Dr G. A. Niro, Division of Gastroenterology, Casa Sollievo della Sofferenza Hospital, Viale Cappuccini, 71013 San Giovanni Rotondo (FG), Italy. E-mail: email@example.com
Background: Delta virus (HDV)-related chronic hepatitis is difficult to treat.
Aims: To evaluate the efficacy of lamivudine 100 mg daily on serum HDV-RNA, hepatitis D virus antibodies and alanine aminotransferase levels, liver histology, and on hepatitis B surface antigen seroconversion.
Methods: Thirty-one hepatitis B surface antigen-positive, HDV-RNA-positive patients with ALT ≥ 1.5 upper normal level and compensated liver disease were randomized (1:2 ratio) to placebo (group A, n = 11) or lamivudine (group B, n = 20) for 52 weeks; thereafter, all patients were given lamivudine for 52 weeks and followed up for 16 weeks.
Results: Twenty-five patients (81%) completed the study. No patient was HDV-RNA-negative at week 52; three patients (11%) were negative at week 104. Two of them remained HDV-RNA-negative at week 120, and one lost the hepatitis B surface antigen without seroconversion. Paired pre-treatment and week 104 liver biopsies were available from 19 patients: of which three of seven (43%) from group A and two of 12 patients (17%) from group B had a ≥2 point decrease in the Ishak necroinflammatory score.
Conclusion: A sustained complete response was achieved in 8% of hepatitis D virus-infected patients treated with lamivudine and a partial histological response in 26% of them. Hepatitis D virus viraemia was unaffected, even in patients when hepatitis B virus replication was lowered by lamivudine therapy.
Hepatitis D virus (HDV) infection is associated with severe liver disease that progresses to cirrhosis and often requires liver transplantation.1, 2 Although the incidence of HDV infection has much diminished in the industrialized world,3 there is still a sizeable number of chronic HDV carriers with active liver disease who might benefit from anti-viral treatment.
Currently, there is no effective therapy for chronic delta hepatitis. About 40–70% of these patients treated with interferon (IFN) experience a biochemical response during therapy at a rate proportional to the dose, 9 million units (MU) three times a week being more effective than 3 MU thrice weekly.4 However, although the biochemical response is usually accompanied by a transient virological response, there is a high rate (60–70%) of post-treatment relapse.5, 6 Aside from IFN, other therapeutic approaches, such as acyclovir, ribavirin and thymosin, have been unhelpful.7–9
Despite evidence that hepatitis B virus (HBV) replicative activity usually is inhibited by HDV and HBV-DNA can be detected in serum of most hepatitis D patients only with sensitive assays,10 HBV is necessary to the thriving of HDV by providing it with complementary biological functions.11 On these premises in the present study the efficacy of lamivudine has been evaluated in patients with chronic delta infection in order to establish whether treatment can achieve further HBV suppression and thus remove biological support to HDV and ameliorate hepatitis D.
Patients and methods
In a multicentre, double-blind, randomized study, 31 patients with chronic delta hepatitis, referred to three Medical Centers (Torino and San Giovanni Rotondo in Italy, and Hannover in Germany) were enrolled. Inclusion criteria were the presence of anti-HDV and hepatitis B surface antigen (HBsAg) in serum for at least 3 months prior to screening, detectable HDV-RNA at enrolment, and alanine aminotransferase (ALT) values ≥1.5 and ≤10 times the normal value. Patients were excluded if they were HCV- or HIV-positive, or had received anti-viral, cytotoxic, corticosteroid or immunomodulatory treatment within 6 months prior to inclusion. Hospital ethics committees approved the study; written informed consent was obtained from all patients prior to participation and the study protocol was conforming to the ethical guidelines.
Patients were randomly assigned by computer-generated code on a ratio of 2:1 to receive either 100 mg lamivudine monotherapy or a matching oral placebo once daily. The study was double-blind for the first 52 weeks of treatment. Thereafter, all patients received on an open-label basis lamivudine for a further 52-week period and were followed up for 16 weeks off therapy.
Laboratory and virological testing
Assessments for efficacy and safety were carried out at baseline, at weeks 4 and 8, and thereafter every 8 weeks until week 104. Off-treatment follow-up assessments were carried out every 4 weeks for 16 weeks.
At each clinical visit routine laboratory tests included serum ALT and aspartate aminotransferase (AST), serum direct and total bilirubin, albumin, and complete blood counts. Serological markers of HBV, HDV, HCV and HIV were tested by commercial enzyme immunoassays (Abbott Laboratories, North Chicago, IL, USA; Sorin Biomedica, Saluggia, Italy; Ortho Diagnostic Systems, Raritan, NJ, USA). Samples were tested for HDV-RNA, as previously described,10 by a semiquantitative single and nested polymerase chain reaction (PCR) assay, with a sensitivity of approximately 1000 genomes for single PCR, and 1–10 genomes for nested-PCR. Serum HBV-DNA levels were determined by a commercial quantitative PCR assay (Amplicor HBV Monitor Test, Roche Diagnostics, GmbH Mannheim, Germany) with a sensitivity threshold of about 1000 copies/mL. In addition, at week 104, in viraemic patients, resistance was evaluated by means of line probe assay (INNO LiPA HBV-DR, Innogenetics, Ghent, Belgium).
A liver biopsy was performed within 24 months before screening, and at the end of the second course of treatment. A single pathologist, blinded to study treatment, assessed all biopsies. The presence of intrahepatic HDAg, HBsAg and HBcAg was evaluated by immunohistochemistry; changes in histological activity were evaluated by the modified Ishak score.12
The primary efficacy end-point was the ability of lamivudine monotherapy to influence eradication of serum HDV-RNA and induce loss of HBsAg with seroconversion to anti-HBs. Secondary efficacy objectives included both the biochemical response, defined as the normalization of ALT, and the histological response, defined as a reduction of ≥2 points in the modified Ishak index. Necroinflammatory activity and fibrosis were also evaluated and scored within each group of patients.
Data were analysed using the spss statistical package (version 6.1.3; SPSS Inc., Chicago, IL, USA). Continuous variables were expressed as mean values and s.d. Categorical variables were reported as percentage values. Baseline features between the two groups of treatment were compared by means of Student's t-test for continuous variables or chi-squared test (or Fisher's exact test, when needed) for categorical variables. Histological scores of liver biopsies were considered as categorical variables and expressed as median values with interquartile range (75th percentile to 25th percentile). Mann–Whitney U-test was used to compare the pre-treatment scores between the two groups of patients. Wilcoxon matched pairs signed ranks test was used to evaluate the differences in histological scores within each group of patients. P-value of <0.05 was considered statistically significant.
A total of 31 patients were randomized to treatment, 11 to placebo and 20 to lamivudine. Mean age at enrolment was 42.6 years (range: 24–58); 24 patients (77%) were males. Route of infection was unknown in 71% of cases; six patients (19%) were transfusion recipients, one patient (3%) recalled a past history of intravenous drug abuse, one was vertical/perinatal, and the last one reported intrafamilial HDV diffusion. Serum immunoglobulin M (IgM) anti-HDV tested positive in 28 patients (90%), and anti-HBe antibodies were detectable in 29 patients (93%). High levels of HDV-RNA were found, at screening, in 29 of 31 patients who tested positive by single PCR amplification. Twelve patients (39%) were positive for HBV-DNA, with individual levels ranging widely from 1027 to 13.9 × 106 copies/mL.
At screening, mean histological scores for group A were 6 ± 2.4 and 4.1 ± 2.1 for grading and staging, respectively. For group B, the mean histological scores were 6.2 ± 3.2 and 3.1 ± 0.9 for grading and staging respectively. At immunoperoxidase evaluation of baseline biopsy, HDV antigen was detected in 15% of cells (range: 0.1–30) in those patients treated initially with placebo and thereafter with lamivudine (group A), and in 8% (range: 0.1–30) in patients who received a 2-year course of lamivudine (group B), the HBsAg in 5% (range: 0.1–10) and 10% (range: 0.1–30), respectively, and the HBcAg was negative in all but one patient from group B (0.1% of cells). The two treatment groups, sorted out according to lamivudine treatment duration, were well matched for demographic, biochemical and virological parameters, and scoring of pre-treatment biopsies (Table 1). Twenty-nine patients (94%) completed the randomized phase of the trial, 28 (90%) patients completed the second course of open-label lamivudine administration and 25 (81%) the follow-up period.
Table 1. Demographic data of 31 patients with chronic hepatitis D (anti-HDV-positive), divided in the two treatment groups
Patients received placebo (group A) or lamivudine (group B) for 52 weeks, thereafter all patients were given lamivudine for further 52 weeks.
ALT, alanine aminotransferase; HDV, hepatitis D virus; IgM, immunoglobulin M; PCR, polymerase chain reaction.
Number of patients
Age (years; mean ± s.d.)
41.7 ± 8.6
43.1 ± 9.5
Gender [male; n (%)]
Transfusion recipient, n (%)
Duration of HDV (years; mean ± s.d.)
3.8 ± 3.6
7.4 ± 5.8
Previous anti-viral therapy, n (%)
ALT (×ULN; mean ± s.d.)
3.5 ± 2.9
3.2 ± 1.9
IgM anti-HDV-positive, n (%)
HDV-RNA-positive I PCR, n (%)
HBV-DNA [>1000 copies/mL; n (%)]
Grading score, median (Q3–Q1)
Fibrosis score, median (Q3–Q1)
All patients tested positive for HDV-RNA at week 52. At week 104 HDV viraemia was no longer detectable in three patients (11%), one from group A and two from group B. At week 120 (after 16 weeks off-treatment) HDV-RNA returned detectable in the patient from group A, while in the two patients from group B a sustained response was maintained. Clinical and virological outcome of responders are reported in Figures 1 and 2.
Nineteen paired biopsies were available from seven group A patients and 12 group B patients. In three patients from group A (43%) and in two from group B (17%) a ≥2 point decrease in the Ishak necroinflammatory score was registered; in the remaining 14 non-responders the score either worsened or did not change (Table 2). Fibrosis score did not change in 74% of patients (six patients from group A and eight from group B).
Table 2. Histological changes in necroinflammatory activity, fibrosis score and immunohistochemistry from baseline to week 104
Group A, n (%)
Group B, n (%)
a Unavailable from two patients, one in each treatment group.
HBsAg, hepatitis B surface antigen.
Number of subjects with pre- and post-treatment biopsy
Better than baseline
No change from baseline
Worse than baseline
Better than baseline
No change from baseline
Worse than baseline
Immunohistochemistrya (HDAg and HBsAg)
Better than baseline
No change from baseline
Worse than baseline
In both group of patients a progressive decline of mean ALT levels was observed until week 52 (from a value of 143 ± 118 to 98 ± 53 U/L in group A, and from 131 ± 78 to 86 ± 41 U/L in group B); thereafter ALT fluctuated slightly over the baseline value. At the end of follow-up (week 120) biochemical sustained response was achieved in four patients (16%), two in each treatment group.
Of the 31 HDV-positive patients, 12 had a concomitant HBV viraemia: five of them were treated with lamivudine for 52 weeks, and the remaining seven patients for 104 weeks and HBV-DNA levels became undetectable in five of them (42%); this change did not result in normal ALT levels or amelioration of liver histology. The seven HBV-DNA-positive patients at the end of therapy had variable HBV viraemia values ranging from low levels (<2500 copies/mL in four patients) to 105–106 copies/mL (in three patients). No drug-resistant HBV variant was identified by the INNO LiPA HBV-DR assay in amplified HBV polymerase region.
No difference in the viral markers was reported between treatment groups, at exception of a single patient who became HBsAg-negative (Figure 2): a 36-year-old man with a past history of drug abuse. HDV-RNA tested positive at week 52 following the first year of lamivudine therapy, but became undetectable at week 104. This patient was shown to lose serum HBsAg at week 120, without a concomitant development of anti-HBs.
The incidence of adverse events on therapy was similar in the treatment groups. Two drug-related events were reported, both in subjects from group A. One of these events was recorded during the initial 52 weeks phase: serum amylase increased in one patient in absence of clinical and ultrasound signs of pancreatitis. The other event (migraine) was recorded both during the randomized and the open-label lamivudine phase of treatment. There were no adverse events that led to premature discontinuation of investigational products or withdrawal of patients from the study.
The spectrum of HDV liver disorders varies from symptomless infection without biochemical signs of liver damage to fulminant hepatitis and disease conductive to cirrhosis.
In HDV-infected patients, a concomitant productive HBV infection has been identified as a major factor promoting liver damage. It was shown that patients with chronic HDV infection and high levels of HBV-DNA in blood have more severe disease than those with low viraemia;13 likewise, in the time course of HBV/HDV reinfection in liver transplants, HDV became detectable earlier than HBV but the onset of recurrent hepatitis D invariably followed the reactivation of HBV.10
Therefore, in this study we treated HDV-infected patients with lamivudine, on the premise that, by inhibiting the synthesis of the helper HBV the drug might control HDV liver damage.
The results were nevertheless disappointing. Hepatitis delta virus RNA became negative in serum only in two treated patients (8%) at week 120 and ALT normalized only in four patients (16%). All the patients but one remained HBsAg-positive at the end of the follow-up. In the single patient who lost the HBsAg, HDV-RNA became negative after 24 months of treatment and viraemia remained undetectable throughout the follow-up.
However, in our patient clearance of HBsAg may have been spontaneous rather than therapy induced. The interaction between HBV and HDV leading to the clearance of the HBsAg has been widely reported by many investigators.14–16 In a study,17 up to 10% of HBsAg carriers originally coinfected with HDV, lost serum markers of HDV and became HBsAg-negative during 1–6 years of follow-up.
Lamivudine was nevertheless effective in reducing serum HBV-DNA levels. HBV-DNA, detectable at baseline in 12 of 31 patients (39%) was detectable in seven of 28 patients (25%) at the end of the study. Similar results were reported by Lau et al.18 in five patients treated with lamivudine monotherapy; despite a rapid fall of serum HBV-DNA, which became undetectable even by polymerase reaction in four, all patients remained HBsAg- and HDV-RNA-positive with no improvement of ALT and liver histology. The limited benefit for HDV-infected patients despite the regression of serum HBV is likely due to the lack of interference of lamivudine with the synthesis of intrahepatic cccDNA which regulates protein production, including the synthesis of the HBsAg.19 Thus, unless the HBsAg is also cleared, inhibition of HBV, however profound, is unlikely to lead to eradication of HDV. This conclusion is consistent with the knowledge that HDV does not need HBV for replicating but critically needs the HBsAg to assembly into virions and propagates its infection to hepatocytes.20, 21 In contrast to the patients treated by Lau et al.,18 we observed a ≥2 point decrease in necroinflammation in five and no histological change from baseline in seven of 19 patients treated for 1–2 years. The significance of the amelioration or stability of necroinflammation despite lack of a virological response is unclear; these findings are similar to the improvement of liver necroinflammation in lamivudine-treated HBV patients, whose serum HBV titre was not modified by therapy.
In conclusion, lamivudine was effective in inhibiting or lowering serum HBV replication but this virological achievement did not result in amelioration of HDV infection, presumably because therapy had no effect on intracellular HBV cccDNA and on production of the HBsAg and thus did not influence either HDV propagation and replication. Future trials with anti-viral agents capable to block intrahepatic HBV cccDNA22 and production of HBsAg or to interfere selectively with the life cycle of HDV23 might be more efficacious.
The present trial was supported by Glaxo Smithkline that provided both lamivudine and placebo.