Treatment of chronic hepatitis D: New advances, old challenges


  • Patrizia Farci M.D.

    Corresponding author
    1. Department of Medical Sciences, University of Cagliari, Cagliari, Italy
    • Department of Medical Sciences, University of Cagliari, Policlinico Universitario, SS 554 Bivio Sestu, 09042 Monserrato, Cagliari, Italy
    Search for more papers by this author
    • fax: (39) 070-510064

  • See Articles on Pages 713 and 728.

  • Potential conflict of interest: Nothing to report.

Chronic hepatitis D is the least common but the most severe form of viral hepatitis,1 caused by one of the most interesting and unusual human pathogens, hepatitis D virus (HDV),2 which is a defective RNA virus that requires the helper function of hepatitis B virus (HBV) for its assembly and transmission. HDV is the smallest animal virus and the only one to possess a circular RNA genome and a single structural protein, hepatitis delta antigen, encapsidated by the hepatitis B surface antigen (HBsAg).3 It does not encode its own polymerase, but exploits a host cellular enzyme for its replication.4 Because of its obligatory link with HBV, infection with HDV occurs only in persons who simultaneously harbor HBV. Worldwide, it has been estimated that 15 million HBsAg carriers are infected with HDV. Although over the past decade there has been a dramatic decline in the prevalence of HDV in Southern Europe,5, 6 residual disease is still present in patients infected when HDV was endemic. These patients pose a major therapeutic challenge because most of them have advanced liver disease.


HDV, hepatitis delta virus; HBV, hepatitis B virus; PCR, polymerase chain reaction; PEG-IFNα, pegylated interferon-alpha; IFNα, interferon-alpha; ALT, alanine aminotransferase.

The unique replication process of HDV, particularly the lack of a virus-specific polymerase, and its high pathogenic potential make chronic hepatitis D a challenging target for antiviral therapy. No specific inhibitor of HDV has so far been developed and in spite of the vital relation between HDV and HBV, drugs that specifically block HBV have little or no effect on HDV replication. This is most likely due to the fact that the only helper function that HBV provides to HDV is the HBsAg envelope, which is efficiently expressed in most HBV carriers regardless of the level of HBV replication. Thus, an effective anti-HDV therapy would require a marked suppression of HBsAg expression, but current therapies for HBV do not achieve this. Lamivudine, a nucleoside analog that potently inhibits HBV replication, has no efficacy on HDV viremia or liver disease activity in patients with chronic hepatitis D.7, 8 Likewise, other antiviral agents such as suramin, acyclovir, ribavirin, and famcyclovir, as well as immunomodulating agents such as steroids, thymosin, levamisole, and thymic humoral factor-gamma 2 have proved ineffective against HDV.9 Clevudine, another nucleoside analog that potently suppresses HBV replication,10 was recently shown to suppress HDV replication in the woodchuck model.11 Such suppression correlated with decreased levels of woodchuck hepatitis virus surface antigen, consistent with the hypothesis that targeting surface antigen may be an important antiviral strategy against HDV.11 However, data on clevudine in humans with chronic hepatitis D are lacking. Encouraging results, albeit preliminary, were also obtained with prenylation inhibitors, which block a post-translational modification of the large HDV antigen that is a critical determinant for viral assembly.12 Using a novel transgenic mouse model for HBV that supports high levels of HDV replication, prenylation inhibitors were shown to be highly effective in clearing HDV viremia, thereby opening new perspectives for a novel class of potential antiviral agents.13

The only option currently available for the treatment of chronic hepatitis D is interferon-alpha (IFNα),14 which is also the only licensed drug for this disease. In the early 1990s, controlled clinical trials15–18 provided evidence that IFNα is effective in chronic hepatitis D, but the rate of relapse is high and the efficacy is proportional to the dose and duration of treatment. Patients treated with 9 million units (MU) for 12 months responded significantly better than those receiving 3 MU, with alanine aminotransferase (ALT) normalization in 71% of patients at the end of treatment and in 50% after 6 months of follow-up.16 Although HDV RNA, as measured by polymerase chain reaction (PCR), remained detectable, the levels of viremia in the high-dose group decreased significantly (up to 4 logs) at the end of treatment. Only recently the long-term effects of IFNα have been investigated, providing evidence of a beneficial effect that extends beyond the termination of therapy. High doses of IFNα significantly improved the long-term clinical outcome and survival of patients with chronic hepatitis D even though the majority had active cirrhosis before treatment.19 However, the therapy of chronic hepatitis D is not yet satisfactory. Several strategies have been explored to improve the efficacy of IFNα, including longer duration of treatment or even continuous therapy for up to 12 years,20 but most of the patients still failed to clear HDV and the rate of relapse remains high. Moreover, these alternatives are poorly tolerated. More recently, a few small-size studies investigated the efficacy of standard IFNα alone or in combination with ribavirin or lamivudine. In contrast with the results obtained in chronic hepatitis C, the addition of ribavirin failed to show any beneficial effects,21, 22 whereas combination with lamivudine was of limited benefit23 (Fig. 1).

Figure 1.

Rate of sustained virological response in recent clinical trials in patients with chronic hepatitis D treated with standard or pegylated IFNα, alone or in combination with lamivudine or ribavirin. In all studies, HDV RNA was assessed by nested PCR assays with a detection sensitivity ranging from 10 to 1,000 genome copies per mL. PEG-IFNα indicates pegylated interferon alpha; IFNα, standard interferon alfa; LAM, lamivudine; RBV, ribavirin.

Two studies in this issue of HEPATOLOGY evaluated the efficacy and safety of pegylated IFNα (PEG-IFNα), alone24 or in combination with ribavirin,25 for the treatment of chronic hepatitis D. Given the increased response rates observed with PEG-IFNα in chronic hepatitis C and B, it was natural to assess its effectiveness in chronic hepatitis D. Castelnau et al. evaluated the efficacy of PEG-IFNα2b monotherapy, 1.5 μg/kg weekly for 12 months, in 14 patients with chronic hepatitis D.24 A virologic response, defined as undetectable HDV RNA by qualitative PCR, was seen in 57% of the patients at the end of treatment and persisted in 43% during follow up (Fig. 1). The rate of ALT normalization was higher at the end of follow up (57%) than at the end of therapy (36%), indicating that in some patients it was achieved during post-treatment follow up. Using real-time PCR to quantify the levels of viremia, the authors found that in patients with a virologic response at the end of treatment, HDV RNA levels significantly decreased during the first 3-6 months of treatment, becoming undetectable by month 6 in 6 of 8 responders, whereas the changes were not significant in nonresponders. However, two end-of-treatment responders relapsed after cessation of treatment, indicating that an early viral response does not differentiate responders from relapsers. Of note, there was a delayed decrease of HDV RNA in two nonresponders, which became evident at the end of the treatment period. These patients might represent “slow responders” who would benefit from a more prolonged course of IFNα therapy.

In the second study,25 Niro et al. report the results of a randomized controlled trial comparing the efficacy and safety of PEG-IFNα alone (1.5 μg/kg/week) for 72 weeks (16 patients) or in combination with ribavirin (800 μg/day) for 48 weeks followed by PEG-IFNα monotherapy for 24 additional weeks (22 patients). Clearance of HDV RNA at the end of treatment occurred in 19% of patients receiving monotherapy and in only 9% of those receiving combination therapy (P = NS). As previously documented with standard IFNα,16 the rate of virological response increased in both groups during post-treatment follow-up to reach 25% and 18%, respectively (Fig. 1). This delayed response was likely due to the long-term immunomodulatory effects of IFNα, as seen in chronic hepatitis B.26 The rates of biochemical response were similar in the two groups of patients (25% versus 27% at the end of follow up) and the overall rate of relapse was high (60%).

The studies by Castelnau et al. and Niro et al. provide the first evidence that PEG-IFNα is well tolerated and effective in the treatment of chronic hepatitis D, even though most patients were unresponsive to a previous course of conventional IFNα. Moreover, in the second study, 85% of the patients had cirrhosis. Thus, in line with the superior results obtained in chronic hepatitis B and C, these preliminary studies suggest the use of PEG-IFNα as the standard care in chronic hepatitis D. Long-acting IFNα, given once weekly, is also likely to provide better compliance for the long-term treatment required in chronic hepatitis D. The addition of ribavirin to PEG-IFNα does not seem to result in a higher benefit. However, larger controlled trials are needed to better evaluate the benefit of PEG-IFNα and to define the optimal treatment schedule in IFNα-naïve patients as well as in previous nonresponders to standard IFNα, even though the changing epidemiology of HDV has made it increasingly difficult to recruit IFNα-naïve patients. Although the major goal in the treatment of chronic hepatitis D is eradication of both HDV and HBV, a one-year course of standard IFNα is associated with only a 10%-20% chance of sustained HDV clearance14 and 10% chance of HBsAg elimination.27 The rate of sustained virological response of 43%, as measured by qualitative PCR, reported by Castelnau et al. using PEG-IFNα is the highest so far documented and is markedly higher than that reported by Niro et al. This discrepancy likely reflects the different proportion of patients with cirrhosis, who are less likely to respond to treatment, enrolled in the two studies (28% versus 74%), highlighting the importance of patient selection when comparing clinical trials. One encouraging note is the high rate of virological response reported by Niro et al. in the post-hoc analysis of 8 patients who were naïve to previous therapy (3/8, 37.5%). In these two studies, HBsAg was eliminated in only one patient, although the follow up was presumably too short for such an event to occur.

These studies also confirm the importance of using quantitative assays for viremia in treatment monitoring, as previously reported.19, 28 However, HDV RNA detection still relies on home-made assays, and it is urgent to develop standardized PCR assays for the detection and quantification of HDV viremia. Castelnau et al. show that HDV RNA quantification identifies patients with an early virological response as well as those with a late decrease in viremia, who might benefit from a more prolonged course of therapy. Even with the use of quantitative methods for viremia, however, there is currently no way to predict who will have a sustained viral loss or will relapse after cessation of therapy. Although patients with a shorter duration of disease may respond better to therapy, clear predictors of response have not yet been identified. Thus, treatment with IFNα or PEG-IFNα at high doses should be offered to all patients with compensated chronic hepatitis D as soon as the diagnosis is made, and continued for at least 1 year before a patient is regarded as a nonresponder. One of the crucial challenges is to decide when treatment should be stopped in a patient with a good response because a significant decline in viremia may lead to a delayed virological response while a loss of serum HDV RNA during treatment does not preclude relapse. Because the major goal is eradication of both HDV and HBV, in patients with a significant decline in HDV RNA after 6-12 months of therapy, IFNα should be continued as long as possible until HDV RNA and HBsAg disappear. Likewise, given the high rate of relapse, therapy should be continued in patients with undetectable HDV RNA until the loss of HBsAg. These recommendations should be applied to both IFNα-naïve as well as to nonresponders to standard IFN who are retreated with PEG-IFNα. Although the rate of virologic response to PEG-IFNα in previous nonresponders differs in the two studies, more than 20% of these difficult-to-treat patients achieved a sustained virologic response, suggesting that a course of PEG-IFNα should be offered to all patients with well-compensated liver disease. The main concerns with long-term treatment are side effects, which are common during standard and PEG-IFNα, particularly in patients with cirrhosis, as shown in the study of Niro et al. Therefore, continuing medical monitoring is essential for the early detection and management of medical and psychiatric complications, and the dose of IFNα after month 12 may need to be tailored according to patient tolerance, ALT response, and whenever possible HDV RNA levels. Although the first results obtained with PEG-IFNα in chronic hepatitis D have been encouraging, the low rate of response and the high rate of relapse emphasize the need to determine predictors of response to IFNα therapy as well as the need to identify, through innovative molecular approaches, new antivirals for the treatment of chronic hepatitis D. Until then, we will continue to tackle the old challenges posed by HDV.