antigen [HBeAg] seroconversion and HBV DNA undetectability) in all patients (HBeAg-positive and HBeAg-negative) with genotype A versus D, and in HBeAg-positive patients with genotype B versus C .
Loss of serum hepatitis B surface antigen (HBsAg) is the closest to a cure outcome in CHB and might be achievable at a substantial rate after a limited course of IFN or pegylated interferon-α2a (PEG-IFN-α2a) therapy [3,4]. The clearance of HBsAg confers excellent long-term prognosis, even in patients who develop cirrhosis during the early phases of the disease . Influence of HBV genotypes on HBsAg loss after PEG-IFN-α2a therapy has not been well determined. Finally, on-treatment quantification of HBsAg was recently shown to be an excellent tool to predict sustained response to PEG-IFN-α2a in HBeAg-negative patients . The aim of this study was to assess the influence of HBV genotypes on the kinetics of serum HBsAg in HBeAg-negative patients treated with PEG-IFN-α2a.
A total of 48 consecutive HBeAg-negative patients were evaluated. They were described in a previous publication . CHB was documented by the presence of HBsAg in serum for >6 months and by liver biopsy showing histological features of CHB compatible with HBV infection (Metavir score). Patients were treated with PEG-IFN-α2a at a dose of 180 μg/week for 48 weeks and were followed for an additional 48 weeks after the end of treatment (EOT). EOT response was defined as undetectable serum HBV DNA at the EOT (48 weeks). Sustained virological response (SVR) was defined as undetectable serum HBV DNA at 24 weeks after treatment cessation. All patients gave their informed consent before liver biopsy.
HBV genotype was determined using the TRUGENE™ HBV Genotyping Kit (Siemens Medical Solutions Diagnostics, Saint-Denis, France). Serum HBV DNA was measured using the TaqMan® PCR assay (COBAS® TaqMan® ; Roche Diagnostics, Meylan, France) at baseline, during therapy (weeks 12, 24 and 48) and during follow-up (weeks 72 and 96). Serum HBsAg was quantified at the same intervals using the Abbott Architect HBsAg QT (Abbott Diagnostics, Rungis, France), a chemiluminescent microparticle immunoassay. The concentration of HBsAg in patients was determined using a previously generated Architect HBsAg calibration curve (range 0.05–250 IU/ml). Samples were diluted 1:20 and 1:500 with the Architect HBsAg diluent in order to expand the upper limit of the dynamic range from 250 to 125,000 IU/ml.
Continuous variables were summarized as mean ±sd and categorical variables as frequency (percentage). The Spearman’s rank correlation was used to quantify the association between continuous or ordered categorical variables, whereas the χ2 test was used for nominal categorical variables. Analysis of variance (standard or non-parametric as required) was used to test for the association between continuous outcomes and nominal categorical variables. Statistical significance was determined at a P-value of <0.05. Data handling and analysis were performed with SPSS software for Windows (version 13.0; SPSS, Inc., Chicago, IL, USA)
Overall, 40 (83%) patients were male, 32 (67%) patients were Caucasian, 10 (21%) patients were African and 6 (12%) patients were Asian. The distribution of HBV genotype was A 27%, B 17%, C 12%, D 29% and E 14%. Mean serum HBV DNA level was 6.9 ±1.5 log10 copies/ml. Mean serum HBsAg level was 3.6 ±0.6 log10 IU/ml. Serological tests for hepatitis C virus, hepatitis delta virus and HIV were negative in all patients. Liver histology showed severe fibrosis (Metavir score F3–F4) in 50% of patients. Of all 48 patients, 30 (62%) showed an EOT response, but only 12 (25%) patients achieved SVR.
Pretreatment serum HBV DNA levels were not different according to HBV genotype. Mean ±sd levels were 7.0 ±1.7, 6.2 ±1.8, 7.7 ±1.7, 6.7 ±1.4 and 6.9 ±1.4 log10 copies/ml for genotypes A, B, C, D and E, respectively (P=0.669). During the treatment course, serum HBV DNA decreased in all genotypes without significant difference (Figure 1). At treatment week 12, mean ±sd decreases were 3.1 ±1.6, 2.8 ±1.6, 2.8 ±2.5, 2.6 ±2.4 and 3.8 ±1.4 log10 copies/ml for genotypes A, B, C, D and E, respectively (P=0.808). At treatment week 24, mean ±sd decreases were 3.8 ±2.4, 3.8 ±1.5, 3.3 ±4.1, 3.4 ±2.3 and 4.2 ±1.5 log10 copies/ml for genotypes A, B, C, D and E, respectively (P=0.966). At treatment week 48, mean ±sd decreases were 4.3 ±2.4, 3.9 ±2.0, 3.9 ±4.5, 3.7 ±2.4 and 3.9 ±2.1 log10 copies/ml for genotypes A, B, C, D and E, respectively (P=0.984).-
Pretreatment serum HBsAg levels were significantly different according to HBV genotype (P<0.001). Genotypes A and C had the highest levels (4.0 ±0.3 and 4.1 ±0.7 log10 IU/ml, respectively). Genotype B had the lowest levels (2.7 ±0.6 log10 IU/ml). Genotypes D and E had intermediate levels (3.6 ±0.5 and 3.6 ±0.4 log10 IU/ml).
During the treatment course, serum HBsAg levels decreased in all genotypes with different kinetics (Figure 2A). At treatment week 12, mean ±sd decreases were similar between genotypes A, B, C, D and E (0.2 ±0.3, 0.2 ±0.5, 0.2 ±0.5, 0.1 ±0.8 and -0.06 ±0.4, respectively [P=0.848]); however, at treatment week 24 and week 48, genotype A showed a greater decrease (0.7 ±0.8 [P<0.005] and 1.3 ±1.6 log10 IU/ml [P<0.001], respectively) compared with non-A genotypes (genotype B 0.5 ±0.6 and 0.7 ±0.7 log10 IU/ml, respectively, genotype C 0.2 ±0.5 and 0.4 ±1.0 log10 IU/ml, respectively, genotype D 0.4 ±0.8 and 0.6 ±0.9 log10 IU/ml, respectively, and genotype E 0.2 ±0.7 and 0.4 ±0.9 log10 IU/ml, respectively; Figure 2B). Moreover, genotypes B and D also showed greater decrease (P<0.01) when compared with genotypes C and E. Of note, pretreatment serum alanine amiontransferase level was also correlated with HBsAg decrease, in contrast to age, gender, ethnicity, pretreatment HBV DNA and HBsAg levels, and liver histology (Table 1).-
ALT, alanine aminotransferase; HBV, hepatitis B virus; HBsAg, hepatitis B surface antigen; NA, not applicable.
Analyses were repeated for patients who developed EOT response and showed the same profile. Genotype A showed a greater decrease in serum HBsAg at treatment weeks 24 and 48 (1.2 ±0.9 [P<0.001] and 2.2 ±1.6 log10 IU/ml [P<0.001], respectively) compared with non-A genotypes (genotype B 0.4 ±0.5 and 0.6 ±0.8 log10 IU/ml, respectively, genotype C 0.2 ±0.5 and 0.2 ±0.4 log10 IU/lml, respectively, genotype D 0.6 ±0.8 and 0.9 ±0.8 log10 IU/lml, respectively, and genotype E 0.4 ±1.0 and 0.6 ±1.2 log10 IU/lml, respectively). Analyses were also repeated for SVRs (Table 2 and Figure 3A and 3B), and showed the same tendency when genotype A and non-A genotypes were compared (Figure 3C).
Serum hepatitis B virus (HBV) DNA levels are presented in log10 copies/ml and hepatitis B surface antigen (HBsAg) levels are presented in log10 IU/ml.
Serum HBsAg levels showed different kinetics after treatment discontinuation according to HBV genotype (Figure 2). Genotypes A and D showed a continuous decrease. In genotype A, mean levels were 2.3 ±1.7 and 2.2 ±1.8 log10 IU/ml at follow-up weeks 24 and 48, respectively, compared with 2.6 ±1.5 log10 IU/ml at the EOT. In genotype D, mean levels were 3.0 ±0.8 and 2.8 ±0.9 log10 IU/ml at follow-up weeks 24 and 48, respectively, compared with 3.0 ±1.1 log10 IU/ml at the EOT. By contrast, other genotypes showed a rebound in serum HBsAg. In genotype B, mean levels were 2.0 ±1.7 and 2.1 ±1.8 log10 IU/ml at follow-up weeks 24 and 48, respectively, compared with 1.9 ±1.5 log10 IU/ml at the EOT. In genotype C, mean levels were 3.8 ±1.3 and 3.9 ±1.4 log10 IU/ml at follow-up weeks 24 and 48, respectively, compared with 3.7 ±1.4 log10 IU/ml at the EOT. In genotype E, mean levels were 3.7 ±1.2 and 3.5 ±1.4 log10 IU/ml at follow-up weeks 24 and 48, respectively, compared with 3.2 ±1.3 log10 IU/ml at the EOT.
This study provides important information about the influence of HBV genotypes on serum HBsAg kinetics in HBeAg-negative patients treated with PEG-IFN-α2a. Because of the unique distribution of HBV genotypes in Asian and Western countries, assessment of the therapeutic relevance of HBV genotype in large clinical trials has been limited by the geographical origin of the population included in these trials . Interestingly, the main HBV genotypes (A–E) were represented in this single-centre study, reflecting the effect of continuous population migration from Asia and Africa to Europe.
The results of this study show that HBV genotype has a strong influence on serum HBsAg kinetics (Figure 2); however, these results should be analysed with caution given the low number of patients leading to a high risk of type 1 error. Interestingly, this influence was independent of HBV DNA kinetics, which was not significantly different according to HBV genotypes in this study (Figure 1). Moreover, this influence seems to be independent of response to treatment as the same tendency was observed when patients who achieved EOT and SVR were assessed separately (Figure 3).
Patients with genotype A showed a steeper decrease of HBsAg after treatment week 24 compared with other genotypes and achieved >1 log10 IU/ml decrease by the end of therapy. However, patients with genotypes B and D also had decreased HBsAg levels during treatment, although less pronouncedly than patients with genotype A, but greater than patients with genotype C and E, achieving a substantial decrease (>0.5 log10 IU/ml) by the EOT. These findings suggest that CHB patients infected with genotype A might be the best candidates for a 48-week course of PEG-IFN-α2a therapy when HBsAg loss is considered as a major end point. However, patients with genotypes B and D might also benefit from this therapeutic strategy or from a combination with potent analogues [7,8] to consolidate the decrease in serum HBsAg. Furthermore, our study might lead to a larger adaptive therapy protocol in which decisions on how long to treat are based on HBsAg kinetics measured at key intervals. Importantly, serum HBsAg continued to decrease in genotype A as well as genotype D during follow-up after the end of PEG-IFN-α2a therapy, highlighting again that PEG-IFN-α2a therapy should not be restricted to patients with genotype A. Interestingly, the influence of genotype A on HBsAg decrease was similarly found in a large recent study ; however, HBsAg decrease was assessed at the EOT only and comparison between genotypes was not analysed according to treatment response in this study. In HBeAg-positive patients treated with PEG-IFN-α2a, a higher rate of HBsAg loss was also reported in genotype A versus non-A genotypes, probably reflecting a difference in HBsAg kinetics .
Finally, the molecular virological factors that contribute to clinical and therapeutic differences among HBV genotypes remain unclear. An IFN sensitivity region within the HBV genotype has been investigated with negative results . In HBeAg-positive patients treated with conventional IFN, a high rate of HBeAg clearance was observed in genotype A versus non-A genotypes, and was shown to be associated with the emergence of core promoter mutations . However, genotype A patients were shown to exhibit a high rate of HBeAg and HBsAg loss, not only as a result of IFN-based therapy but also spontaneously .
In conclusion, HBV genotype has a strong influence on serum HBsAg kinetics during PEG-IFN-α2a therapy in HBeAg-negative patients. Future clinical trials should be designed according to HBV genotype to better elucidate the therapeutic relevance of these findings.
PM advises, is a consultant for or is in the speakers’ bureau of Roche, Schering–Plough, Gilead, Bristol–Myers Squibb, Novartis, Vertex and Tibotec. MP has received grants from Roche, Schering–Plough and Gilead. All other authors declare no competing interests.