Peginterferon alpha-2b plus adefovir induce strong cccDNA decline and HBsAg reduction in patients with chronic hepatitis B


  • Karsten Wursthorn,

    1. Department of Medicine, University Hospital Hamburg–Eppendorf, Hamburg, Germany
    Search for more papers by this author
    • K. Wursthorn and M. Lutgehetmann contributed equally to this work.

  • Marc Lutgehetmann,

    1. Department of Medicine, University Hospital Hamburg–Eppendorf, Hamburg, Germany
    Search for more papers by this author
    • K. Wursthorn and M. Lutgehetmann contributed equally to this work.

  • Maura Dandri,

    1. Department of Medicine, University Hospital Hamburg–Eppendorf, Hamburg, Germany
    Search for more papers by this author
  • Tassilo Volz,

    1. Department of Medicine, University Hospital Hamburg–Eppendorf, Hamburg, Germany
    Search for more papers by this author
  • Peter Buggisch,

    1. Department of Medicine, University Hospital Hamburg–Eppendorf, Hamburg, Germany
    Search for more papers by this author
    • Potential conflict of interest: Dr. Zollner received travel grants from Gilead. Dr. Petersen received grants from Gilead. Dr. Buggisch is on the Speakers' Bureau for and received grants from Schering-Plough. He also received grants from Gilead.

  • Bernhard Zollner,

    1. Department of Microbiology, University Hospital Hamburg–Eppendorf, Hamburg, Germany
    Search for more papers by this author
    • Potential conflict of interest: Dr. Zollner received travel grants from Gilead. Dr. Petersen received grants from Gilead. Dr. Buggisch is on the Speakers' Bureau for and received grants from Schering-Plough. He also received grants from Gilead.

  • Thomas Longerich,

    1. Institute of Pathology, University of Heidelberg, Heidelberg, Germany
    Search for more papers by this author
  • Peter Schirmacher,

    1. Institute of Pathology, University of Heidelberg, Heidelberg, Germany
    Search for more papers by this author
  • Frauke Metzler,

    1. Department of Medicine, University Hospital Hamburg–Eppendorf, Hamburg, Germany
    Search for more papers by this author
  • Myrga Zankel,

    1. Essex Pharma GmbH, Munich, Germany
    Search for more papers by this author
  • Conrad Fischer,

    1. Gilead Sciences Inc, Foster City, CA
    Search for more papers by this author
  • Graeme Currie,

    1. Gilead Sciences Inc, Foster City, CA
    Search for more papers by this author
  • Carol Brosgart,

    1. Gilead Sciences Inc, Foster City, CA
    Search for more papers by this author
  • Joerg Petersen

    Corresponding author
    1. Department of Medicine, University Hospital Hamburg–Eppendorf, Hamburg, Germany
    • Medizinische Klinik und Poliklinik I, Zentrum für Innere Medizin, Universitätsklinikum Hamburg—Eppendorf, Martinistr. 52, D-20246 Hamburg, Germany
    Search for more papers by this author
    • Potential conflict of interest: Dr. Zollner received travel grants from Gilead. Dr. Petersen received grants from Gilead. Dr. Buggisch is on the Speakers' Bureau for and received grants from Schering-Plough. He also received grants from Gilead.

    • fax: (49) 40-42803-8065


Hepatitis B virus (HBV) covalently closed circular DNA (cccDNA) is responsible for persistent infection of hepatocytes. The aim of this study was to determine changes in intrahepatic cccDNA in patients with chronic hepatitis B (CH-B) during 48 weeks of antiviral therapy and its correlation to virological, biochemical, and histological parameters. Twenty-six HBsAg-positive CH-B patients received combination treatment with pegylated interferon alpha-2b (peg-IFN) and adefovir dipivoxil (ADV) for 48 weeks. Paired liver biopsies from before and at the end of treatment were analyzed for intrahepatic HBV-DNA. Median serum HBV-DNA had decreased by −4.9 log10 copies/mL at the end of treatment and was undetectable in 13 individuals (54%). Median intrahepatic total HBV-DNA and cccDNA had decreased by −2.2 and −2.4 log10, respectively. Changes in intracellular HBV-DNA positively correlated with HBsAg serum reduction and were accompanied by a high number of serological responders. Eight of 15 HBeAg-positive patients lost HBeAg, and five developed anti-HBe antibodies during treatment. These eight patients exhibited lower cccDNA levels before and at the end of therapy than did patients without HBeAg loss. Four patients developed anti-HBs antibodies. ALT normalized in 11 patients. The number of HBs-antigen- and HBc-antigen-positive hepatocytes was significantly lower after treatment, suggesting the involvement of cytolytic mechanisms. In conclusion, combination therapy with peg-IFN and ADV led to marked decreases in serum HBV-DNA and intrahepatic cccDNA, which was significantly correlated with reduced HBsAg. (HEPATOLOGY 2006;44:675–684.)

Infection with hepatitis B virus (HBV) causes acute and chronic hepatitis and is strongly associated with the development of cirrhosis and hepatocellular carcinoma. Immediately after infection of hepatocytes, the viral DNA is transferred to the nucleus, where the viral polymerase is removed, and the double-stranded, open circular DNA is converted to a covalently closed circular DNA molecule (cccDNA). During chronic HBV infection (CH-B), cccDNA accumulates in hepatocyte nuclei, apparently at a level of about 5–50 copies per cell, where it persists as a minichromosome and functions as the template for the transcription of viral genes.1 The RNA pregenome, in addition to producing capsid and polymerase proteins, becomes encapsidated and is reverse-transcribed. A particularity of the hepadnavirus life cycle is that DNA-containing nucleocapsids can either recycle back to the nucleus to amplify and maintain the pool of cccDNA or become enveloped and secreted into the blood, where new viral particles can spread to other hepatocytes.2, 3 Because cccDNA is the transcriptional template of the virus, it is required for maintenance of HBV infection.

Evidence from the woodchuck hepatitis virus system indicated that the pool of cccDNA persisted even when viral production was strongly reduced by the presence of nucleoside analogues.4, 5 Woodchuck studies6, 7 and recent clinical trials8 demonstrated that significant reduction in the cccDNA pool was achieved only after long-term antiviral therapy. As HBV polymerase inhibitors do not have a direct effect on cccDNA formation, the decreased intrahepatic cccDNA levels are supposed to be derived from the lack of sufficient recycling of viral nucleocapsids to the nucleus because of the strong inhibition of viral DNA synthesis in the cytoplasm and less incoming viruses from the blood. Furthermore, experiments with HBV-replicating transgenic mice9 and chimpanzees10, 11 have shown that inflammatory cytokines can efficiently suppress viral replication through noncytolytic immune-mediated mechanisms that also contribute to diminishing the cccDNA reservoirs of infected cells. However, these processes may require a very long time, given the long half-life of both hepatocytes and cccDNA in the absence of cell division.4 On the other hand, histological data indicate that liver cell death and regeneration both occur in transiently and chronically infected livers, suggesting that clearance of hepadnavirus infection is accompanied by immune-mediated destruction of infected hepatocytes.12, 13 Yet it has been shown that very low levels of cccDNA can persist indefinitely in a small number of liver cells, possibly explaining the lifelong immune response to HBV despite clinical resolution of infection14 and the clinical relapse of hepatitis B during chemotherapy and other immunosuppressive regimens.

The nucleoside analogues lamivudine and adefovir dipivoxil (ADV) are potent inhibitors of HBV replication.15 Because of the stability of cccDNA in infected cells, to achieve viral eradication, many years of drug administration may be required, which eventually leads to selection of resistant viruses.16, 17 Thus, monitoring cccDNA in liver samples may become an end point of antiviral therapy if a threshold can be determined below which host-mediated mechanisms or combination antiviral therapy are likely to establish control of viral replication in most patients.

Despite the crucial role of cccDNA during persistent infection and the importance of understanding resolution mechanisms, scant data from patients have been reported. Detection and quantification of cccDNA from liver biopsies were recently demonstrated using a selective real-time polymerase chain reaction (PCR) assay.8, 18 Those studies found that cccDNA level strongly correlated with total intracellular HBV DNA level and that, in general, intrahepatic HBV-DNA levels were substantially lower after HBeAg seroconversion in patients receiving oral ADV for 1 year.8 Interestingly, a significant, 0.8 log decrease in cccDNA copies per cell was demonstrated after nucleotide analogue therapy. However, histological analysis of liver biopsies suggested that changes in cccDNA occurred mainly by noncytolytic mechanisms because the number of HBV antigen-positive cells remained unchanged at the end of treatment.8 In a recent study by Sung et al.,18 it was shown that intrahepatic cccDNA and total HBV-DNA levels were superior to serum HBV-DNA as predictors of sustained virological response to antiviral therapy in patients who received either lamivudine monotherapy or combination therapy of lamivudine and pegylated interferon alpha-2b (peg-IFN) for 1 year. Because liver biopsies were taken only at the end of therapy, the impact of combination therapy on intracellular viral DNA reduction could not be investigated.

To evaluate the efficacy of combination therapy with parenteral peg-IFN and oral ADV, 26 HBsAg-positive CH-B patients were included in an open-label single-center pilot study. Paired biopsies were collected before and after 48 weeks of treatment to gain insight into intrahepatic cccDNA persistence, mechanisms of cccDNA clearance, and a possible noninvasive serum surrogate parameter for cccDNA.


CH-B, chronic hepatitis B infection; cccDNA, covalently closed circular DNA; HBeAg, hepatitis Be antigen; anti-HBe, anti–hepatitis Be antibody; HBsAg, hepatitis Bs antigen; ADV, adefovir dipivoxil.

Patients and Methods

Patient Characteristics and Study Design.

Twenty-six CH-B patients (20 men, 6 women) with a median age of 34 years (range 19–55) were prospectively recruited in a single-center pilot study of combination therapy with peg-IFN alpha-2b (PegIntron, Schering-Plough, Kenilworth, NJ) and ADV (Hepsera, Gilead Sciences, Foster City, CA). The study was approved by the German Federal Ethical Committee. All patients had the documented presence of hepatitis B surface antigen (HBsAg+) in serum for more than 6 months. At the screening visit, which preceded the start of therapy by at most 4 weeks, patients had to have elevated levels of serum HBV-DNA (>105 copies/mL) and of alanine aminotransferase, ALT (>1.2 times upper limit of normal). Both drugs were given simultaneously for 48 weeks, ADV as 10 mg orally daily and 12 kDa peg-IFN at a dose of 1.5 μg/kg body weight subcutaneously once a week.

Liver biopsies were taken before (n = 26) and after 48 weeks of combination therapy (n = 23) for histology and intrahepatic viral DNA analysis. Biopsy specimens were stored at −80°C until experimental analysis. All patients were seen and monitored every 4 weeks for biochemical parameters and serological markers of hepatitis. End-of-treatment virological response was defined as undetectable serum HBV-DNA level (<100 copies/mL). Serological response was defined as HBeAg seroconversion (disappearance of HBeAg, appearance of anti-HBe), and biochemical response was defined as ALT normalization at the end of treatment.

Genotyping and Viral Resistance Assay.

HBV-DNA genotyping was performed using the commercial INNO-Lipa HBV Genotyping assay (Innogenetics, Belgium). The HBV reverse-transcriptase domain was sequenced at baseline and at the end of treatment. Corresponding sequences were checked for mismatches. If there were viral nonresponse (>1 × 105 copies/mL), an HBV genomewide sequence scan was performed.

Virological, Biochemical, and Serological Analyses.

Serum aminotransferases were determined by a sequential multiple autoanalyzer. HBeAg and anti-HBe were measured by enzyme-linked immunosorbent assays (Sanofi Diagnostics Pasteur, Freiburg, Germany). Sera were stored at −80°C until evaluation.

Two assays were used to quantify HBsAg levels in patient sera. First, HBsAg concentration was determined in serum samples with a quantitative assay of electro-immunodiffusion (QIE, Laurell method), which relies on a combination of electrophoresis and immunoprecipitation procedures.19 Briefly, serum samples were loaded on an agarose gel containing HBsAg antibodies. Precipitates formed in electrophoresis could be measured and correlated with a known standard HBsAg run in parallel. The sensitivity of the assay ranges between 0.5 and 50 μg/mL HBsAg. Serum samples that turned out to have HBsAg levels above 50 μg/mL were diluted 1:10 with phosphate-buffered saline solution (PBS) for quantitative measurement. For quantitative detection of HBsAg levels below 1 μg/mL, we used the highly sensitive immunoassay Architect ci8200 integrated system (Abbott Diagnostics, Germany). Samples were measured pure and diluted 1:10 in HBsAg Manual Diluent (Abbott Diagnostics), as suggested by the manufacturer. The sensitivity of the Architect assay ranges between 0.05 and 250 IU/mL HBsAg, which corresponded to a range of approximately 0.1–500 ng/mL HBsAg when the purified HBsAg (Hytest, Finland) was used as the standard. Purified HBsAg (Hytest, Finland) was also used in the Laurell quantitative assay for comparison.

Serum HBV DNA was quantified as previously described.20 The assay detects HBV-DNA using real-time fluorescent-probe PCR (TaqMan, Roche, Mannheim, Germany), with a lower limit of quantification of 100 copies/mL.

Intrahepatic HBV-DNA Evaluation.

Intrahepatic HBV-DNA species were analyzed as previously described.8 In short, after extraction of DNA from liver biopsies using a MasterPure DNA purification kit (Epicentre, Biozym, Germany), each DNA sample was used for amplification in the Light Cycler system (Roche Diagnostics, Mannheim, Germany) with specific fluorescence hybridization probes. After treatment with plasmid-safe DNase, cccDNA-specific primers were used to amplify and quantify the cccDNA. Intrahepatic HBV-DNA values were measured in three independent assays and normalized for cellular DNA content using a commercially available beta-globin gene kit (Roche DNA Control Kit; Roche Diagnostics) as previously described.8

Histological Analysis and Immunostaining for HBcAg and HBsAg.

Inflammation and fibrosis in the liver biopsies were assessed using the modified hepatic activity index (mHAI—the Ishak score),21 the Desmet score,22 and the semiquantitative scoring system for fibrosis (SSS—the Chevallier score23). Histological samples from baseline and week 48 liver biopsies were blinded for analysis and independently evaluated by two pathologists (P.S. and T.L.), and a final consensus was reached.

Immunostaining for HBsAg and HBcAg was performed on a Dako Auto-Stainer using a Dako Envision+ system with peroxidase and 3,3′-diaminobenzidine chromogen (Dako, Hamburg, Germany). Primary rabbit polyclonal antibodies to HBcAg (Dako), at a 1:600 dilution, and monoclonal antibodies to HBsAg (Zymed, San Francisco, CA), at a 1:2,000 dilution, were used.

Statistical Analysis.

Statistical comparisons were performed using XLStat version 7.5 (Adinsoft, Brooklyn, NY). For nonparametric pairwise comparisons, the Spearman rank correlation coefficient was used, and the Wilcoxon rank sum test was used for two-group comparisons. The Mann-Whitney Test was used for comparison of two independent groups. A P value < .05 was considered significant. All statistical tests were two-sided.


Clinical Characteristics.

Twenty-six patients with chronic HBV infection received combination treatment with ADV and peg-IFN for 48 weeks. Fifteen patients were positive for HBeAg before treatment (Table 1). At baseline, HBeAg-positive patients had a median ALT level three times over the upper limit normal, ULN (range 1.2–15.1), and median serum HBV DNA was 4 × 107 copies/mL (range 1 × 106–2 × 109 copies/mL). Among HBeAg-negative patients, the median ALT level was 1.6 (range 1–5), and median serum HBV-DNA was 1 × 104 copies/mL (range 2 × 103–4 × 105 copies/mL) (Table 1). Although all HBeAg-negative patients fulfilled the criteria at the screening visit (HBV DNA > 1 × 105, ALT > 1.2), fluctuations in serum HBV DNA and ALT levels were observed in some patients at baseline.24 Genotype distribution was as follows: eight patients were infected with genotype A, one patient with genotype C, 14 patients with genotype D, and two patients with genotype G. One patient could not be genotyped. Before the start of combination therapy, all patients had a liver biopsy, with a median necroinflammatory score of 2 (range 1–3) and fibrosis score of 1 (range 0–4).

Table 1. Patient Characteristic at Baseline and After 48 Weeks of Combination Therapy
Patient CharacteristicsBaselineWeek 48
Median Age, years3435
Serum HBV DNA (copies/mL, median)7.5 × 1061.0 × 102
 HBeAg-pos.4 × 107 (1 × 106–2 × 109)3 × 103 (<100–2 × 108)
 HBeAg-neg.1 × 104 (1 × 103–4 × 105)<100 (<100–9 × 103)
ALT (×ULN, median)3.11.1
 HBeAg-pos.3.1 (1.2–15.1)1.2 (0.7–2.4)
 HBeAg-neg.1.6 (1–5)1.1 (0.5–2)

Twenty-four patients were followed until the end of treatment, whereas two patients were lost to follow-up because they left the country (patients 22 and 24). After 48 weeks of therapy, virological response occurred in 13 of 24 patients (54%), and median serum HBV-DNA was 100 copies/mL (range <100-2 × 108 copies/mL), indicating a median 4.9 log10 reduction from baseline.

As shown in Table 2, 8 of the 15 patients who were initially positive for HBeAg became HBeAg-negative (53%) by the end of therapy, and five of these eight patients also developed anti-HBe over the course of treatment (33%). Of the patients with HBeAg loss, five had genotype D at baseline, two had genotype G, and one had genotype C. Furthermore, four patients (16%) lost HBsAg and developed anti–hepatitis Bs antibody (anti-HBs), as shown in Table 2. The genotypes these patients had were distributed equally. ALT improved in 21 patients (84%). Median ALT levels decreased to 1.1× ULN (range 0.5–2.4), and 11 patients (46%) showed ALT normalization (Tables 2 and 3). Paired liver biopsies were available from 23 patients. Histological analysis revealed that median inflammation score dropped by 1 point to 1 (range 0–2; the Desmet score), whereas the median fibrosis score remained stable at 1 (range 0–4), shown in Table 3. Nevertheless, fibrosis decreased in eight patients (35%), and inflammation regressed in 10 patients (43%). Altogether, histological improvement was observed in 13 of 23 patients (56%). None of the patients developed an ADV-associated resistance mutation.

Table 2. Summary of Effects of Peg-IFN and ADV in CH-B Patients After 48 Weeks of Treatment
Changes in Serological and Histological ParametersResults
Serum HBV DNA reduction (median)−4.9 log
Serum HBV DNA <100 copies/mL13/24 (54%)
Intracellular total HBV DNA reduction−2.2 log
Intrahepatic cccDNA reduction (median−2.4 log
Serum HBsAg reduction (median)−0.6 log
HBsAg seroconversion4/24 (17%)
HBeAg loss8/15 (53%)
HBeAg seroconversion5/15 (33%)
ALT improvement21/24 (84%)
ALT normalization11/24 (46%)
Inflammation score improved10/23 (43%)
Fibrosis score improved8/23 (35%)
Table 3. Correlation of All Serological, Virological, and Histological Parameters Measured for Each Patient
  No.GTALT (ULN)HBsAg (μg/mL)HBV-DNA (copies/mL)Total HBV DNA (copies/cell)cccDNA (copies/cell)HBsAg (% pos. cells)HBcAg (% pos. cells)Desmet
  • *

    Patients with HBsAg seroconversion.

  • ND, measurement not determined because of material shortage or biopsy not available; Δ, values under limit of detection (3 cccDNA copies /10,000 cells); GT, genotype.

  • Necroinflammatory (grading) and fibrosis (staging) scores are given according to Desmet′s system. Patients 22 and 24 were lost to follow up. Patient 3 denied second liver biopsy.

HBeAg PositiveHBeAg Seroconversion2D5.21.0350*1 × 108<1 × 102139.600.030.280Δ9502003120
9C1.90.8150*3 × 107<1 × 102120.760.043.970.01100501110
10D3.11.34311.22 × 1071 × 10360.280.470.370.4820200103111
17D15.10.721335 × 107<1 × 1021270.382.371.470Δ60254001212
20D6.21.51210.43 × 107<1 × 102684.060.202.790.046031502231
HBeAg Loss21G5.22.0740*6 × 107<1 × 10265.970Δ2.580Δ10502121
18G2.61.0260.172 × 1074 × 103700.542.140.641.05905090502222
4D1.21.4100.55 × 1069 × 1038.260.130.040Δ6015502222
HBeAg Persistency5A4.11.41521.24 × 1073 × 103234.331.0010.825.79152515151102
15D3.70.7988.61 × 1075 × 106225.501.603.820.836055051100
7D3.11.01681 × 1062 × 103149.460.971.870.7470254002110
8A8.50.810032.46 × 107<1 × 10253.000.500.940.4010020.01100
23D3.11.5999.53 × 108<1 × 1021090.201.1216.850Δ4030501100
25D3.12.4134872 × 1092 × 108913.8523.7054.020.22802040701112
26D2.61.263222 × 1083 × 1042153.010.06195.920Δ70302011000
 HBeAg Negative16A1.01.16.60*2 × 103<1 × 1021.751.000.210.2700002244
1D5.01.81616.63 × 1051 × 10211.590.371.820Δ4015003233
3A4. × 104<1 × 1025.22ND0.05ND10ND0ND1ND1ND
6A1.01.1171.84 × 103<1 × 1021.020.540.060Δ4020002133
11D3. × 104<1 × 10218.370.370.360Δ201002210
12A1. × 104<1 × 10221.810.200.180.551ND3ND2233
13D1. × 103<1 × 1023.750.070.210Δ150001011
14A2. × 1059 × 102104.880.421.160.7730252002110
19A2.61.9153.34 × 1055 × 10228.720.570.040Δ55002132
22ND1.6ND12ND8 × 103ND15.01NDNDND20ND0ND1ND1ND

Levels of Intracellular Total HBV DNA and cccDNA Before Therapy.

As shown in Fig. 1 and Table 3, our analyses revealed that before therapy, HBeAg-positive CH-B patients had significantly higher median intrahepatic total HBV-DNA (225 copies/cell; range 8.26–2,153) compared with that of HBeAg-negative patients (13.3 copies/cell; range 1–104.8; P < .001; Fig. 1A). Median intrahepatic cccDNA level in participating patients was 1.05 copies/cell (range 0.036–195). Similar to total intrahepatic HBV-DNA, median intracellular cccDNA level was significantly higher in HBeAg-positive patients (2.58 copies/cell; range 0.036–195) than in HBeAg-negative patients (0.2 copies/cell; range 0.039–1.82; P < .01), as shown in Fig. 1B. Interestingly, as shown in Fig. 1C, in the 8 of 15 HBeAg-positive patients who became HBeAg-negative or developed anti-HBe at the end of 48 weeks of combination therapy, pretreatment cccDNA levels (1.05 copies/cell; range 0.036–3.97) were significantly lower than in patients without HBeAg loss or seroconversion (10.8 copies/cell; range 0.94–195; P < .05), although differences in intrahepatic total HBV DNA between these two groups were not significant (130 and 234.3 copies/cell, respectively).

Figure 1.

Comparison of baseline levels of intracellular total HBV DNA and cccDNA between HBeAg-positive, HBeAg-negative patients, and patients who lost HBeAg during therapy. (A) Intrahepatic total HBV DNA per cell at baseline in HBeAg-positive and -negative patients. (B) Intrahepatic cccDNA per cell at baseline in HBeAg-positive and -negative patients. (C) Pretreatment cccDNA levels of patients with HBeAg loss versus HBeAg persistency after 48 weeks of combination therapy. Each point represents a single patient, with bars indicating the median values.

Our analysis revealed that cccDNA levels at baseline were not significantly predictive for HBsAg seroconversion or for virological or biochemical response with ALT normalization.

Effect of 48 Weeks of Combination Therapy on Intracellular HBV DNA and cccDNA.

After 48 weeks of combination therapy, median intracellular total HBV-DNA was 0.47 copies/cell (range <0.0002–23.7 copies/cell), which corresponded to a 2.2 log10 decrease (P < .0001), shown in Fig. 2, whereas median copy number of cccDNA per cell was 0.004 (range <0.0003–5.79 copies/cell). This equates to a −2.4 log10 change (P < .01) in cccDNA levels achieved after 48 weeks of therapy (Fig. 2). Because of the marked cccDNA reduction induced by combination therapy, cccDNA became undetectable (3 copies/10,000 cells) in 11 patients (9 of 11 patients were HBeAg-negative at week 48; Table 3). On the other hand, total intracellular HBV-DNA was still detectable in 22 of 23 patients analyzed after treatment. Our data indicate that oral ADV therapy in combination with the immune modulator peg-IFN not only induced inhibition of HBV replication, but also the loss of a comparable magnitude of cccDNA.

Figure 2.

Effect of 48 weeks of combination therapy with peg-IFN and ADV on levels of total intrahepatic HBV DNA and cccDNA. Each point represents a single patient, with bars indicating the median values. Changes in individual patients are shown by connecting lines.

Levels of total intrahepatic HBV-DNA were clearly lower in patients with evidence of HBsAg seroconversion than in patients with persistent HBsAg (median 0.03 and 0.5 copies/cell, respectively). Although the number of patients who were HBsAg seroconverters was small (4 of 24 patients), this difference was statistically significant (P = .001). Similar differences cccDNA level were also observed between HBsAg-positive and anti-HBsAg+ patients (median 0.04 and 0.002 copies/cell, respectively).

Changes in HBsAg Levels.

Before therapy, all 26 patients were HBsAg positive, with levels ranging from 0.5 to 134 μg/mL and a median of 16.1 μg/mL (Table 3). HBsAg titers were measured every 4 weeks. Even if some fluctuations occurred in surface antigen concentrations measured during the course of treatment, patients had a median 72% reduction in serum HBsAg, with a median change of −0.55 log10 at the end of therapy. After 48 weeks of treatment, the median HBsAg titer was 4.5 μg/mL (range 0–87 μg/mL). The drop in serum HBsAg was highly significant (P < .001). As shown in Fig. 3, the HBsAg decline was clearly evident in 19 patients (trend lines), whereas values remained relatively stable or worsened slightly in five patients (patients 1, 5, 7, 17, and 20) until the end of therapy. HBsAg was under the limit of detection (Architect assay) for all four patients with HBsAg seroconversion (patients 2, 9, 16, and 21) at the end of therapy.

Figure 3.

Changes in HBsAg levels analyzed in patient serum samples obtained during the 48 weeks of therapy. Each line represents the trend line of quantitatively measured HBsAg for each individual patient (n = 24) over the 48 weeks of antiviral therapy. Serum samples were taken every 4 weeks. Quantitative measurement of HBsAg was performed using the Laurell method and Architect system for serum samples below 1 μg/mL HBsAg. Shown are trend lines (A) from patients with all HBsAg measurements remaining lower than 30 μg/mL and (B) from patients with at least one quantitative measurement found above 30 μg/mL.

Before the commencement of treatment, serum HBsAg in the eight HBeAg-positive patients who either lost HBeAg or seroconverted to anti-HBe was not significantly lower compared to those patients who remained HBeAg positive after 48 weeks (P = .13). Furthermore, we did not observe a significant difference in serum baseline HBsAg levels between patients who developed a virological response (n = 13) and patients who did not (n = 11) at the end of treatment (P = .66).

As shown in Fig. 4, we observed a significant correlation (P < .001; r = 0.625) between the concentration of HBsAg found in serum and the amount of total intracellular HBV-DNA in patients, both at baseline and at the end of therapy. A comparable tendency was also found regarding cccDNA concentration and level of serum HBsAg at baseline (P = .006; r = 0.547). However, this correlation was not statistically significant at the end of treatment (P = .6), possibly because cccDNA levels were below the lower limit of detection in 11 of 23 patients.

Figure 4.

Correlation between levels of total intrahepatic HBV-DNA and serum HBsAg titers taken at the time of biopsy. Each point represents a single patient. Reported are the data obtained for each patient both at baseline (n = 26) and at the end of treatment (n = 22*). The bar indicates the regression line. *Correlation was not performed in one patient because of undetectable HBV DNA levels at 48 weeks.

Immunohistology of Liver Biopsies for HBV Antigens.

Immunohistochemical staining was performed for each liver specimen before and after 48 weeks of combination treatment to determine the number of HBsAg- and HBcAg-positive hepatocytes (Fig. 5). As reported in Table 3, a mean of 36% of the hepatocytes stained positive for HBsAg in the baseline analysis of the liver biopsies (range 0–95). In liver samples analyzed at week 48, 14% of the hepatocytes were HBsAg positive, indicating a mean 2.5-fold reduction. Only a minority of the hepatocytes (16%) stained HBcAg positive before antiviral therapy, and the number of HBcAg-positive cells dropped to a mean of 7% at week 48, which corresponded to a 2.3-fold reduction.

Figure 5.

Reduction of the overall number of hepatocytes with detectable levels of HBcAg and HBsAg in a representative liver biopsy obtained from one patient before and after 48 weeks of antiviral therapy. Immunohistological staining with specific HBsAg or HBcAg antiserum was performed on formalin-fixed paraffin-embedded sections. Arrows indicate areas of HBcAg-positive staining.


Chronic HBV infection is often associated with the development of severe complications, such as cirrhosis and hepatocellular carcinoma (HCC), and currently available treatments are suboptimal because of the high relapse rate after cessation of treatment. At present, two main groups of therapeutic agents have been approved: immune modulator interferon alpha and three HBV polymerase inhibitors, lamivudine, adefovir dipivoxil, and entecavir (U.S.). Although interferon acts by stimulating cytokine production to control viral replication and promotes hepatocyte clearance by inducing cytotoxic T-cell activity, nucleos(t)ide analogues suppress viral replication at the level of DNA synthesis. However, interferon therapy inhibited HBV replication in approximately one-third of the CH-B patients, and prolonged therapy with nucleos(t)ide analogues was likely to lead to the selection of drug-resistant mutations in most patients.25, 26 These limitations underscore the need for alternative therapeutic approaches.

In this study, we explored the strength in antiviral and clinical effects of peg-IFN when used in combination with ADV. The major aim of this study was to analyze changes in intrahepatic total HBV-DNA and cccDNA levels in a prospective manner during 48 weeks of combination therapy in CH-B patients. Intrahepatic HBV-DNA level represent the active replicative form of the virus, and therefore, the amount of liver HBV-DNA may represent a better predictive factor for sustained virological response compared with more standard evaluations of treatment response, such as serum parameters and histology.

Recent studies have shown that suppression of HBV-DNA in serum does not indicate clearance of the virus in hepatocytes,18, 27 and the long half-life of cccDNA in infected hepatocytes is responsible for viral reactivation following drug cessation.4, 28 During antiviral therapy with ADV monotherapy, cccDNA could be reduced by nearly 1 log10 and intracellular total HBV DNA by 1.6 log10.8 However, the overall number of hepatocytes positive for hepatitis B antigens remained stable in that study, suggesting that in the setting of nucleotide monotherapy cccDNA reduction was mainly driven by the suppression of HBV DNA synthesis, which depleted cytoplasmic nucleocapsids available to replenish the nuclear cccDNA.29 Cytokines and noncytolytic TH-1 immune responses may also have been involved.30

In this study, cccDNA reduction was stronger and decreased by 2.4 log10, whereas total intrahepatic HBV-DNA decreased by 2.2 log10 after 48 weeks of combination treatment. Of note, combination treatment with peg-IFN and ADV strongly reduced the number of HBsAg- and HBcAg-positive hepatocytes, by 2.5- and 2.3-fold, respectively. Data available from animal models and clinical trials indicate that the rates of total intracellular HBV-DNA and cccDNA loss differ significantly under therapy with nucleos(t)ide analogues, the latter being much slower.4, 8 However, when immune-mediated mechanisms contribute to the clearance or destruction of infected cells, the magnitude of loss in both types of intrahepatic HBV-DNA might actually become similar.31 The reduction of the number of infected cells observed in this study suggests a clearance mechanism involving cytopathic effects. Although cytolytic mechanisms may have played an important role in the clearance of infected cells in the months preceding the second biopsy, the histological analysis performed on paired liver biopsies indicated that reduction of infected cells at the end of therapy was not accompanied by an increased necroinflammatory score (mHAI and Desmet). On the contrary, a general histological improvement could be observed in most patients.

Another contributor to the greater cccDNA reduction observed in this study could be a generally stronger antiviral response triggered by combination therapy, although this cannot be ascertained completely without control groups of patients receiving peg-IFN alone and ADV monotherapy.

Strong suppression of HBV-DNA levels to less than 100 copies/mL was achieved (median reduction of 4.9 log10) in 54% (13 of 24) of the treated patients. Although the number of patients enrolled in this study was small and the potency of such a combination therapy, in terms of serological outcome needs to be validated by a larger sample size, 48 weeks of combination therapy led to a high number of serological responders, as 53% (8 of 15) of the initially HBeAg-positive patients lost HBeAg and 33% (5 of 15) developed anti-HBe antibodies during treatment. The rates of HBeAg loss and seroconversion observed in this study were similar to those found in studies involving a larger number of CH-B HBeAg-positive individuals treated with peg-IFN alone or in combination with lamivudine.32 Analyses of biochemical parameters also revealed that 46% of patients reached a biochemical response with normal ALT values after 48 weeks of therapy. Similar rates of patients with ALT normalization have been reported in studies involving patients treated with IFN alone or in combination with nucleoside analogues.32–34 HBsAg seroconversion occurred in 4 of 26 patients. This number of HBsAg seroconverters is relatively high compared to earlier observations with therapy with peg-IFN alone or in combination with lamivudine, where 16 of 542 patients showed HBsAg seroconversion by the end of treatment,32 but it could be overestimated because of the smaller number of patients. Our analysis revealed that median baseline cccDNA levels were significantly lower (P < .01) in patients developing a serological response than in patients who did not show HBeAg loss or seroconversion. It will be important to investigate in a larger sample size whether a lower copy number of cccDNA is predictive of seroconversion.

A major obstacle to the study of cccDNA remains the requirement for repeated liver biopsy specimens from patients. Furthermore, a certain variability in cccDNA levels in liver biopsy samples should be taken into account. The distribution of cccDNA copy number may diverge among hepatocytes and in different parts of the liver. Therefore, it would be important to investigate whether cccDNA levels could be correlated with noninvasive serum surrogate parameters. Consistent with previous observations,8 we found a median 0.55 log10 reduction in HBsAg levels after 48 weeks of therapy. Interestingly, a significant correlation (P < .001) was found between HBsAg titers and amount of intrahepatic viral DNA both at baseline and at the end of treatment. Although a similar relationship was also found for cccDNA levels before therapy, this correlation was not statistically significant at the end of treatment (P = .6). A reason for this discrepancy could be the high number of patients (11 of 23) with cccDNA below the detection limit at the end of treatment. Further studies are warranted to establish the strength of such a correlation between levels of serum HBsAg and intrahepatic HBV-DNA, which could be used as a surrogate parameter to monitor the progressive clearance of cccDNA in the liver, and hence, efficacy of antiviral therapy. Interestingly, we observed in 3 patients (patients 17, 5, and 25), despite a reduction in cccDNA levels, a worsening in liver histology and an increase or maintaining of HBsAg levels. This observation raises the question of whether serial monitoring of HBsAg levels in the serum in the future might act as an additional marker for evaluating the response in patients during antiviral therapy.

In summary, our study revealed that 48 weeks of combination therapy with peg-IFN alpha-2b and ADV effectively diminished the amount of intrahepatic cccDNA. The reduction of HBV-DNA in the liver was accompanied by a significant loss of HBcAg- and HBsAg-positive hepatocytes and by a decrease in HBsAg levels in serum.