About 0.8 million people die because of hepatitis B virus (HBV) infection each year. In around 5% of infected adults, the immune system is ineffective in countering HBV infection, leading to chronic hepatitis B (CHB). CHB is associated with hepatocellular carcinoma, which can lead to patient death. Unfortunately, although current treatments against CHB allow control of HBV infection, they are unable to achieve complete eradication of the virus. Cytokines of the IFN family represent part of the innate immune system and are key players in virus elimination. IFN secretion induces the expression of interferon stimulated genes, producing proteins that have antiviral properties and that are essential to cell-autonomous immunity. IFN-α is commonly used as a therapeutic approach for CHB. In addition, IFN-γ has been identified as the main IFN family member responsible for HBV eradication during acute infection. In this review, we summarize the key evidence gained from cellular or animal models of HBV replication or infection concerning the potential anti-HBV roles of IFN-γ with a particular focus on some IFN-γ-inducible genes.

Interferon-gamma-1b (IFN-γ-1b) improves alpha interferon (IFN-α) inhibition of hepatitis C virus (HCV) replication in replicon system. We described virological response after addition of IFN-γ to a combination of ribavirin/peginterferon (PEG-IFN)-α-2a or α-2b.

In this non-comparative, multicenter trial, patients chronically infected by HCV who were nonresponders to a previous treatment by PEG-IFN and ribavirin were restarted on a regimen of PEG-IFN-α-2a (180 μg/week) + ribavirin (1000-1200 mg/day) for 16 weeks. If HCV-RNA decreased less than 2 log(10) copies/mL (nonresponders), and if PEG-IFN-α-2a and ribavirin dosages were unchanged while tolerance was good, IFNγ-1b (100 μg three times per week) was added for the last 32 weeks of treatment. Virological response was evaluated at week 28 (12 weeks after initiation of IFN-γ-1b).

Among the 48 patients started on dual therapy, 23 patients (47%) were nonresponders at week 12 and received IFN-γ-1b from week 16 onward. Their mean HCV-RNA (log(10) IU/mL) was 6.83 at baseline, 5.81 at week 12, and 5.63 at week 28. No patient reached undetectable HCV-RNA at week 28 (upper bound of 95% confidence interval: 14.8%); none had a decrease > 1 log(10) IU/mL. One case of grade 4 neutropenia was reported.

Among the strictly selected nonresponders, IFN-γ-1b (at a dosage of 100 μg thrice a week) in combination with PEG-IFN-α-2a and ribavirin failed to show virological efficacy.

Treatment of HBV chronic carriers using interferon (IFN)-α or nucleoside/nucleotide analogues fails to suppress viral infection. Type-II IFN-γ has been shown to inhibit HBV replication. The goal of the present work was to evaluate the antiviral efficacy against HBV of a thermostable IFN-γ variant isolated using Massive Mutagenesis and thermoresistant selection (THR) technologies.

The thermostability of wild-type (wt) and S63C IFN-γ was determined in vitro and in vivo. Activation of the IFN-γ responsive element by wt and S63C IFN-γ was tested using a luciferase assay. HepG2.2.15 cells constitutively expressing HBV were used to analyse the antiviral activity of wt and S63C IFN-γ against HBV replication. Intracellular HBV DNA was detected by Southern blot and quantified by real-time PCR analyses.

S63C IFN-γ was shown to be more thermostable and had a longer half-life than wt IFN-γ. Both wt and S63C IFN-γ displayed a similar capacity to activate the IFN pathway. The treatment of HepG2.2.15 cells with wt or S63C IFN-γ induced the inhibition of HBV viral replication. After heating, S63C IFN-γ displayed better conservation of its antiviral activity against HBV when compared with wt IFN-γ.

These results confirm that the THR method can be used to isolate mutants with enhanced thermostability and demonstrate that a thermostable IFN-γ variant presents antiviral properties against HBV replication. This molecule could provide a new strategy to treat patients who do not respond to antiviral therapy.

Interferon (IFN) plays a central role in the innate and adaptive antiviral immune responses. While IFN-alpha is currently approved for treating chronic hepatitis B and hepatitis C, in limited studies, IFN-gamma has not been shown to be effective for chronic hepatitis B or C. To identify the potential mechanism underlying the differential antiviral effects of IFN-alpha and IFN-gamma, we used cDNA microarray to profile the global transcriptional response to IFN-alpha and IFN-gamma in primary human hepatocytes, the target cell population of hepatitis viruses. Our results reveal distinct patterns of gene expression induced by these 2 cytokines. Overall, IFN-alpha induces more genes than IFN-gamma at the transcriptional level. Distinct sets of genes were induced by IFN-alpha and IFN-gamma with limited overlaps. IFN-alpha induces gene transcription at an early time point (6 h) but not at a later time point (18 h), while the effects of IFN-gamma are more prominent at 18 h than at 6 h, suggesting a delayed transcriptional response to IFN-gamma in the hepatocytes. These findings indicate differential actions of IFN-alpha and IFN-gamma in the context of therapeutic intervention for chronic viral infections in the liver.

To evaluate the in vitro anti-HBV activity of recombinant human IFN-gamma, alone and in combination with lamivudine.

A recombinant baculovirus-HBV/HepG2 culture system was developed which could support productive HBV infection in vitro. Expression of HBsAg and HBeAg in infected HepG2 culture medium was detected by commercial enzyme immunoassays. HBV DNA replication intermediates were detected in infected cells by Southern hybridization and viral DNA load was determined by dot hybridization.

IFN-gamma at 0.1 to 5 microg/L efficiently down regulated HBsAg expression in transduced HepG2 cells. At 5 microg/L, IFN-gamma also suppressed HBV DNA replication in these cells. While treatment with a combination of lamivudine and IFN-gamma showed no additive effect, sequential treatment first with lamivudine and then IFN-gamma was found to be promising. In this culture system the best HBV suppression was observed with a pulse of 2 micromol/L lamivudine for two days, followed by 1 microg/L IFN-gamma for another four days. Compared to treatment with lamivudine alone, the sequential use of 0.2 micromol/L lamivudine for two days, followed by 5 microg/L IFN-gamma for six days showed a 72% reduction in HBV cccDNA pool.

This in vitro study warrants further evaluation of a combination of IFN-gamma and lamivudine, especially in IFN-alpha non-responder chronic hepatitis B patients. A reduced duration of lamivudine treatment would also restrict the emergence of drug-resistant HBV mutants.

Currently, there are no effective therapies available for patients with chronic hepatitis C who have failed to respond to optimal interferon alfa-based regimens. The aims of this pilot study were to assess the antiviral activity and safety of interferon gamma in chronic hepatitis C.

Patients with chronic hepatitis C, genotype 1, who had not responded to or who had relapsed after therapy with interferon alfa and ribavirin were enrolled in a trial of interferon gamma 1b given in doses of 100, 200 or 400 microg subcutaneously three times weekly for 4 weeks. Frequent blood samples were obtained for HCV RNA levels.

Fourteen patients were enrolled. Geometric mean HCV RNA levels remained unchanged. Serum aminotransferase levels also did not change, while there were significant decreases in neutrophil counts (-41% from baseline) and hematocrit (-5%). Low grade fever and malaise were common with the first injection of interferon gamma, but no serious side effects were encountered.

Although relatively well tolerated, interferon gamma in doses of 100-400 microg thrice weekly had no effect on HCV RNA levels in patients with chronic hepatitis C who had failed to achieve a sustained response to interferon alfa-based therapies.

Phospholipase A(2) plays a role in cholesterol gallstone formation by hydrolyzing bile phospholipids into lysolecithin and free fatty acids. This study investigated its effects on cholesterol crystallization in model bile systems. Supersaturated model bile solutions with different cholesterol saturation indexes (1.2, 1.4, and 1.6) were prepared using cholesterol, taurocholate, and egg yolk phosphatidylcholine, soybean phosphatidylcholine, palmitoyl-oleoyl phosphatidylcholine, or palmitoyl-linoleoyl phosphatidylcholine. Then the effect of digestion of phosphatidylcholine by phospholipase A(2) on bile metastability was assessed by spectrophotometry and video-enhanced differential contrast microscopy. Addition of phospholipase A(2) caused the release of free fatty acids in a time-dependent manner. Cholesterol crystallization was enhanced by an increased crystal growth rate in model bile containing hydrophilic species such as soybean or palmitoyl-linoleoyl phosphatidylcholine, consisting predominantly of polyunsaturated fatty acids. Because phospholipase A(2) enhanced cholesterol crystallization in bile containing hydrophilic phosphatidylcholine species, but not hydrophobic phosphatidylcholine species, release of polyunsaturated fatty acids by hydrolysis may be responsible for such enhancement. Therefore, the role of phospholipase A(2) in cholesterol gallstone formation depends on the phospholipid species present in bile, so that phospholipid species selection during hepatic excretion is, in part, crucial to the cholesterol stone formation.

Acute and chronic liver diseases related to hepatitis viruses are the main indications for liver transplantation. The risk of viral reinfection after transplantation is the main limiting factor in these indications. The risk of viral B reinfection is: 80% in the absence of prophylaxis; is related to the presence of active viral B replication prior to transplantation; is higher in patients with chronic liver disease, rather than with fulminant hepatitis; and is higher in patients with hepatitis B virus (HBV)-related liver disease alone rather than in those with HBV-hepatitis delta virus (HDV) infection. Post-transplant long-term passive antibody to hepatitis B (anti-HB) immunoprophylaxis reduces the risk of HBV recurrence to 30% in patients with HBV cirrhosis, and to less than 10% in those with fulminant hepatitis B. Patients with HBV-HDV liver disease receiving passive anti-HB immunoprophylaxis are at low risk of HBV recurrence (10-15%), but at high risk of HDV recurrence (80%). However, HDV reinfection of the graft has no clinicopathological consequence in the absence of concomitant HBV reinfection. The five year survival of patients transplanted for HBV cirrhosis and for HDV cirrhosis at the Hepatobiliary Center, Hôpital Paul Brousse is 72% and 85%, respectively. Hepatitis B virus reinfection of the graft is characterized by a high level of viral replication, and a chronic outcome. Antiviral treatments such as ganciclovir, adenine arabinoside monophosphate, famcyclovir, and lamivudine have a place after transplantation and may stop HBV replication, ganciclovir, famcyclovir and lamivudine should be continued for several months and in some cases indefinitely. Hepatitis C virus reinfection is almost constant, assessed by the persistence of hepatitis C virus (HCV)-RNA in the serum in 90% of cases. Acute lobular hepatitis appeared in 75% of patients at a median of 4 months post transplantation with a range of between 23 days and 4 years. In our series, the 5 year actuarial rate of HCV acute hepatitis on the graft, chronic hepatitis, and cirrhosis, is 75, 60, and 8%, respectively. Hepatitis C virus RNA level is dramatically increased after transplantation and seems to correlate with the occurrence of acute hepatitis on the graft. A positive relation between genotype 1b and prevalence and severity of HCV hepatitis on the graft have been suggested in European series. There is no demonstrated way to prevent HCV reinfection. The use of interferon for the treatment of HCV hepatitis on the graft was disappointing due to a poor antiviral effect and the occurrence of chronic rejection episodes in some patients. Promising results of the combination of interferon and ribavirine have been reported and need confirmation. The 5 year survival of patients transplanted for viral C cirrhosis at the Hepatobiliary Center, Hôpital Paul Brousse is 78%. In conclusion, patients with HBV cirrhosis and without HBV replication are candidates for liver transplantation. Long-term passive anti-HB prophylaxis is the best way to prevent HBV recurrence. Patients with HBV replication should be included in protocols using combinations of antiviral treatments and passive anti-HB immunoprophylaxis. Viral C reinfection is frequent, but medium-term survival is good. However, long-term graft and patient survival remains unknown and methods to prevent and treat HCV reinfection on the graft are needed.

Aggregation and fusion of non-micellar particulate species, such as unilamellar vesicle and phospholipid lamellae, are believed to precede the nucleation of cholesterol crystals in bile. However, little is known about the time sequence relationship between transformation of non-micellar particles and the initial appearance of cholesterol crystals, as no adequate technique is available for assessing such transformations quantitatively. We have developed a novel method for quantitatively estimating vesicle transformation in supersaturated model bile systems, using a spectrophotometric technique to determine the time sequence relationship between such transformations and cholesterol crystal nucleation. We also investigated the potency of a given effector substance on this transformation. This method permits simultaneous quantitative determination of vesicle aggregation and of cholesterol crystal growth. Maximal vesicular aggregation as determined from turbidity, coincided with initiation of cholesterol crystal nucleation. The addition of divalent cations, Ca2+ and Mg2+, to the model bile solutions promoted vesicle aggregation and cholesterol crystal nucleation and growth. In contrast, apolipoproteins A-1 and A-2 retarded such processes. These data were highly reproducible and reliable. The method described is easy to perform, provides reproducible results and permits the determination of the potency of effector substances on vesicle transformation and on the nucleation of cholesterol crystals.

Interferon-alpha is currently the only therapy approved for treatment of chronic HBV in the United States and Europe. Interferon-alpha therapy causes loss of HBeAg and HBV DNA in approximately a third of treated patients, and the loss of these markers of active viral replication is associated with improvements in hepatic histology and ALT levels. However, the long-term effects of interferon-alpha on morbidity and mortality, and especially on the incidence of the complications of chronic HBV infection, remain to be defined. The currently available treatment for chronic HBV is far from perfect. Interferon therapy is usually associated with significant side effects and requires subcutaneous administration. Additionally, there are a large number of patients who either fail to meet criteria for treatment, or who, with therapy, fail to respond (at least 60% of all patients). Moreover, interferon treatment is expensive (approximately $5,000 for a 16 week course of 5MU daily). Hence the search continues for effective, orally-available and cost-efficacious therapy. Of the agents available, the nucleoside analogues appear to have the greatest promise. The availability of cell culture systems and animal models for studying potential anti-HBV drugs will aid in the future development of these agents. Therapeutic vaccines, and combination therapies (given either concurrently or sequentially) may also play a future role in the management of chronic HBV infection. While prevention of disease must be a primary goal in the war against this common infection, a continued focus must be maintained on the treatment of the approximately 300 million individuals world-wide with established chronic HBV infection.

The effects of tumor necrosis factor-alpha and/or interferon-gamma on the replication of hepatitis B virus were examined using HB611 cells. These cells were derived from human hepatoblastoma cells, Huh6, by integrating hepatitis B virus DNA, and produce hepatitis B virus continuously. Each of the cytokines inhibited hepatitis B virus replication in the cells assessed as the amount of episomal hepatitis B virus DNA, without a decrease in cell viability. When the two cytokines were administered together, the inhibitory effect became greater. Incubation of the cells with 1,000 U/ml tumor necrosis factor-alpha decreased HBV DNA replicative intermediates by 55%, and that with 1,000 U/ml interferon-gamma decreased these by 51%. Furthermore, incubation with 1,000 U/ml tumor necrosis factor-alpha and 1,000 U/ml interferon-gamma in combination decreased HBV DNA replicative intermediates by 71%. In contrast, the amount of hepatitis B virus RNA and secretion of hepatitis B e antigen were not apparently reduced by the cytokines, and 2',5'-oligoadenylate synthetase activity was not detected in the supernatant. These results suggest that tumor necrosis factor-alpha and interferon-gamma inhibit hepatitis B virus replication by blocking some step in reverse transcription and that the 2',5'-oligoadenylate synthetase is not involved in the mechanism underlying the inhibition by these two cytokines.

The concanavalin A-bound glycoproteins in human gallbladder bile have recently been demonstrated to be strong promoters of cholesterol crystal nucleation. In the present study, we investigated the mechanism(s) whereby such promoters affect cholesterol crystal nucleation and/or growth, and compared these mechanisms with those of another promoter, calcium ion. Concanavalin A-bound glycoproteins were isolated from the Helix pomatia-unbound fraction of gallbladder bile from stone-free patients, and determined by electrohoresis to consist of six subclasses (MW 143, 98, 80, 58, 50, and 40 kDa). A cholesterol crystal growth assay showed that concanavalin A-bound glycoproteins both accelerated nucleation time and increased growth rate, whereas calcium ion affected nucleation time only. In the presence of both concanavalin A-bound glycoproteins and calcium ion, both cholesterol nucleation and growth were markedly enhanced. A gel permeation chromatographic study revealed that concanavalin A-bound glycoproteins shifted a considerable amount of cholesterol from micelles to vesicles, whereas calcium ion did not. These results suggest that concanavalin A-bound glycoproteins promote cholesterol crystal nucleation and growth, partly by shifting cholesterol from stable micelles to metastable nonmicellar fractions in bile. In contrast, calcium ion promotes these processes by other mechanisms and, therefore, enhances the effect of concanavalin A-bound glycoproteins.

Recent advances have been made in the treatment of chronic viral hepatitis, mainly with recombinant interferon (IFN) alpha. However, the present treatment of chronic viral hepatitis is not entirely satisfactory because the efficacy is inconstant and/or incomplete. In chronic hepatitis B IFN-alpha induces a sustained interruption of hepatitis B virus (HBV) replication, with a HBeAg to anti-HBe seroconversion in about 30% of patients. Patients most likely to respond are those with no immunosuppression, HBV infection acquired during adulthood or active liver disease with low HBV replication. Responders usually show a significant decrease in serum HBV DNA levels during the first 2 months of therapy, followed by a significant increase in the level of aminotransferases. New nucleoside analogues might be useful in combination with IFN-alpha in the treatment of those who do not respond to IFN therapy. In chronic hepatitis B-D, the rate of sustained response to IFN-alpha therapy is low. To be effective, IFN-alpha must be used at a high dosage (9-10 mega units) with a long duration (1 year). In chronic hepatitis C, IFN-alpha at a dosage of 3 mega units over 6 months, induces a sustained response in about 20% of patients. A higher dosage of IFN (5-10 mega units) and a longer duration of treatment increases the rate of sustained response but is associated with poor tolerance. Non-responders to a first course of IFN do not respond to a second course of treatment. In patients who respond but relapse after treatment, the rate of sustained response after a second course of IFN needs to be assessed. Ribavirin, which has a significant antiviral effect on hepatitis C virus, might be useful in combination with IFN-alpha. At the dosage (3-6 mega units) usually used, IFN-alpha is relatively well tolerated. In about 10% of the patients therapy is interrupted, mainly because of severe fatigue, thyroid dysfunction or depression.

All interferons display antiviral properties, but gamma-interferon especially has an immunomodulatory effect and may induce autoimmune phenomena. Therefore the formation of autoantibodies was investigated in patients with chronic hepatitis B treated with gamma-interferon. Eleven patients (all HBs-Ag and HBe-Ag positive) were treated for 6 months with recombinant gamma-interferon. The following antibodies were tested: anti-nuclear antibodies, smooth muscle antibodies, anti-actin, anti-mitochondrial antibodies of subgroup anti-M2 and anti-M9 as well as naturally occurring antibodies, antibodies to liver-kidney microsomes, vascular endothelial cell antibodies, sarcolemmal antibodies, parietal cell antibodies, thyroglobulin antibodies and antibodies to laminin and keratin. All patients produced autoantibodies during therapy. The maximum antibody formation and the highest titres were observed in the period between the 3rd and 6th month after therapy began. The cumulative frequencies of the different antibody specificities were as follows: n = 6 anti-nuclear antibodies, n = 7 smooth muscle antibodies, n = 1 anti-actin, n = 12 antibodies to laminin or keratin, n = 6 endothelial cell antibodies/sarcolemmal antibodies, n = 6 anti-mitochondrial antibodies, n = 1 antibodies to liver-kidney microsomes, n = 2 thyroglobulin antibodies, n = 4 parietal cell antibodies. Antibodies persisted in six patients over a period of 3 months (two cases of parietal cell antibodies and one case of antibodies to liver-kidney microsomes) and were still detectable in three patients 6 months after therapy. In three patients new antibody formation occurred 1 month after therapy. So far, clinical signs of an autoimmune disorder have not appeared in any of the patients.(ABSTRACT TRUNCATED AT 250 WORDS)

Interferon treatment of hepatitis B and C virus (HBV, HCV) infections has been hampered by overall initial response rates of < 50%, a relapse rate that is > 50% for patients with chronic HCV, and rare responses in individuals with chronic HBV who are immunosuppressed or immunologically tolerant to the HBV. Because of these difficulties, the efficacy of other therapeutic agents is being vigorously explored. Among the immunomodulatory agents being evaluated, thymosin appears to be a promising new therapy for HBV. Results from an ongoing multicenter trial evaluating thymosin are expected next year. A variety of nucleoside analogues with antiviral activity against the HBV have also been identified. Several of the more active agents deserve further study in clinical trials. In chronic HCV infection, only interferon therapy has been extensively studied. Ribavirin alone may have some value, but its precise role in the treatment of chronic HCV will require additional testing. Interferon therapy for patients with chronic HBV or HCV infection represents an important first step in the treatment of these disorders. In the absence of an ideal antiviral agent, however, combinations of the available antiviral and immunomodulatory agents or synergistic combinations of antiviral agents need to be studied in order to achieve better therapeutic responses.

The objective was to determine the proportion of patients with chronic hepatitis B in whom hepatitis B virus DNA is demonstrated by polymerase chain reaction after HBeAg to anti-HBe or HBsAg to anti-HBs spontaneous or therapeutically induced seroconversion. Polymerase chain reaction was performed on serum 6 and 12 mo after HBeAg to anti-HBe seroconversion in 12 patients and 2, 6 and 12 mo after HBsAg to anti-HBs seroconversion in 13 patients. Polymerase chain reaction was performed on liver tissue after HBeAg to anti-HBe seroconversion in five patients and after HBsAg to anti-HBs seroconversion in one patient. Serum HBV DNA was demonstrated by polymerase chain reaction in 83% of patients 6 or 12 mo after HBeAg to anti-HBe seroconversion and in 58%, 31% and 15% of patients at 2, 6 and 12 mo, respectively, after HBsAg to anti-HBs seroconversion. Liver HBV DNA was demonstrated by polymerase chain reaction in all patients tested. Our results show that (a) a reduced level of hepatitis B virus replication persists in most of the patients after HBeAg to anti-HBe seroconversion and might be predictive of reactivation, and (b) in contrast, hepatitis B virus replication progressively disappears in most of the patients after HBsAg to anti-HBs seroconversion.

A pilot study was designed to determine the tolerance and effectiveness of natural or recombinant gamma interferon in patients with chronic hepatitis B. Sixteen patients received 0.5 to 3.0 million units (MU) per day of gamma interferon (IFN-gamma) for 7 days. Nineteen chronic hepatitis B patients who were treated with 5-6 MU leukocyte-derived alpha interferon (IFN-alpha) daily served as controls. All completed the treatment schedule. IFN-gamma exerted mild, but significant inhibitory effects (P less than .05) on serum DNA polymerase levels. However, the changes were significantly less (P less than .001) than those seen with IFN-alpha therapy when compared with percent change from basal values. In contrast, serum 2', 5'-oligoadenylate synthetase (2-5 AS) activities were markedly enhanced to a similar extent during therapy with both IFNs. Serum beta 2-microglobulin values were significantly increased by administration with both IFNs, although higher values were seen with IFN-gamma. Five patients received 1 MU IFN-gamma for 28 consecutive days and their HBeAg levels similarly decreased as those seen in patients treated with IFN-alpha. Side effects seemed to be greater during IFN-gamma therapy than IFN-alpha despite the lower doses used. The antiviral effect on serum HBV levels appeared less with IFN-gamma than with IFN-alpha. Alternatively immunomodulatory functions may have been enhanced with IFN-gamma in patients with chronic HBV infection.

Nineteen Chinese patients with chronic hepatitis B virus (HBV) infection, seropositive for HBV e antigen (HBeAg) and HBV DNA on at least three occasions in 6 months, were randomised to receive either recombinant human interferon-gamma (rIFN gamma) 0.1 mg/m2 intramuscularly thrice weekly for 16 weeks (n = 11) or no anti-viral therapy (controls, n = 8). Five patients in the treatment group and four patients in the control group had persistently elevated serum alanine aminotransferases (ALT) of over two times the upper limit of normal before entering into the trial. rIFN gamma had no or minimal inhibitory effect on serum HBV DNA during treatment and no patient developed e-seroconversion or sustained loss of serum HBV DNA. Hepatitic flare, which occurred in a proportion of patients responding successfully to interferon-alpha (IFN alpha) therapy, was not observed with rIFN gamma treatment. Side-effects included pyrexia and mild headache that showed tachyphylaxis and were well tolerated by all patients. In the control group, one patient with elevated pre-entry serum ALT lost serum HBV DNA and seroconverted to anti-HBe. Another patient with elevated ALT lost serum HBV DNA transiently during therapy. In the dose given, rIFN gamma was safe but had no apparent anti-viral effects in Chinese patients with chronic HBV infection.

At present immune-based therapies are not an alternative to alpha-interferon in the treatment of chronic hepatitis B. Patients who do not respond to alpha-interferon should probably be followed without re-treatment. Re-treatment using a short course of prednisone before alpha-interferon deserves further study. Future investigation of immune-based therapies should be done in the context of available antiviral therapies. Alternatives need to be propounded and tested, especially alternatives that cause us to question assumptions and to rethink our current beliefs about chronic HBV immunopathogenesis.