Pathogenesis of and New Therapies for Hepatitis D

Abstract

Hepatitis delta virus (HDV) infection of humans was first reported in 1977, and now it is now estimated that 15–20 million people are infected worldwide. Infection with HDV can be an acute or chronic process that occurs only in patients with an HBV infection. Chronic HDV infection commonly results in the most rapidly progressive form of viral hepatitis; it is the chronic viral infection that is most likely to lead to cirrhosis, and it is associated with an increased risk of hepatocellular carcinoma. HDV infection is the only chronic human hepatitis virus infection without a therapy approved by the Food and Drug Administration. Peginterferon alpha is the only recommended therapy, but it produces unsatisfactory results. We review therapeutic agents in development, designed to disrupt the HDV life cycle, that might benefit patients with this devastating disease.

The hepatitis Delta virus (HDV) was first identified in humans in 1977, when a cohort of patients with hepatitis B virus (HBV) presented in Italy with severe hepatitis.¹ Immunohistochemical analyses of liver biopsies and sera from these patients revealed a novel antigen pattern in the nuclei of liver cells, which researchers called the δ antigen.^{1, 2} Initially, this δ antigen was believed to be a new HBV antigen, but it was quickly found to be separate and associated with the HBV surface antigen (HBsAg).¹ Transmission experiments in chimpanzees led to further characterization of the δ antigen (HDAg)—a structural component of a distinct infectious pathogen with a low molecular weight RNA genome that required HBV for its life cycle.^{2, 3} Subsequent cloning and sequencing of this genome confirmed that this was a unique RNA virus; its classification is still used today, as the only member of a separate genus, deltavirus.⁴

HDV is estimated to affect 15–20 million people worldwide from all age groups. Infection with this pathogen can lead to an acute or chronic disease, but only in individuals with HBV infection. Although HDV infection varies among geographic regions, it is believed that approximately 5% of the HBV-infected population worldwide is coinfected with HDV.⁵ Chronic infection with HDV is considered to be the most severe form of human viral hepatitis infection. Compared with other chronic viral hepatitis infections, HDV infection has been described to progress more rapidly, is more likely to lead to cirrhosis, and is associated with increased risk of hepatocellular carcinoma (HCC). HDV infection is the only chronic human hepatitis virus infection without a Food and Drug Administration-approved therapy.

Virology

The HDV particle is approximately 36 nm in diameter and consists of a 1.7 kb single-stranded circular RNA genome of negative polarity.^{2, 6, 7} Eight genotypes of HDV have been identified (Figure 1); the separation in nucleotide sequence of the studied region is typically <16% among isolates of the same genotype and up to 40% over the full-length sequence among genotypes.^8–10 Genotype 1 is found worldwide and is the predominant virus in North America, Europe, Australia, and the Middle East.^9–11 Genotype 2 has been identified predominantly in Asia, Southeast Asia, and Russia.^12–15 Genotype 3 has been identified in South America.^16–18 Genotype 4 has been described in Japan and Taiwan.^{15, 19} Finally, genotypes 5–8 have been predominantly found in Africa.^{9, 10, 20}

Figure 1. Map of the Prevalence of HDV Infection.

World map shows estimated prevalence of HDV infection and genotype distribution.

The HDV genome encodes 1 protein that exists in 2 forms: the small delta antigen (SHDAg) and the large delta antigen (LHDAg) (Figure 2a). The LHDAg is identical to the SHDAg except that it contains an extra 19 amino acids at its carboxyl terminus. This extension results from a specific RNA-editing event that occurs during genome replication.²¹ The addition of these 19 amino acids alters the carboxy terminus of the protein, which includes a CXXX-box motif (C=cysteine, X=one of the last 3 amino acids at the carboxyl terminus of the LHDAg). This CXXX-box motif is a substrate for farnesyltransferase, an enzyme that adds a farnesyl group to the cysteine of the CXXX-box.²² The farnesylation of the CXXX-box is essential for virion assembly.²³ The complete HDV particle comprises a complex of the viral genome and both HDV antigen isoforms, all surrounded by a lipid envelope in which HBsAg proteins are embedded.²⁴ The 3 isoforms of HBsAg (small, middle, and large [or pre-S1]) that are embedded are the same as those found on the HBV virion and are myristoylated at the N-terminus—a modification required for cell entry.²⁵ HDV does not encode its own envelope proteins, but instead utilizes the HBV envelope proteins (Figure 2b) for HDV assembly and infection of new cells. The presence of HBV, either via coinfection or superinfection, is therefore essential for HDV propagation in humans.

Figure 2. HBV and HDV Structure.

(A) The small delta antigen (SHDAg) and the large delta antigen (LHDAg) are each derived from the single expressed reading frame in the HDV genome. The SHDAg and LHDAg differ based on a specific RNA editing event, occurring during genome replication—this results in the addition of 19 amino acids to the SHDAg carboxyl terminus. (B) HBV and HDV share envelope proteins. HDV uses the envelope proteins of HBV for assembly and infection of new hepatocytes. Therefore, the presence of HBV infection is required for HDV infection and propagation in humans.

HDV genome replication occurs solely through RNA-dependent RNA replication, without DNA intermediates or chromosomal integration events. Additionally, HDV does not encode its own RNA-dependent RNA polymerase but requires cell polymerases for its replication. Notably, HDV appears to recruit RNA polymerase II to replicate its RNA genome, in an RNA-dependent manner.²⁶ It is believed that the double stranded-like structure of the HDV genomic RNA, created by the high intramolecular RNA folding into a collapsed circle, or rod-like structure, provides a template similar to double-stranded DNA, which enables recruitment of RNA polymerase II.²⁷ Upon uncoating, the virus genome is transported to the liver cell nucleus where genome replication occurs.

RNA replication occurs via a double-rolling circle mechanism²⁸ by RNA polymerase II, which associates with the SHDAg to create linear multimeric copies of antigenomic RNA. The plus-sense antigenomic RNA is complementary to, and is of opposite polarity to, that of the incoming, negative-sense, genomic RNA. The linear multimeric copies of antigenomic RNA then undergo intramolecular cleavage by the antigenomic ribozyme to form linear monomers that are then ligated, forming closed circular antigenomes which serve as templates for production of genomic RNA via a similar mechanism.²⁶ As replication continues, an ADAR1-mediated RNA editing event takes place on the antigenomic RNA, which modifies the amber stop codon of the reading frame encoding SHDAg. This results in translation proceeding to the next downstream stop codon, adding an extra 19 amino acids with resultant expression of the LHDAg.²⁹ The addition of the extra 19 amino acids significantly changes the function of the delta antigens—the SHDAg is required for HDV RNA genomic replication whereas the LHDAg functions as a potent trans-dominant inhibitor of HDV RNA replication.³⁰ Moreover, only the LHDAg mediates packaging with HBsAg into virions. Therefore, RNA editing event is an important part of the HDV life cycle that serves to turn off RNA replication and turn on packaging of newly created HDV genomes, virus assembly, and release.³¹

Epidemiology

Although an estimated 15–20 million people are infected with HDV worldwide, the global prevalence is not known. Certain areas of high prevalence have been described and include: Central Africa (15%–50%)^32–34, West Africa (17%–30%)^{35, 36}, the Mediterranean basin (27%)³⁷, the Middle East (Iran, 7.8%)³⁸, Northern Asia (Mongolia, 26%–60% in adults and >6% in school-aged children)^39–41, Eastern Europe (Romania, 20%)⁴² (Parts of Russia, 22%)⁴³, certain areas of Southeast Asia (Vietnam, 15%)⁴⁴, and the Amazon basin of South America (13%–29%).^{45, 46}

Initial reports of endemicity, with prevalence rates >20%, were described in Southern Europe after HDV’s initial discovery in Italy. However, rates of HDV infection in this region have decreased substantially since the 1990s as a result of disposable syringe practices, HBV vaccination programs, socioeconomic improvements, and deaths of individuals with chronic infection.^47–50 However, despite these improvements, new patients with HDV continue to be identified in this region, likely due to migration from endemic areas.^51–53

Reporting of HDV prevalence in the US continues to be incomplete. In a study in Northern California, the prevalence of HDV among HBV-infected patients was 8%. However, only 42% of the 1191 HBsAg-positive subjects in this study were tested for HDV.⁵⁴ In a Midwestern population of HBV-infected patients, only 12% of 1007 HBsAg-positive patients were tested for an HDV antibody—among those tested, 3.3% were positive for the HDV antibody.⁵⁵ In a Veterans administration study, where suboptimal testing for HDV was also identified, 3.4% of HBsAg-positive adults were reported to have HDV infection.⁵⁶

Among injection-drug users (IDUs) in the US, the prevalence of HDV infection has increased. In an IDU cohort in Baltimore, Maryland, approximately 50% of HBV-infected individuals were found to be infected with HDV.⁵⁷ In a cohort of IDUs assessed by the Urban Health Study in San Francisco, 36% of HBsAg-positive individuals were found to have HDV viremia.⁵⁸ So, even in developed countries, subgroups of HBV-infected patients (such as IDUs) are at high risk for HDV infection. Also, immigration from endemic areas continues to be a significant risk factor for HDV infection in the developed world.

Pathogenesis

HBV infection is required for productive HDV infection in humans. HDV transmission in humans occurs either via co-infection with acute HBV or as a super-infection in patients with HBV infection. The exact mechanisms of interaction between these viruses have not been elucidated, but HDV reduces certain activities of HBV and viremia.^59–62 Retrospective studies have found the presence of HDV infection to reduce HBV DNA levels in patients’ blood.^62–64 In a study that followed patients with HBV and HDV coinfection for up to 8 years, researchers found 3 predominant patterns of HDV RNA and HBV DNA in blood samples from patients with active viral replication: high levels of HDV RNA and low levels of HBV DNA (54% of the patients), low levels of HDV RNA and high levels of HBV DNA (30% of patients), and equal levels of HDV RNA and HBV DNA (15% of patients).⁶⁵

Little is known about the pathogenesis of HDV infection. Although HBV is not directly cytopathic to infected hepatocytes, HDV may have direct cytopathic effects specifically related to the SHDAg.^66–68 Additionally, innate and adaptive immune responses could be involved in mediating liver damage. The LHDAg promotes an inflammatory response, possibly by activating STAT3 and NF-kB, although this might help clear HDV-infected cells.^{69, 70} The effects of this inflammatory response include endoplasmic reticulum stress and necroinflammation, along with possible increases in production of reactive oxygen species, which might ultimately promote development of HCC.⁷⁰ HDV may also have the ability to interfere with interferon (IFN) alpha signaling, by blocking activation and translocation of STAT proteins, which may contribute to the persistence of HDV and impair IFN alpha-based therapies. The adaptive immune response to HDV infection has been described as weak. These features combined might allow chronic HDV infection to induce rapid, progressive liver disease.

Clinical Features of Acute Infection

Acute HDV infection can occur via either HBV co-infection (simultaneous infection with both viruses during the same exposure) or super-infection (infection with HDV after an established HBV infection, such as in an HBsAg-positive individual). The clinical course of an acute co-infection is similar to that of acute HBV infection. However, co-infection causes more severe disease, with an increased risk of acute hepatic failure.⁷¹ Additionally, acute co-infections cause a biphasic course in alanine aminotransferase (ALT) levels—2 peaks are often observed, several weeks apart, as HBV infection must first be established before HDV can begin to spread.^{71, 72} Acute superinfection of HDV has a more severe clinical course compared to acute co-infection, with an increased risk of acute liver failure.⁷³ In patients with known HBV infection, acute superinfection may be mistaken for an HBV flare, whereas in patients with undiagnosed viral hepatitis, an acute superinfection may be mistaken for acute HBV infection. Similar to HBV mono-infections, HDV infection becomes chronic in fewer than 5% of patients who become co-infected as adults, but it becomes chronic in most people co-infected during the neonatal period, and in more than 90% of cases of super-infections.^{74, 75}

Chronic Infection and Disease Progression

Once chronic HDV infection has been established (based on detection of the HDV antibody more than 6 months after infection, with detectable HDV RNA in serum or HDAg, detected by immunohistochemistry) levels of transaminases are usually increased to a greater extent than in patients with HBV mono-infection. Compared with chronic HBV or HCV infection, chronic HDV infection leads to more severe liver disease, with increased rates of fibrosis progression.^{76, 77} Chronic HDV infection progresses to cirrhosis within 2 years in 10%–15% of cases⁷⁵, and within 5 to 10 years in up to 80% of cases.⁷⁸ Despite the rapid progression to cirrhosis, markers that can be detected noninvasively that have been validated for monitoring fibrosis in patients with HBV or HCV infection, such as the aspartate aminotransferase to platelet ratio index or fibrosis-4 index, do not appear to be reliable for monitoring fibrosis in patients with chronic HDV infection.⁷⁹ The performances of commercial noninvasive fibrosis assays (such as hepascore, fibroscore, or fibrosure) have not been evaluated in patients with HDV infection. Although transient elastography has been approved for the staging of liver disease, it has not been evaluated in patients with HDV infection. It is not known whether the extent of hepatic inflammation caused by HDV affects transient elastography measurements.

Outcomes related to chronic HDV infection have been reported to be more severe than those of chronic HBV infection. The risk of HCC, as a complication of cirrhosis, is up to 3-fold higher in co-infected patients compared to patients with HBV mono-infection (Table 1).^{78, 80–83} Patients with HDV infection are also 2-fold more likely to develop hepatic decompensation than patients with only HBV infection,^{84, 85} and are more likely to die from hepatic decompensation or HCC.^{84, 85} HDV genotype 1 appears to be more virulent than genotype 2⁸⁶, although additional factors, such as ethnicity and location, are likely to affect virulence.

Table 1.

Risk of HCC in Patients with HDV Infection

Geographic area	Study type	Follow-up period	Number of patients with only HBV infection	Number of patients positive for anti-HDV	Results pertaining to HCC risk in patients with HDV compared to HBV mono-infection	Reference
Western Europe	Longitudinal study of patients with HBV infection and compensated cirrhosis	6.6 years (median)	161	39	HCC risk increased 3.2-fold in patients positive for anti-HDV	ref78
Jordan	cohort study of HBsAg-positive patients	8 years	195	20	Higher prevalence of HDV in patients with HCC (10/15, 67%). Subjects positive for anti-HDV developed HCC an average 10 years earlier than anti-HDV negative patients	Ref83
Greece	case-control	9 years	117	9	10% of patients with HCC (n=87) were HDV-positive compared to 0 patients without HCC	ref148
Italy	cohort study	2 years	14	9	Patients with HDV infection and cirrhosis developed HCC at a significantly younger age than patients with only HBV infection and cirrhosis	ref149
Japan	cohort study	121 months (mean)	1058	69	Patients positive for anti-HDV developed HCC at a rate of 7.84 per 1000 person-years compared to 2.73 in patients with only HBV infection	ref150
England	retrospective, cross-sectional	6 years	880	82	Risk of HCC was 9.7% in patients with HDV coinfection compared to 7.8% of patients with only HBV infection	Ref80
Sweden	registry analysis	12 years	8510	327	Compared to patients infected with only HBV (reference population) patients with HDV co-infection had an increased risk for HCC (standardized incidence ratio, −6.11)	Ref82
United States	retrospective analysis of Veterans Health Administration population		25,603	73	Patients positive for anti-HDV developed HCC with a 2.9-fold higher incidence rate than patients negative for anti-HDV; HDV infection was independently associated with HCC (odds ratio, 2.1).	Ref56

Diagnosis and Tests

A moderate to high index of suspicion for HDV infection should be maintained in patients with specific risk factors. Screening for HDV infection continues to be inadequate in the US.^{54, 56, 58} Risk factors include a history of intravenous drug use, high-risk sexual behaviors, infection of a first-degree relative, and immigration from endemic regions. In patients with acute infections, HDV antigen can be measured by ELISA or radioimmunoassay. However, the virus can be detected only during the first 2 weeks of acute infection, and is only transiently detectable thereafter.^87–89 Acute infection with HDV induces innate and adaptive immune response, so immunoglobulin M (IgM) and G (IgG) can also be detected, in immune-competent patients.⁹⁰ Anti-HDV IgM typically appears 2–3 weeks after the onset of symptoms and disappears by 2 months after an acute infection (although it may persist for as long as 9 months in HDV superinfection).

Assays to detect HDV IgM might be used to identify patients with acute infections, but IgM has been reported to increase during disease flares in patients with chronic HDV infection. HDV IgG and HDV total antibodies persist in serum after resolution of acute HDV infection and in patients with chronic coinfection. Today, these antibody tests are typically utilized as initial screening tests for the detection of HDV infection (HDV Total Antibody and HDV IgM Antibody). A quantitative microarray antibody capture (Q-MAC) assay that quantifies the amount of anti-HDV IgG in serum has been reported to accurately identify patients with HDV infection in Mongolia and the US.^{41, 58}

After detection of HDV antibody, serum should be tested for HDV RNA, to confirm infection. Hybridization assays for HDV RNA have been largely replaced by qualitative or semiquantitative real-time PCR assays, due to their improved sensitivity and a lower limit of detection (as low as 10 genomes/ml).⁹¹ However, results obtained from different laboratories are often not comparable, due to the range in sensitivity of the assays.^{34, 92} Notably, a World Health Organization international RNA standard is now available, allowing for reporting of results in international units (IU). High-diversity HDV genotypes (such as genotypes 6–8) are often challenging to detect and quantify, although pan-genotypic assays have become available.⁹¹ New assays have been explored for HDV RNA quantitation^{41, 93, 94}, and a quantitative HDV RNA assay has recently become commercially available in the US.

Another method of HDV detection, albeit invasive, is the intrahepatic detection of HDAg, in which liver tissue is collected and analyzed by immunohistochemistry. Although the reported sensitivities of this test for identification of HDV vary⁹⁵, in cases for which a liver biopsy is obtained, these samples might be analyzed for HDV—the virus can be detected via immunohistochemistry in liver cell nuclei. However, with the availability of serologic assays, liver immunohistochemistry analyses are not often used.

IFN-based therapies

Although professional societal guidelines have recommended pegylated IFN (PegIFN) alpha for treatment of chronic HDV infection, there is no satisfactory or Food and Drug Administration-approved therapy for this potentially devastating disease.^96–98 Although patients with HDV infection have been treated with PegIFN, its administration is limited and typically avoided in patients with cirrhosis, active autoimmune disease, or certain psychiatric disorders. A summary of the results from clinical trials of IFN-based therapies is presented in Supplementary Table 1.

Initially, the effects of IFN alpha-2a in patients with HDV infection were studied groups that received 9 million international units IFN 3 times per week (high-dose), 3 million international units IFN 3 times per week (low-dose), or no therapy for 1 year (control).⁹⁹ A complete response was defined by normalization in level of ALT and negative results from PCR analyses of HDV RNA—this was achieved by 50% of patients in the high-dose IFN group. In the low-dose group, 21% of patients had a complete response and none of the patients in the no-therapy group had a complete response. During a follow-up period of up to 48 weeks after therapy, all patients had relapsed. Interestingly, in a subsequent analysis of the same cohort, with follow-up period of as long as 14 years after therapy, survival was significantly longer for patients who received high-dose IFN compared to patients who received low-dose IFN or patients who did not receive any therapy.¹⁰⁰ Notably, achieving a 2 log₁₀ decline in HDV RNA at end of treatment was associated with the significant increase in survival. There was no difference in long-term outcomes between the low-dose IFN group and the no therapy group, neither of which achieved the mean 2 log₁₀ decrease in HDV RNA at end of treatment.

The effects of different durations of PegIFN alpha-2b therapy, and combinations with other drugs, were evaluated in patients with chronic HDV infection. Among patients given PegIFN alpha-2b (1.5 ug/kg/week) for 1 year, 43% were HDV-negative after a median post-therapy follow-up time of 16 months (range, 6–42 months).¹⁰¹ PegIFN is therefore at least as effective as standard IFN for the treatment of chronic HDV infection. However, extending therapy with PegIFN alpha-2b for 72 weeks¹⁰², or administering PegIFN alpha 2a with increasing doses up to 360 mcg/wk for up to 5 years^{103, 104}, did not increased rates of sustained virologic response (<35%). Interestingly, 1 patient who received IFN alpha therapy for 12 years was reported to have a complete virologic response (negative results from PCR tests of HDV RNA with HBsAg seroconversion).¹⁰⁵ Despite the variations of responses (ranging from 20% to 40%) to PegIFN alpha, levels of HDV RNA at 24 weeks of therapy can identify patients most likely to have a response to 48 weeks of therapy.¹⁰⁶

Combination therapies with PegIFN alpha have been investigated, without much success. The combination of ribavirin with PegIFN alpha for 48 weeks followed by PegIFN monotherapy for an additional 24 weeks did not improve patient outcomes, compared to PegIFN monotherapy for 72 weeks.¹⁰² Nucleos(t)ide analogue therapy alone^{107, 108} has shown no benefit, and combination of nucleos(t)ide analogue with PegIFN did not provide any benefit compared with PegIFN monotherapy^109–112.

New Therapeutic Approaches

Although PegIFN alpha can be effective in patients with HDV infection, this drug has significant side effects, because IFN receptors are expressed by many different cell types. In contrast, receptors for PegIFN lambda (a type III IFN) are restricted to specific cell types, including liver cells.¹¹³ IFN lambda was found to be expressed in response to viral infections and to have antiviral activity in mice.¹¹⁴ IFN lambda induces expression of IFN-stimulated genes to induce a broad-spectrum anti-virus immune response. IFN lambda binding to type III IFN receptors results in receptor dimerization, leading to activation of signal transduction pathways mediated by Janus kinase (JAK) and signal transducer and activator of transcription (STAT), similar to pathways stimulated by IFN alpha. In early-stage trials of patients with liver disease, PegIFN lambda reduced HBV and HCV virus levels in blood, similar to PegIFN alpha, but with significantly fewer side effects.^{115, 116} However, the limited utility of IFN-based therapies for patients with HBV infection and the efficacy of direct antiviral agents in patients with HCV infection has reduced the number of studies into the effects of PegIFN lambda.

IFN alpha and lambda were found to reduce markers of intrahepatic HDV infection in mice with humanized livers infected with HBV and HDV.¹¹⁷ In these mice, PegIFN lambda reduced HDV viremia by 1.2 log₁₀ and serum levels of HBsAg by 0.4 log₁₀. Given its tolerability profile and anti-HDV activity, PegIFN lambda is being tested in patients with chronic HDV infection (Table 2). In randomized, open-label, multi-center trial, 33 patients with chronic HDV infection are receiving PegIFN lambda (120 or 180 μg weekly) for 48 weeks.¹¹⁸ An interim analysis has revealed that tolerability was better than that of PegIFN alpha and that both doses of PegIFN lambda have activity against HDV—some patients were found to be negative for HDV RNA, by PCR analysis, at week 8 of therapy.¹¹⁸

Table 2.

Results from Clinical Trials of Therapeutic Agents for HDV Infection

Therapeutic agent	Treatment duration	Baseline HDV load	Biochemical outcome measured	Virologic outcome measured	Serologic outcomes measured	Reference and number of patients
PegIFN lambda (120 or 180 μg/week)	48 weeks	4.5 log10 IU/mL*	NR	6 of 10 (60%) with >/= 2 log10 decline at week 24 4 of 10 (40%) PCR negative by week 24	NR	Ref148, n=33
Daily, subcutaneous Myrcludex B (2 mg/day) for 24 weeks then PegIFN alpha monotherapy for 24 weeks	48 weeks	104,14 (copies/ml)*	normalized level of in 6/8 patients	102.47	No change in HBsAg	ref124
Daily, subcutaneous Myrcludex B (2 mg/day) for 24 weeks then PegIFN alpha monotherapy for 24 weeks	48 weeks	104.14 (copies/ml)*	normalized level of ALT in 1/8 patients	101.62	NR	Ref124
PegIFN alpha monotherapy	48 weeks	104.21 (copies/ml)*	normalized level of ALT in 1/8 patients	102.11	NR	ref 124
Myrcludex B (2 mg, 5 mg, or 10 mg plus tenofovir (245 mg/day) vs tenofovir (245 mg/day) monotherapy	24 weeks	NR	normalized levels of ALT in 42.8% of patients (2 mg), 50% (5 mg), 40%(10mg), or 6.6% (tenofovir only)	reduction of 1.75 log10 IU/ml^ (2 mg) reduction of 1.60 log10 IU/ml^ (5 mg) reduction of 2.70 log10 IU/ml^(10mg) 0.015 log 10 IU/ml^ (tenofovir)	No change in HBsAg	ref125 n=120
Intravenous REP2139-Ca (500 mg/week) for 15 weeks, then intravenous REP2139-CA (500 mg/week plus PegIFN alpha (180 μg)for 15 weeks, then PegIFN alpha for 33 weeks	63 weeks	2.7×104to 2.3×107 IU/ml	normalized level of ALT in 7/12 patients	Undetectable in 7/12 patients after removal of all antiviral therapy	HBsAg reduction >5 log in 6 with REP monotherapy	Refs134–136 n=12
Lonafarnib (100 mg twice daily) vs lonafarnib (200 mg twice daily) vs placebo	4 weeks	median 9.27 ×105 IU/ml	no change	0.73 log10 IU/mL* vs 1.54 log10 IU/ml*	No change	ref141 n=14
Lonafarnib monotherapy (100–300 mg, 2 or 3 times per day)	12 weeks	5.19–5.90 log10 IU/ml*	mean decrease in level of ALT, from 107 to 56 U/L in all groups at week 4	+0.03 log10 IU/ml* (200 mg twice daily) decrease of 1.78 log10 IU/ml* (300 mg twice daily) decrease of 1.31 log10 IU/ml* (100 mg 3 times per day)	No change	ref144 n=15
lonafarnib (100–300 mg, twice daily) with PegIFN alpha 180 μg/w	8 weeks	5.36–6.53 log10 IU/ml*	NR	decrease of 2.97 log10 IU/ml* (100 mg twice daily)	No change	ref 144
lonafarnib 100 mg twice daily plus ritonavir (100 mg/day)	8 weeks	6.56 log10 IU/ml*	normalized level of ALT in 3/3 patients by week 4	decrease of 3.19 log10 IU/ml*	No change	ref 144
lonafarnib 25 or 50 mg plus ritonavir (100 mg) twice daily, with or without PegIFN alpha or lonafarnib (75 mg or more) plus ritonavir (100 mg) twice daily	12 or 24 weeks	NR	normalized level of ALT in 9/15 patients at week 24	1.74 log10 IU/ml* (lonafarnib 25 mg plus ritonavir) decrease of 5.57 log 10 IU/ml* (lonafarnib 25 mg plus ritonavir and PegIFN)	NR	ref145 n=55
Lonafarnib (50 mg, 75 mg, or 100 mg) plus ritonavir (100 mg) daily vs placebo	12 or 24 weeks	4.58 log10 IU/ml^	normalized level of ALT in 10 of 21 patients	0.18 to-3.70 log10 IU/ml^ depending on virus kinetics	NR	ref146 n=21
Dose escalation starting lonafarnib (50 mg twice daily) plus ritonavir (100 mg) twice daily vs lonafarnib (100 mg) twice daily plus ritonavir (100 mg) twice daily	24 weeks	4.60 log10 IU/ml*	normalized level of ALT in 8 of 15 patients	1.58 log10 IU/ml*	NR	ref147 n=15

Symbols:

^*, Mean;

^{^}, Median

Abbreviations: IU, international units; NR, not reported

As previously observed, PegIFN lambda therapy was associated with transient increases in levels of transaminases and bilirubin in some patients. Although these increases might reflect a beneficial antiviral response, levels resolved with dose reduction and were not associated with hepatic decompensation. If the study finds that PegIFN lambda produces comparable rates of response to PegIFN alpha, PegIFN lambda increased tolerability could make it an attractive option for treatment of HDV infection—either as a monotherapy or in combination with other experimental therapies.

Inhibitors of virus entry

HBV and HDV bind to the solute carrier family 10 member 1 (SLC10A1 or NTCP) and enter hepatocytes, so inhibitors of this protein are being developed as therapeutics (Figure 4). The 48 N-terminal amino acids of the preS1 or L-HBsAg envelope protein are required for receptor binding¹¹⁹. Myrcludex B, a myristoylated lipopeptide derived from the pre-S1 domain of the HBV envelope, inhibits binding of HBV and HDV to NTCP and virus entry into hepatocytes in culture and in mice with humanized livers.^{61, 120,121}.

Myrcludex B has been tested in clinical trials. A phase 1b/2a trial found a decrease of at least a 1 log₁₀ in HDV RNA in patients who received Myrcludex B for 24 weeks¹²³. Twenty-four patients with detectable HDV RNA in serum were given daily subcutaneous injections of Myrcludex B, alone or in combination with PegIFN alpha and compared to PegIFN alpha alone for 24 weeks. A reported interim analysis at the end of 24 weeks of treatment demonstrated no significant changes in HBsAg levels, the study’s primary endpoint. However, HDV RNA declines of 1.67 (Myrcludex B group), 2.6 (Myrcludex B plus PegIFN alpha group), and 2.2 log₁₀ (PegIFN alpha group) were observed (Table 2). Two subjects in the Myrcludex B group, 5 in the Myrcludex B plus PegIFN group and 2 in the PegIFN group achieved serum HDV RNA levels below the lower limit of quantification during the reported interim analysis.¹²⁴ Serum HBV DNA levels decreased to near or below-detectable levels in all patients in the Myrcludex B plus PegIFN group, whereas no changes were observed in patients from the other groups of this study.¹²⁴ After 24 weeks of therapy, patients continued to receive PegIFN alpha monotherapy for a total of 48 weeks, which did not provide additional benefits, with response rates similar to those described for PegIFN alpha monotherapy, and with rebound following cessation of therapy. These findings indicate that Myrcludex B-mediated inhibition of viral entry can have inhibitory activity against HDV in humans. However, higher doses and longer treatment durations may be required, along with other drugs, for virus eradication. Additionally, pre-treatment levels of HDV RNA in participants were relatively low (1×10³–1s10⁵ copies/ml) and it is unclear whether similar results can be achieved in patients with higher serum levels of HDV RNA.

More recently, end of study results were reported from a multicenter, open-label phase 2 trial to assess the safety and efficacy of Myrcludex B in combination with tenofovir in patients with chronic HDV (Table 2).¹²⁵ In this study, 120 patients received tenofovir (245 mg/day) and were randomly assigned to groups given tenofovir monotherapy or once-daily subcutaneous injections of Myrcludex B (2 mg, 5 mg, or 10 mg) for 24 weeks, followed by 24 weeks of tenofovir monotherapy. The primary endpoint was a reduction in HDV RNA by 2 logs, or negative results from assays for HDV RNA in serum. At the end of therapy, median decreases in levels of HDV in the groups that received Myrcludex B ranged from 1.6 log to 2.7 logs; the highest dose (10 mg) produced the greatest decrease in median serum level of HDV RNA. Levels of ALT normalized in as many as 50% of patients. At 12 weeks of follow up, HDV RNA relapse occurred in as many as 80% of the subjects who responded to Myrcludex B. Notably, levels of bile acids were reported to increase during therapy. This finding was reproduced in healthy volunteers—plasma bile acid exposure increased 19.2-fold, without signs of cholestasis.¹²⁶ However, further studies of Myrcludex B are needed in patients with chronic HDV infection.

Inhibitors of HBsAg secretion

Nucleic acid polymers (NAPs) are phoshorothioated oligonucleotides with activities against a variety of infectious agents. Although their specific mechanisms are not understood, their activity appears to be dependent on polymer length and hydrophobicity.¹²⁷ NAPs have inhibitory activity against herpes simplex virus^{128, 129}, arenaviruses¹³⁰, and HCV.¹³¹ In ducks infected with the duck HBV, 28 days administration of the NAP REP2055 reduced serum levels of duck HBsAg and duck HBV DNA and increased anti-duck HBs antibodies; levels of duck HBsAg and duck HBV DNA remained at low or undetectable levels in serum through 16 weeks after administration.¹³² (see Figure 3). REP 2055 was found to have tolerability issues early-phase trails, so REP2139 replaced REP2055 and was studied in patients with HBeAg-positive chronic HBV infection.¹³³ In a phase 2 study, 3 of 12 patients given REP2139 had decreases in serum levels of HBsAg to below the limit of quantitation during 20–35 weeks of weekly intravenous infusion, resulting in a decrease in HBV DNA of 4–6 log₁₀. ¹³³

Figure 3. HDV Life Cycle and Therapeutic Targets.

Myrcludex-B attempts was designed to block entry of HDV into hepatocytes (A). REP2055 and REP2139-Ca were designed to inhibit HBsAg secretion (B). Lonafarnib inhibits HDV prenylation, required for virus packaging and secretion (C).

These encouraging results led to a small cohort study in patients with HBV and HDV coinfection. Patients were given weekly intravenous REP2139-Ca (500 mg) for 15 weeks, followed 15 weeks of REP2139-Ca (250 mg) combined with PegIFN alpha, and finally, PegIFN alpha monotherapy for 33 weeks (NCT02233075)(see Table 2).¹³⁴ In this study, 4 of 12 patients had decreases in serum levels of HBsAg (as much as 5 log₁₀ by week 15) and HDV RNA (below the limit of detection)¹³⁵. After discontinuation of REP1239-Ca and continuation of pegIFN, levels of HbsAg rebounded in 6 patients. However, after discontinuation of therapy, 5 of 12 patients had increases in levels of HDV RNA, with 3 of the 5 returning to baseline levels.¹³⁴. A long-term follow-up study of these patients is underway to assess their response to maintenance therapy.¹³⁶

Although these results are promising, little is known about the safety of long-term administration of NAPs. Given the mechanism of action of NAPs against HBV, with inhibition of HBsAg secretion in animals, this raises the question of whether long-term administration would increase the risk of HCC.^{132, 137}

Virus assembly and packaging inhibitors

The HDV lifecycle involves the covalent addition of prenyl lipids to proteins (prenylation).¹³⁸ Prenylation promotes membrane association of proteins and also mediates interactions between proteins.¹³⁸ Agents that inhibit the covalent addition of a farnesyl prenyl lipid group to the C-terminus of HDV large delta antigen disrupt its ability to interact with and form secreted particles with HBsAg (Figure 3).^{23, 139} Inhibition of farnesyltransferase inhibits secretion of virions and inhibit HDV in cultured hepatocytes and in mice.^{23, 140} The prenylation inhibitor lonafarnib was tested in patients with HDV infection at doses of 100 mg and 200 mg twice daily for 28 days, compared with placebo (Table 2).¹⁴¹ Both groups given lonafarnib had significant reductions in serum HDV RNA compared with patients given placebo. Patients in the 100 mg group had a mean decrease in HDV RNA of 0.73 log₁₀ IU/mL whereas patients in the 200 mg group had a mean decrease of 1.54 log₁₀ IU/mL. In aggregate, serum concentrations of lonafarnib correlated with change in HDV RNA (r²=0.78). However, patients given 200 mg had gastrointestinal side effects, including nausea, diarrhea, and weight loss. No patients were negative for HDV RNA during the 28-day study period, and HDV RNA levels returned to baseline levels by the end of the follow-up period.

Lonafarnib is metabolized by the cytochrome P450 family 3 subfamily A member 4 (CYP3A4), so researchers investigated whether addition of low-dose ritonavir (an inhibitor of CYP3A4) could increase the bioavailability and efficacy of lonafarnib. The addition of ritonavir to lonafarnib might allow for patients to be given lower doses of lonafarnib, which would reduce gastrointestinal side effects, and increase the amount of lonafarnib absorbed into the blood, increasing its pharmacokinetic and tolerability parameters. Ritonavir is given with HIV combination therapies for this purpose.^{142, 143} The combination of lonafarnib and ritonavir was tested in 4 phase 2 studies of patients with HDV infection (called the LOWR HDV studies).

In the open-label LOWR HDV-1 study, 15 patients were given lonafarnib monotherapy (100–300 mg) 2 or 3 times daily for up to 12 weeks, lonafarnib (100–300 mg) twice daily with PegIFN alpha (80 mcg, weekly) for 8 weeks, or lonafarnib (100 mg, twice daily) in combination with ritonavir (100 mg daily) for 8 weeks.¹⁴⁴ This study found that lonafarnib was safe in patients with HDV infection, for up to 12 weeks, and that the addition of ritonavir increased serum levels of lonafarnib, even though patients were given lower doses. The combination of lonafarnib and ritonavir reduced the mean level of HDV RNA by 2.4 log₁₀ at week 4 of administration and by 3.2 log₁₀ at week 8. The effects of lonafarnib were dose dependent—patients with the highest levels of serum lonafarnib had the greatest log₁₀ decreases in HDV RNA (r=0.68). The study also found the combination of lonafarnib (100 mg twice daily) with PegIFN alpha (180 mcg weekly) to be a feasible combination for future studies, whereas higher doses would have reduced tolerability. No subjects in this study tested negative for HDV RNA PCR, but 1 patient had a post-treatment transient increase in ALT, and during the follow-up period tested negative for HDV RNA in the PCR assay and had normalization of ALT. Sequencing analyses of HDV isolates found no evidence for resistance variants.

In the open-label, dose-optimization LOWR HDV-2 study, 55 patients were given combinations of lonafarnib with ritonavir, with or without PegIFN alpha, for 12, 24, or 48 weeks.¹⁴⁵ Patients were assigned to high-dose groups (given doses of lonafarnib ranging from 75 mg, twice daily or more, plus ritonavir), low-dose groups (given 25 or 50 mg lonafarnib twice daily with ritonavir, 100 mg, twice daily), or low-dose combination groups (given 25 or 50 mg lonafarnib twice daily with ritonavir, 100 mg, twice daily and PegIFN alpha, 180 mcg, weekly). Because responses in the high-dose groups were comparable to those in the low-dose groups, the study was extended beyond 12 weeks for only the low-dose groups. At 24 weeks, 6/12 patients (50%) who received lonafarnib (50 mg) had levels of HDV RNA below limit of quantification or a decrease of 2 log₁₀ or more (defined as responders). Patients given PegIFN alpha with lonafarnib (25 or 50 mg twice daily plus ritonavir) had the highest proportion of responders: 8 of 9 patients (89%) had levels of HDV RNA below limit of quantification or a decrease of 2 log₁₀ or more week 24. Importantly, 60% and 88% of patients in the all-oral and PegIFN alpha combination groups, respectively, had normalized levels of ALT at week 24.5, of 5 patients (100%) with low baseline viral load (4 log₁₀ IU/ml or lower) were responders to all-oral lonafarnib (50 mg and 100 mg ritonavir, twice daily). Adverse events for these lonafarnib (25 mg or 50 mg) regimens were predominantly mild to moderate. Again, as observed in LOWR HDV-1, several patients had post-treatment transient increases in levels of ALT, but later tested negative for HDV RNA in the PCR assays, and levels of ALT normalized.

The LOWR HDV-3 and 4 studies have been completed with end of study reports presented. The LOWR HDV-3 study was a randomized, double-blind study in which 21 patients received a daily all-oral combination of lonafarnib and ritonavir; some were assigned to groups given 24 weeks of lonafarnib and ritonavir and others to groups given 12 weeks of placebo and then 12 weeks of lonafarnib and ritonavir.¹⁴⁶ Patients were randomly assigned to groups given lonafarnib (50, 75, or 100 mg) with ritonavir (100 mg) once daily for 24 weeks (n=12) or 12 weeks of placebo followed by lonafarnib (50, 75, or 100 mg) once daily for 12 weeks (n=9). After 12 weeks, the median decrease in HDV RNA from baseline was 1.6 log₁₀ IU/mL in the lonafarnib (50 mg) group, 1.33 log₁₀ in the lonafarnib (75 mg) group, and 0.83 log₁₀ in the lonafarnib (100 mg) group. During this study, 6 patients achieved a decrease in serum level of HDV RNA of 2 log or more; HDV RNA levels became undetectable in 1 subject, below 14 IU/ml in 3 subjects, and <250 IU/ml in 2 subjects. Levels of ALT normalized in 4 of 6 subjects, and in 47% of the entire cohort. The combination was safe and tolerable for the 6 months of administration.

The LOWR HDV-4 study was an open-label, dose-escalation study in which 15 patients with chronic HDV infection were given lonafarnib with ritonavir for 24 weeks, to investigate the effects of rapid step-wise increases of lonafarnib to high doses.¹⁴⁷ All patients received a baseline dose of lonafarnib (50 mg) twice daily with ritonavir (100 mg) twice daily. If this was tolerated, the dose of lonafarnib as increased to 75 mg twice daily after 4 weeks, followed by an increase to 100 mg twice daily after another 1 week, if tolerated. Based on the end of study reports, lonafarnib dose escalation to 100 mg twice daily was possible in 10 of 15 patients (66%), and maintenance doses of 100 mg until the end of dosing were possible for 5 of 15 patients (33%). Interestingly, of these 5 patients, 1 tested negative for HDV RNA in the PCR assay and another had a decrease in HDV RNA to the lower limit of quantification. At the end of treatment, the mean decrease in HDV RNA for all patients was 1.7 log₁₀ IU/mL. Levels of ALT normalized in 53% of patients.

Future Directions

HDV infection is a worldwide problem in all age groups and results in the most rapidly progressive form of chronic viral hepatitis. There are knowledge gaps related to worldwide and US prevalence, as a result of suboptimal HDV screening. Although clinical outcomes, including HCC and hepatic decompensation, appear to be more severe in patients with HBV and HDV co-infection than in patients with HBV mono-infection, HDV is the only chronic hepatotropic viral infection without a satisfactory therapy. With advances in our understanding of the HDV life cycle, strategies to target hepatocyte entry, virus prenylation and assembly, and increase the anti-HDV immune system are being developed and tested. These agents might be successful as monotherapies or work in combination with other treatments for viral hepatitis. Nevertheless, there is great interest in eradicating a virus that was discovered more than 40 years ago and curing this devastating disease.

Abbreviations:

HDV

hepatitis delta virus

HBV

hepatitis B virus

HBsAg

hepatitis B surface antigen

RNA

ribonucleic acid

HCC

hepatocellular carcinoma

US

United States

SHDAg

small hepatitis D antigen

LHDAg

large hepatitis D antigen

DNA

deoxyribonucleic acid

VA

veterans administration

IDU

injection drug user

ALT

alanine aminotransferase

EOT

end of treatment

HCV

hepatitis C virus

NAP

nucleic acid polymer

HSV

herpes simplex virus