Review article: chronic hepatitis B – anti-viral or immunomodulatory therapy?

Authors


  • This commissioned review article was subject to full peer-review.

Prof. H. L. A. Janssen, Department of Gastroenterology and Hepatology, Erasmus MC, University Medical Center Rotterdam, ‘s Gravendijkwal 230, Room Ha 204, Rotterdam 3015 CE, The Netherlands.
E-mail: h.janssen@erasmusmc.nl

Abstract

Aliment Pharmacol Ther 2011; 33: 501–513

Summary

Background  First-line treatment options for chronic hepatitis B (CHB) consist of nucleos(t)ide analogues with a high barrier to resistance (entecavir and tenofovir) or the immunomodulatory agent peginterferon (PEG-IFN). The optimal choice for individual patients remains controversial.

Aim  To review treatment options for CHB, with a focus on deciding between prolonged nucleos(t)ide analogue therapy or a finite course of PEG-IFN.

Methods  A comprehensive literature search was undertaken.

Results  Long-lasting, treatment-maintained suppression of hepatitis B virus (HBV) DNA without resistance is achievable in most patients by entecavir or tenofovir. A sustained off-treatment response is, however, unlikely and long-term therapy must be anticipated. PEG-IFN offers a higher rate of sustained response in a subgroup of patients, but is frequently complicated by side effects. Pre-treatment predictors of response, including HBV genotype, alanine aminotransferase and HBV DNA levels, aid in selecting patients for PEG-IFN therapy. Furthermore, on-treatment markers such as quantitative hepatitis B surface antigen may be applied to identify nonresponders early during the PEG-IFN treatment course, thereby preventing unnecessary treatment.

Conclusions  Both nucleos(t)ide analogues and PEG-IFN can be prescribed as first-line treatment options for CHB. However, PEG-IFN should only be considered for patients with a high chance of response based on pre-treatment and on-treatment factors.

Introduction

Chronic hepatitis B (CHB) remains a major cause of liver disease worldwide. Although the global prevalence has fallen since the introduction of effective vaccination programmes, there are still more than 350 million people who are chronically infected with the hepatitis B virus (HBV).1 Progression of HBV-related liver disease to cirrhosis, hepatic decompensation and hepatocellular carcinoma (HCC) is estimated to result in 0.5–1.2 million annual deaths.1

The immunomodulatory agent interferon (IFN)-alfa has been a mainstay in the treatment of CHB since it was licensed for this indication in the early 1990s. In the last decade, great strides have been made in the treatment of CHB with the introduction of nucleos(t)ide analogues (NA) and pegylated forms of IFN (PEG-IFN). The increasing number of treatment options has added to the complexity of anti-viral therapy for CHB, leading to the development of multiple international practice guidelines in recent years.2–4 All of these guidelines support both NA and PEG-IFN as first-line treatment options, but the optimal choice for individual patients remains controversial. This review will focus on deciding between NA and PEG-IFN therapy.

Management of chronic hepatitis B

Phases of infection

Patients with CHB may present in any one of four, not necessarily sequential, phases of infection.5 During the immunotolerant phase, which is common among perinatally infected patients, the hepatitis B e antigen (HBeAg) is detectable and the serum HBV DNA level is high (>100 000 copies/mL or >20 000 IU/mL), whereas the serum alanine aminotransferase (ALT) concentration is within the normal range and liver histology shows minimal inflammation. In the immunoactive phase, the host immune response results in a decline in HBV DNA level, elevated ALT level and hepatic inflammation on liver biopsy. During this phase, HBeAg loss and seroconversion to anti-HBe can occur. HBeAg seroconversion is often followed by the inactive carrier state characterised by a low serum HBV DNA level (<10 000 copies/mL or <2000 IU/mL) and normalisation of ALT. However, in a significant proportion of HBeAg-negative patients, viral replication recurs or persists at higher levels, resulting in HBeAg-negative CHB. This phase of the infection develops through presence of viral strains harbouring mutations in the precore or basal-core promoter region that reduce or abolish the expression of HBeAg.6

Patients eligible for anti-viral treatment

There continues to be a debate as to when anti-viral treatment should be initiated in patients with precirrhotic HBV-related liver disease.7 The guidelines of the American Association for the Study of Liver Diseases (AASLD) and those of the Asian-Pacific Association for the Study of the Liver (APASL) advocate treatment for patients who remain HBeAg positive with serum HBV DNA levels >20 000 IU/mL in combination with persistently elevated ALT levels (>2 times the upper limit of normal (ULN) over a 3- to 6-month period).3, 4 According to the AASLD guidelines, the same HBV DNA and ALT criteria apply to HBeAg-negative patients, but it is emphasised that treatment should also be considered for these patients if the HBV DNA concentration is >2000 IU/mL combined with a liver biopsy showing moderate-to-severe necroinflammation and/or fibrosis.4 The APASL guidelines recommend to consider anti-viral treatment for HBeAg-negative patients with an HBV DNA level >2000 IU/mL and ALT levels persistently >2 × ULN.3 According to the 2009 HBV guidelines of the European Association for the Study of the Liver (EASL), anti-viral therapy should be considered, irrespective of HBeAg status, if HBV DNA levels are >2000 IU/mL, ALT levels are >1 × ULN and if a liver biopsy shows moderate-to-severe necroinflammation and/or fibrosis.2

Treatment goals

As complete eradication of HBV is rarely achieved due to persistence of covalently closed circular DNA (cccDNA) in host hepatocytes,8 the main goal of anti-viral treatment is to prevent progression to cirrhosis, hepatic decompensation and HCC.9 As these events develop only after decades of infection, clinical studies have focused on various short-term, surrogate outcomes to assess treatment efficacy.4, 9 The most widely used surrogate markers include suppression of HBV DNA levels and HBeAg loss with or without appearance of anti-HBe in HBeAg-positive patients (virologic response), normalisation of serum ALT (biochemical response) and improvement of liver histology (histological response).4

Parameters used to assess treatment response differ between NA and PEG-IFN therapy because of their different modes of action. NA suppress viral replication by competitively inhibiting viral polymerase activity, whereas PEG-IFN has both immunomodulatory and weak anti-viral properties.10, 11 During NA therapy, maintained suppression of HBV DNA is the main goal given the risk of developing anti-viral resistance associated with detectable HBV DNA levels.12 The main goal of PEG-IFN therapy is to achieve immunological control over HBV resulting in a sustained response after a finite treatment course.

In HBeAg-positive patients treated with PEG-IFN, HBeAg seroconversion is frequently used as a primary endpoint as it is associated with increased survival and a reduced risk of developing HCC.13–15 In contrast, suppression of serum HBV DNA below the detection limit of a sensitive polymerase chain reaction (PCR) assay is often applied for NA therapy in HBeAg-positive CHB. In HBeAg-negative patients, suppression of HBV DNA levels in combination with normalisation of ALT is considered the most important treatment goal. Although HBV DNA undetectability is pursued during NA treatment, an HBV DNA level <2000 IU/mL (∼10 000 copies/mL) is often used to define a sustained off-treatment response to PEG-IFN therapy.2 Large population-based studies have shown that an HBV DNA concentration below this threshold is associated with a low risk of cirrhosis and HCC.13, 16, 17 Hepatitis B surface antigen (HBsAg) seroclearance is the ultimate surrogate endpoint as it is associated with improved outcomes, provided that it occurs at a younger age and precedes the development of cirrhosis.18

Approved therapies

Current treatment options can be summarised into NA therapy [lamivudine (LAM), adefovir (ADV), telbivudine (LdT), entecavir (ETV) and tenofovir (TDF)] and IFN-based therapy (conventional IFN and PEG-IFN), which have all been investigated in randomised clinical trials. Virologic and biochemical response rates for HBeAg-positive and HBeAg-negative CHB at 1 year of therapy are summarised in Table 1. These trials used different virologic assays and were not head-to-head comparisons for all the agents. Efficacy results of up to 5 years of continuous treatment are available for several NA. Furthermore, long-term off-treatment follow-up results for PEG-IFN have been reported.

Table 1.   Overview of response rates in HBeAg-positive and HBeAg-negative patients after 1 year of anti-viral therapy
AgentHBeAg positive (%)HBeAg negative (%)
HBeAg seroconversionHBV DNA undetectabilityNormal ALTHBV DNA undetectabilityNormal ALT
  1. These trials were not head-to-head comparisons and different virologic assays were used. HBV DNA undetectability: HBV DNA <300–400 copies/mL.

Peginterferon22–27 10–2534–396338
Lamivudine1836607271
Adefovir1221485172
Telbivudine2360778874
Entecavir2167689078
Tenofovir2176689376

Anti-viral agents

Low barrier to resistance

Lamivudine.  Lamivudine was the first NA approved for the treatment of CHB in the late 1990s. Early randomised controlled trials evaluated its use against placebo, whereas LAM has also served as control arm in trials with newer anti-viral agents. In the treatment of HBeAg-positive disease, HBeAg seroconversion occurred in 18% of patients and HBV DNA was suppressed to undetectable levels (<300 copies/mL) in 36% of patients at 1 year of therapy.19 Extended therapy resulted in increasing cumulative rates of HBeAg seroconversion, reaching 47% at 4 years 20–22. In HBeAg-negative CHB, HBV DNA undetectability (<300 copies/mL) was achieved in as much as 72% of patients after 1 year of LAM therapy.23 However, virologic response rates (HBV DNA <1000 copies/mL) progressively decreased during longer-term continuous LAM therapy to 34% at 4 years.24

Amelioration of necroinflammatory and fibrosis scores on liver biopsy was observed in patients with LAM induced viral suppression,25 and the risk of hepatic decompensation and HCC in patients with advanced liver disease (bridging fibrosis or cirrhosis) was reduced.26 However, LAM is not considered a first-line agent for CHB anymore because of its low barrier to resistance resulting in a high rate of drug-resistant mutations in the tyrosine–methionine–aspartate–aspartate (YMDD) motif of the HBV polymerase. LAM resistance occurred in up to 70% of patients after 5 years of therapy (Figure 1).2 The development of anti-viral resistance is usually followed by a rise in HBV DNA and reactivation of disease.12, 25, 27

Figure 1.

 Rates of anti-viral resistance in nucleos(t)ide analogue naïve patients treated with the five registered nucleos(t)ide analogues within randomised controlled trials and subsequent follow-up studies. The bars represent years of therapy. These trials used different virologic assays and were not head-to-head comparisons for all the agents.

Intermediate barrier to resistance

Adefovir.  Adefovir was the second agent licensed for the treatment of CHB. Treatment with ADV 10 mg daily for 1 year resulted in HBeAg seroconversion in 12% of HBeAg-positive patients and 21% achieved HBV DNA negativity by PCR (<400 copies/mL).28 At 5 years on ADV therapy, HBeAg seroconversion was observed in 48% of patients and 39% had an HBV DNA level <1000 copies/mL.29 After 1 year of ADV therapy in HBeAg-negative CHB, 51% of patients became negative for HBV DNA (<400 copies/mL).30 The percentage of patients receiving prolonged ADV therapy who had HBV DNA <1000 copies/mL increased to 67% at 5 years and improvement of histological findings was observed in >70% of patients.31, 32

The rate of selection for ADV-resistant HBV is lower compared with LAM: ADV-resistant mutations were not present after 1 year. However, following 5 years of ADV therapy-resistant mutations were detected in up to 29% of patients, which, in addition to the low anti-viral potency of ADV, precludes its use as a first-line agent (Figure 1).2

Telbivudine.  Telbivudine demonstrated significantly greater anti-viral efficacy compared with LAM for the treatment of both HBeAg-positive and HBeAg-negative CHB.33 In HBeAg-positive patients, HBeAg seroconversion was observed in 23% and 30% of patients after 1 and 2 years of LdT therapy respectively.33, 34 HBV DNA suppression (<300 copies/mL) was observed in 56% of patients at year 2.33 In HBeAg-negative patients, HBV DNA became undetectable in 88% of patients after 1 year of LdT.33 The corresponding rate at 2 years decreased to 82%.34

Despite the high potency of LdT, virologic response rates deteriorate over time due to the emergence of LdT resistance. LdT resistance mutations are cross-resistant with LAM and were found in up to 22% of patients after 2 years of therapy (Figure 1).2, 12 The role of LdT is therefore limited in the treatment of patients with CHB.

High barrier to resistance

Entecavir.  Entecavir showed greater anti-viral potency than LAM in the treatment of patients with CHB. In HBeAg-positive patients, HBV DNA was suppressed to below the limit of detection (300 copies/mL) in 67% of patients and HBeAg seroconversion occurred in 21% of patients after 1 year.19 Cumulative rates of HBV DNA undetectability increased to 80% and 94% at 2 and 5 years of ETV therapy respectively.35, 36 In patients with HBeAg-negative CHB, HBV DNA was undetectable in 90% of patients after 1 year of therapy.23 The efficacy of ETV in NA naïve patients was recently confirmed in two large investigator-initiated studies. A European cohort study showed that a virologic response (HBV DNA <400 copies/mL or 80 IU/mL) was achieved in 79% of patients even through a short-term follow-up.37 Preliminary results of an Italian study demonstrated that ETV therapy resulted in undetectable levels of HBV DNA in 91% and 97% of HBeAg-negative patients at 1 and 2 years respectively.38

Very low rates of ETV resistance have been reported in NA-naïve patients,2 but resistance developed in up to 53% of LAM-experienced patients after 5 years of treatment (Figure 1). Only one additional mutation in the HBV polymerase is required to develop resistance in the presence of LAM-resistant mutations.12, 39

Tenofovir.  Tenofovir is the most recently approved NA for the treatment of CHB after demonstrating superior anti-viral efficacy compared with ADV for both HBeAg-positive and HBeAg-negative disease.40 In the treatment of HBeAg-positive CHB, HBeAg seroconversion and HBV DNA undetectability (<400 copies/mL) were observed in 21% and 76% of patients at 1 year of therapy respectively.40 At 3 years of therapy, the rate of HBeAg seroconversion increased to 26% and 78% of patients had undetectable HBV DNA.41 In HBeAg-negative patients, HBV DNA undetectability was achieved in 93% of patients after 1 year of TDF therapy,40 and 88% after 3 years (intention-to-treat analysis).42 However, when considering all patients on treatment at year 3, 99% of patients had undetectable HBV DNA. Recently, a large investigator-initiated study showed that 79% of patients who failed previous NA treatment had HBV DNA levels <400 copies/mL after a median follow-up of almost 2 years. However, the presence of ADV, but not LAM, resistance impaired TDF efficacy.43

Furthermore, it was recently shown that TDF resistance is absent after 3 years of continuous therapy (Figure 1), although solid conclusions based on these data are hampered by the study protocol that allowed viremic patients to add emtricitabine at week 72.41, 42

Immunomodulatory agents

Interferon has been used for the treatment of CHB for almost three decades. Modification of IFN through the attachment of a polyethylene glycol (PEG) molecule improved pharmacokinetic and pharmacodynamic properties. PEG-IFN is administered once-weekly by subcutaneous injection, which results in a relatively continuous drug exposure during the dosing interval. Two types of PEG-IFN have been developed: PEG-IFN alfa-2a and PEG-IFN alfa-2b.44 PEG-IFN alfa-2a proved to be at least as effective as conventional IFN in HBeAg-positive patients, with a comparable tolerability profile.45

Peginterferon

The efficacy of PEG-IFN alfa-2a and PEG-IFN alfa-2b for HBeAg-positive CHB has been evaluated in two pivotal trials.46, 47 PEG-IFN alfa-2a monotherapy during 1 year resulted in HBeAg seroconversion in 27% of patients, whereas this was 22% in patients treated with PEG-IFN alfa-2b. Six months after treatment discontinuation, HBeAg seroconversion rates were 32% and 29% respectively. Sustained off-treatment viral suppression (serum HBV DNA <400 copies/mL) was achieved in 7–14% of patients. In both studies, PEG-IFN monotherapy was compared with PEG-IFN and LAM combination therapy. Although combination therapy was associated with more potent viral suppression during the treatment phase, sustained off-treatment response rates were comparable with monotherapy.46, 47 At 6 months after discontinuation of PEG-IFN therapy, HBsAg seroconversion occurred in 3–5% of HBeAg-positive patients.

Long-term follow-up (mean duration 3 years) results of 172 HBeAg-positive patients treated with PEG-IFN alfa-2b ± LAM showed that among initial responders (defined as HBeAg negativity 26 weeks after treatment) HBeAg clearance was sustained in 81% of cases and the rate of HBsAg loss increased to 30%. Overall, HBsAg loss was achieved in 11% of patients at the last visit.48

In HBeAg-negative patients, PEG-IFN alfa-2a monotherapy for 1 year resulted in a sustained off-treatment response (HBV DNA <20 000 copies/mL and normal ALT at 6 months after treatment) in 36% of patients.49 Serum HBV DNA was <400 copies/mL in 19% of patients and 4% had lost HBsAg at 6 months after treatment. In parallel with HBeAg-positive disease, addition of LAM to PEG-IFN was beneficial in terms of on-treatment viral suppression, but did not result in higher sustained off-treatment response rates.49 Although PEG-IFN may be given for <1 year for HBeAg-positive CHB without reducing its efficacy,45 extending therapy beyond 1 year may result in lower relapse rates after treatment discontinuation in HBeAg-negative disease. A recent study showed that a sustained off-treatment response to PEG-IFN alfa-2a, defined as HBV DNA ≤2000 IU/mL at 1 year after treatment discontinuation, was achieved in 31% of genotype D-infected HBeAg-negative patients treated for 2 years compared with 10% of those treated for 1 year (P = 0.01).50

Results of long-term follow-up studies have shown that the off-treatment sustainability of response to IFN-based therapy is lower in patients with HBeAg-negative compared with HBeAg-positive CHB. A virologic response, defined as HBV DNA <10 000 copies/mL at 6 months after PEG-IFN alfa-2a treatment, was durable in 43% of patients with an initial response after 3 years of follow-up. The rate of HBsAg loss increased to 9%.51

Safety

Assessment of safety of both NA and PEG-IFN therapy should consider their mode of action and typical duration of use. NA generally have favourable side-effect profiles. However, all NA theoretically pose a risk for most inhibit not only viral, but also human DNA polymerases.52 In addition, some agents inhibit human mitochondrial DNA polymerases that may result in a clinical syndrome including lactic acidosis, neuropathy and myopathy.52 Fortunately, most NA used for CHB have shown little adverse events during the first years of therapy. Adverse events did not occur more frequently during LAM treatment compared with placebo,53 and the side-effect profile of ETV was indistinguishable from that of LAM.19 In contrast, ADV is known to be nephrotoxic in up to one-third of patients.54 Renal toxicity has also been described in HIV-infected patients treated with TDF, necessitating creatinine monitoring especially during prolonged treatment.52 Nevertheless, TDF was well tolerated and safe during the first year of therapy in patients with CHB.40

One year of PEG-IFN therapy is associated with considerable side effects. The most frequently reported are flu-like syndrome, headache, myalgia, fatigue and local reactions at the site of injection.46, 47, 49 These symptoms typically present early during therapy, whereas neuropsychiatric side effects such as depression and agitation tend to present in the later stages.55 Other clinically relevant side effects, including anorexia, thyroid dysfunction and hepatitis flares, occur less frequently. Decompensated cirrhosis is an absolute contraindication, but PEG-IFN is generally well tolerated and effective in patients with early cirrhosis.56 PEG-IFN has myelosuppressive effects, but neutropenia and thrombocytopenia rarely result in clinically significant infections and bleedings.55 Of particular note is the recently described neuropathy syndrome described in patients treated with a combination of PEG-IFN and LdT, which precludes clinical use of this combination regimen.57 In general, NA show favourable side-effect profiles in comparison to PEG-IFN, but long-term safety data for the new agents are lacking.

Treatment strategies

When to use peginterferon therapy?

Both treatment modalities have substantial advantages and limitations (Table 2).58 PEG-IFN therapy should always be considered for HBeAg-positive CHB, because a sustained off-treatment response is achieved in approximately one-third of patients.46–48 Furthermore, HBsAg loss, which approximates clinical cure, is more likely to be achieved during PEG-IFN therapy compared with NA.48, 51 However, the clinical use of PEG-IFN is compromised by its side effects and costs.55 Selection of patients with the highest probability of achieving a response is therefore essential. This is even more critical in HBeAg-negative disease, where a sustained off-treatment response is achieved in not more than 20% of patients.51

Table 2.   Pros and cons of nucleos(t)ide analogues vs. peginterferon for the treatment of chronic hepatitis B
Nucleos(t)ide analoguesPeginterferon
Daily oral dosingWeekly subcutaneous injection
Well tolerated in the medium termFrequent side effects
Potent HBV DNA suppressionModest HBV DNA suppression
Limited increase in rate of HBsAg lossIncrease in rate of HBsAg loss
Response hardly durable after therapyResponse durable after therapy
Risk of anti-viral resistanceAbsence of anti-viral resistance
Long-term or indefinite therapy may be requiredFinite duration of therapy
Less expensive during first year, possibly equally or more costly after long-term therapyExpensive

Patients should ideally be selected for PEG-IFN therapy based on their individual probability of response. Pros and cons of PEG-IFN can then be weighed on an individual basis. In case of a low baseline probability of response to PEG-IFN, it should be advised to initiate NA therapy. In addition, NA are the treatment regimen of choice for several patient groups, such as patients presenting with hepatic decompensation, those receiving chemotherapy, immunocompromised patients and pregnant women with a very high level of viral replication (HBV DNA >109 copies/mL).59

Baseline prediction of response to peginterferon in HBeAg-positive patients.  In recent years, many reports demonstrated the importance of HBV genotypes for IFN-based therapy.60, 61 HBV has been classified into eight genotypes (A–H) defined by divergence of more than 8% in the entire genome.62 The main HBV genotypes (A–D) have a distinct distribution throughout the world, with a predominance of A and D in Europe and North America and B and C in Asia, and appear to have a strong effect on the outcome of HBV infection.63–65 The association between HBV genotypes and responsiveness to IFN was confirmed in our study, which pooled the data from the two large trials investigating the efficacy of PEG-IFN in HBeAg-positive CHB (n= 721).66 Baseline factors associated with an increased likelihood of a sustained off-treatment response (HBeAg loss and HBV DNA <2000 IU/mL 6 months after treatment) included genotype A, older age, female gender, higher serum ALT levels and lower serum HBV DNA levels (Table 3). A multivariable prediction model was constructed, which allowed for individual prediction of a sustained off-treatment response to PEG-IFN based on the aforementioned baseline data.66 The so-called PEG-IFN treatment index can be found at http://www.liver-gi.nl/peg-ifn, and can guide the clinician when considering patients for PEG-IFN therapy. General recommendations on whether to consider PEG-IFN therapy for patients with HBeAg-positive CHB, based on a predicted probability of response >30%, are shown in Table 4.66

Table 3.   Baseline predictors of response to peginterferon therapy
HBeAg-positive patientsHBeAg-negative patients
  1. Baseline parameters associated with higher response rates to 1 year of peginterferon therapy. Data for HBeAg-negative patients are from single studies and are unconfirmed.

Older ageYounger age
Female genderFemale gender
Lower HBV DNA level (≤9 log copies/mL)Lower HBV DNA level
Higher ALT level (≥2 × ULN)Higher ALT level
HBV genotype A 
Table 4.   Recommendations for considering HBeAg-positive patients for peginterferon
HBV genotypeConsider peginterferon in HBeAg-positive patients with:
  1. Recommendations are based on an average probability of a sustained off-treatment response (HBeAg loss and HBV DNA <2000 IU/mL 6 months after treatment) of more than 30%. Probabilities for individual patients may vary, and other characteristics (including age and gender) should be taken into consideration.

AEither high ALT (≥2 × ULN) or low HBV DNA (≤9 log copies/mL)
B or CBoth high ALT (≥2 × ULN) and low HBV DNA (≤9 log copies/mL)
DPeginterferon generally not recommended

Nonetheless, considerable uncertainty remains on the individual level, and additional predictors of response may augment the baseline prediction model. In parallel with HBV DNA, quantitative assays for HBeAg and HBsAg have become available.67 Fried et al.68 reported that patients with lower baseline HBeAg levels have a higher probability of sustained off-treatment HBeAg seroconversion. HBsAg, another HBV marker, is secreted from the hepatocyte during viral replication as part of the HBV nucleocapsid, or as part of non-infectious viral particles.67 Serum HBsAg levels correlate with the amount of intrahepatic cccDNA, at least in HBeAg-positive patients.69–72 However, baseline serum HBsAg levels were not associated with a sustained off-treatment response in the two large PEG-IFN studies for HBeAg-positive CHB.73, 74

Baseline prediction of response to peginterferon in HBeAg-negative patients.  Currently, data on baseline predictors of response to PEG-IFN in HBeAg-negative CHB are limited (Table 3). A post hoc analysis of HBeAg-negative patients participating in the registration trial of PEG-IFN alfa-2a found that younger age, female gender, higher baseline ALT and lower baseline HBV DNA levels were associated with a higher probability of achieving a sustained off-treatment response (HBV DNA levels <20 000 copies/mL combined with normal ALT at 6 months after treatment).75 Unfortunately, the authors did not provide tools for easy clinical application of their findings, and calculation of the probability of response for individual patients is therefore cumbersome. Moreover, several studies showed that serum HBsAg levels were comparable between responders and nonresponders to PEG-IFN therapy.76–78

Which nucleos(t)ide analogue to use?

Entecavir and TDF are potent NA with a high barrier to resistance and are therefore recommended as first-line treatment options by the practice guidelines of EASL and AASLD.2, 4 In support of these recommendations, a recent study demonstrated that ETV and TDF are the most effective treatment options for CHB during the first year of treatment.79 Furthermore, it was estimated that long-term liver-related mortality may be reduced by 80% if all patients with high HBV DNA and ALT levels are treated with NA with a high barrier to resistance, compared with only 47% if NA with a low barrier to resistance would be used.80 Nevertheless, both AASLD and EASL guidelines do not take into consideration the cost-effectiveness ratio of these agents. A recent study from Spain suggested that TDF has a superior cost-effectiveness profile compared with other NA including ETV, although the cost-effectiveness ratio would be similar in countries with similar costs for both agents.81 For less potent anti-viral agents, the so-called road map model may be applied to adapt therapy in patients at high risk for the development of resistance after a set duration of treatment (detectable HBV DNA at 24 weeks for NA with a low barrier to resistance).3, 82 Nevertheless, even patients with undetectable HBV DNA at 24 weeks of treatment with NA with a low barrier to resistance may develop resistance,83 which is a major drawback for using this approach.

Treatment adaptation

Nucleos(t)ide analogues.  Previous studies have shown that the development of anti-viral resistance is mainly related to the level of ongoing viral replication.83, 84 As long-term oral therapy appears to be necessary in a majority of patients treated with NA, treatment adaptation at an early stage in case of suboptimal suppression of HBV DNA levels is required to prevent anti-viral resistance, in particular when NA with a low barrier to resistance are used.85 In case treatment adaptation is required, a treatment strategy associated with minimal risk of the development of multidrug-resistant HBV by taking into account cross-resistance profiles should be adopted.12 Nevertheless, it is currently unknown whether the road map concept also applies to NA with a high barrier to resistance as evidence for the advantages of adapting treatment in patients with detectable serum HBV DNA levels during therapy with ETV or TDF is lacking.86

According to current guidelines for the treatment of CHB, finite treatment with NA may be a reasonable option in HBeAg-positive patients.2–4 Treatment may be stopped after the achievement of HBeAg seroconversion combined with undetectable HBV DNA and an additional 6–12 months of consolidation therapy.2–4 However, studies on the durability of HBeAg seroconversion during NA therapy have reported conflicting results.87, 88 A recent study conducted in our centre showed that no more than 31% of patients achieved a durable remission (HBeAg negativity and HBV DNA <10 000 copies/mL) after HBeAg seroconversion.89 Furthermore, among patients who discontinued NA therapy after HBeAg seroconversion and at least 6 months of consolidation therapy, only 22% of patients had a sustained response.89 Therefore, ongoing NA therapy, irrespective of the achievement of HBeAg seroconversion, appears to be necessary. NA therapy should thus at least be continued until the occurrence of HBsAg loss, both in HBeAg-positive and HBeAg-negative disease, although HBsAg loss is rarely observed during NA therapy.

Peginterferon.  Even in the face of careful selection of patients for PEG-IFN, a considerable number of patients fail to achieve a response. On-treatment predictors of response to PEG-IFN may guide clinicians in their decision of whether to continue PEG-IFN or switch to NA therapy in specific patients.

In HBeAg-positive disease, the usefulness of HBV DNA levels in prediction of response during the first months of PEG-IFN therapy is limited. Different patterns of decline were observed during PEG-IFN alfa-2b therapy and even patients with a late or post-treatment decline of HBV DNA had a considerable chance of achieving a response.90 High HBeAg levels after 24 weeks of PEG-IFN alfa-2a therapy had a somewhat greater negative predictive value (NPV) upon HBeAg seroconversion in comparison with HBV DNA (96% vs. 86%).68 Recent studies have explored the role of on-treatment HBsAg levels for prediction of response to PEG-IFN in HBeAg-positive CHB. A preliminary report showed that PEG-IFN alfa-2a treated patients with HBsAg levels <1500 IU/mL at week 12 had 51% chance of sustained off-treatment HBeAg seroconversion compared with only 16% of those with HBsAg levels >20 000 IU/mL.73 Unfortunately, a considerable proportion of responders would be lost if one would discontinue therapy in patients with HBsAg levels >20 000 IU/mL. We recently reported that HBeAg-positive patients treated with PEG-IFN alfa-2b who experienced no HBsAg decline from baseline at week 12 of the treatment course had only 3% chance (NPV 97%) of a sustained off-treatment response (HBeAg loss and HBV DNA <10 000 copies/mL at 26 weeks after treatment) and no chance of HBsAg loss (Figure 2a).74 These patients should therefore be advised to discontinue PEG-IFN therapy.

Figure 2.

 (a) Flowchart showing the chance of a sustained off-treatment response (HBeAg negativity and HBV DNA <10 000 copies/mL 26 weeks after treatment) based on the presence of any HBsAg decline at week 12 of peginterferon alfa-2b therapy for 52 weeks in HBeAg-positive patients.74 (b) Flowchart for the probability of a sustained off-treatment response (HBV DNA <10 000 copies/mL and normal ALT 24 weeks after treatment) based on (1) the presence of any HBsAg decline and (2) HBV DNA decline ≥2 log at week 12 of peginterferon alfa-2a therapy for 48 weeks in HBeAg-negative patients.76*SR, sustained response.

In parallel with HBeAg-positive disease, prediction of response to PEG-IFN using HBV DNA levels is difficult in HBeAg-negative patients.91 Recent studies have focused on the predictive value of on-treatment HBsAg levels, the only viral marker that remains detectable in HBeAg-negative patients with suppressed HBV DNA levels. Low HBsAg levels at the end of 1 year of PEG-IFN treatment were associated with the highest probability of achieving a sustained off-treatment response, suggesting that HBsAg decline during therapy may allow discrimination between responders and nonresponders.77 Indeed, high predictive values for on-treatment HBsAg declines at weeks 12 and 24 on sustained off-treatment response (HBV DNA <70 copies/mL at 24 weeks after treatment) were reported in a cohort of 48 patients treated with PEG-IFN alfa-2a for 1 year: only 10% of patients who did not achieve a 0.5-log decline in serum HBsAg from baseline to week 12 of therapy achieved a response (NPV 90%).78 We recently showed that on-treatment decline of HBsAg at week 12 alone was of limited value in prediction of sustained off-treatment response (HBV DNA <10 000 copies/mL and normal ALT 24 weeks after treatment).76 Nevertheless, it was possible to establish a stopping rule by combining early on-treatment declines of HBsAg and HBV DNA (Figure 2b). None of the patients (20% of the study population) without HBsAg decline combined with <2 log HBV DNA decline achieved a sustained off-treatment response (NPV 100%).76 A majority of patients in our study were Caucasians infected with genotypes D and A, so these findings need to be confirmed in patients with genotypes B and C.

Combination therapy

Currently available agents for CHB are licensed for use as monotherapies and also guidelines recommend to start with monotherapy.2, 4 Combination therapy could, however, offer advantages in the treatment of CHB, including a combination of immunomodulatory and anti-viral effects, synergistic anti-viral effects and a higher barrier towards resistance. Furthermore, combination therapy has improved treatment outcomes in many other chronic infections.92 Combining PEG-IFN with NA is the most appealing approach, but the addition of LAM, ADV and ribavirin to PEG-IFN therapy did not result in higher sustained response rates.46, 47, 49, 93, 94 New trials on a combination of PEG-IFN with ETV or TDF are therefore needed. In addition, studies exploring different ways of combining PEG-IFN with NA are indicated as most studies started and ended combination therapy at the same time. Pilot studies have suggested that lowering HBV DNA by NA before PEG-IFN therapy may improve outcomes, but contradictory results have been reported.95, 96

Studies on the efficacy of de novo combination therapy with NA are scarce and only combinations of NA that are no longer considered as first-line treatment were investigated.97–99 Although it was shown that combination therapy can delay the development of resistance,97 there are no data to support de novo combination therapy with NA that have a high barrier to resistance in NA-naïve patients. Moreover, a recent study in patients with incomplete viral suppression during therapy with ADV showed that TDF monotherapy and the combination of TDF and emtricitabine, which is structurally similar to LAM, had similar efficacy.100

Overall, there is insufficient evidence to recommend de novo combination therapy in CHB as first-line therapy. Nevertheless, current guidelines recommend that a combination of NA may be considered in patients at high risk of complications in the event of virologic breakthrough such as those with advanced liver disease.2, 4

Conclusions

The treatment of CHB has greatly improved over the past decade with the introduction of potent NA with a high barrier to resistance and PEG-IFN. The advantages and limitations of both treatment options should be considered when a patient has an indication to initiate anti-viral therapy. NA can be prescribed to all adult CHB patients in whom treatment is indicated and offer easy daily oral dosing, are well tolerated and NA with a high barrier to resistance can maintain suppression of viral replication for prolonged periods of time. However, an off-treatment sustained response is unlikely in a majority of patients. A sustained off-treatment response is more likely to be achieved in a subgroup of patients with a finite course of PEG-IFN therapy, but treatment is frequently complicated by the occurrence of side effects.

Both NA and PEG-IFN can be given as first-line treatment option for CHB. However, PEG-IFN should only be considered for patients with a high chance of response. Predictors of response at baseline, such as HBV genotype, ALT and HBV DNA levels, aid in selecting patients for PEG-IFN therapy, especially in HBeAg-positive disease. Furthermore, recent studies suggest that on-treatment viral markers, in particular quantitative HBV DNA and HBsAg, may be applied to identify patients who are unlikely to benefit from PEG-IFN early during the treatment course, thereby avoiding unnecessary treatment. Patients who are not eligible for PEG-IFN or who have a low probability of response to PEG-IFN based on baseline or on-treatment factors should be advised to initiate or switch to NA therapy.

Acknowledgements

Declaration of personal interests: Vincent Rijckborst and Milan J. Sonneveld have nothing to disclose. Prof. H. L. A. Janssen received grants from and is a consultant for Bristol Myers Squibb, Gilead Sciences, Novartis, Roche and Merck (Schering Plough). Declaration of funding interests: None.

Ancillary