New paradigms for the treatment of chronic hepatitis B


Professor Man-Fung Yuen, Department of Medicine, The University of Hong Kong, Queen Mary Hospital, Pokfulam Road, Hong Kong. Email:


The main goals of chronic hepatitis B treatment should be the long-term suppression of viral replication to minimize disease progression and the risk for the development of hepatocellular carcinoma. Treatment end-points, depending on surrogate markers alone, in particular hepatitis B e-antigen seroconversion, may not be ideal for patients who acquire the disease early in life.

Currently-available drugs include interferons and oral nucleoside/nucleotide analogs. Although interferon therapy provides a finite treatment period, a significant proportion of patients may not respond, and long-term outcome is inconclusive. Long-term efficacy has been demonstrated for both lamivudine and adefovir. However, prolonged nucleoside/nucleotide analog therapy is associated with the emergence of drug-resistant mutations. Therefore, nucleoside/nucleotide analogs with a high genetic barrier and potent antiviral activity, such as entecavir, should be used to reduce the chance of developing drug-resistant mutations. Drugs with a low genetic barrier, including lamivudine and telbivudine, should be used in conjunction with early testing for antiviral response. This can predict favorable outcomes in the long term. The early detection of drug-resistant mutations should prompt clinicians to either add or switch to another agent with a different drug-resistance profile. There are currently no treatment models in the use of combination or sequential therapy in treatment-naïve patients.

To date, long-term treatment appears to be the most effective option. Despite recent advances made with better understanding on the natural history of chronic hepatitis B infection and with newer antiviral drugs available, challenges remain with respect to treatment criteria, treatment end-points, and duration of treatment.


An estimated 400 million people worldwide are infected with chronic hepatitis B (CHB), constituting a major health problem.1 Approximately 1 million people die annually as a result of hepatitis B virus (HBV)-related liver cirrhosis and hepatocellular carcinoma (HCC). Over 70% of CHB patients are of Asian origin. The numbers of people with CHB in the USA and UK have been estimated to be approximately 1.25 million and 180 000, respectively.2

The age of acquisition of HBV plays an important part in determining the natural history of HBV infection, and is best illustrated by the difference observed between Asian and Caucasian patients. The majority of Asian patients acquire HBV either via vertical transmission or early horizontal transmission, that is, within the first few years after birth. This is in contrast to the majority of Caucasian patients who acquire the infection during adolescence or in early adulthood. The course of those infected early in life is characterized by a prolonged immunotolerant phase followed by a prolonged phase of immunoclearance, typically during the third and fourth decades of life. These patients will eventually undergo hepatitis B e-antigen (HBeAg) seroconversion with the development of antibodies to HBeAg (anti-HBe). Disease progression to symptomatic cirrhosis and HCC has been shown to occur in a proportion of patients after HBeAg seroconversion. In fact, the majority of these complications occur in patients who are anti-HBe positive. In contrast, the majority of patients infected during later childhood or adolescence will immediately enter the immunoclearance phase without going through the immunotolerant phase, and the disease usually becomes more quiescent after HBeAg seroconversion. The differences in natural history between these two groups of patients must be considered when considering CHB treatment.

Goals of therapy

The ideal goal of CHB therapy is the complete eradication of HBV. However, despite recent advances in treatment strategies with newer and more potent antiviral agents, complete eradication of HBV is still not possible. The persistence of highly-stable, covalently-closed circular DNA within the nuclei of hepatocytes provides an intracellular reservoir of HBV replication, leading to persistent infection. The virus may also integrate into the host genome. Therefore, the long-term goals of treatment should be the prevention of liver cirrhosis, liver failure, and HCC.

For the purpose of clinical trials, treatment end-points for CHB have focused on more short-term goals, such as normalization of serum alanine aminotransferase (ALT) levels, HBV–DNA suppression, HBeAg seroconversion, and improvement in liver histology. Recent treatment guidelines and algorithms have realized the importance of viral suppression in the prevention of disease progression and HCC.3,4 However, the treatment of CHB remains complex and is dependent on multiple factors. These include the biochemical, virological, and immunological profiles, as well as the underlying disease severity and its natural history.

Treatment choices

There are currently six antiviral agents available for the treatment of CHB: lamivudine, adefovir, entecavir, telbivudine, interferon (IFN)-α, and pegylated IFN (peg-IFN). Newer antiviral drugs, such as tenofovir, will become available, increasing the options available for CHB treatment. The choice of therapy should always take into consideration the antiviral efficacy, risk of developing drug resistance, long-term safety profile, methods of administration, and costs of the agent.3

IFN-α and peg-IFN

Previous meta-analyses have shown IFN-α to be effective in promoting hepatitis B surface antigen (HBsAg) and HBeAg seroconversion, and in suppressing HBV–DNA.5 Peg-IFN has largely surpassed standard IFN in CHB treatment. In the only head-to-head comparison from a phase II study, peg-IFN-α2a was superior to standard IFN in HBeAg clearance, HBV–DNA suppression, and the normalization of ALT. However, the standard IFN in this trial was given at a fixed dose of 4.5 MIU three times a week rather than the conventional doses of 5 MIU/m2 daily or 10 MIU/m2 three times a week.6

Compared with lamivudine, 48 weeks of peg-IFN-α2a (180 μg/week) resulted in a higher rate of HBeAg seroconversion (32% vs 19%, P < 0.001).7 The majority of recent studies have focused on combination or sequential therapy with lamivudine.8 In one small study from India, lowering viral load with lamivudine prior to IFN therapy has been shown to improve sustain viral response when compared to using IFN alone.9 However, most studies have not shown any additional benefit of adding lamivudine to IFN therapy in either HBeAg-positive or HBeAg-negative disease when assessed at 6 months after the cessation of therapies.7,10–14 It should be noted that in all these studies lamivudine was stopped after 48 weeks, irrespective of HBeAg seroconversion or the extent of viral suppression, a practice which is not advocated in any of the current guidelines, and which on its own may be hazardous for presaging post-treatment hepatitis flares. Although no additional antiviral efficacy is observed, there is evidence that combining IFN with lamivudine therapy decreases the development of lamivudine-resistant mutations.12,13

The long-term effectiveness of standard IFN has not been consistently shown. Long-term benefits, including preventing cirrhosis and HCC, were not observed in Japanese and Chinese patients.15,16 In a more recent retrospective study of 233 IFN-treated Taiwanese patients with pretreatment mean ALT levels greater than 175 U/L, IFN was shown to reduce HCC and cirrhosis in HBeAg-positive patients compared to untreated controls. However, a subgroup analysis comparing HBeAg non-seroconverters in the control group, non-seroconverters in the IFN group, seroconverters in the control group, and seroconverters in the IFN group, statistical differences were only observed between the first group versus the third and fourth groups. There were no differences among the latter three groups.17 Also a mean ALT level of 175 U/L at recruitment is an unusually high ALT level for Asian patients with CHB, so that the results obtained from this study can only apply to those patients with ALT levels of >150 U/L. Further results regarding peg-IFN with long off-treatment follow up will be awaited to determine its long-term efficacy.


Lamivudine was the first oral antiviral drug approved for the treatment of CHB. It is effective in inducing viral suppression, normalization of ALT, HBeAg seroconversion, resolving histological hepatitis, and decreasing the progression of liver fibrosis.18,19 The long-term benefits of lamivudine in reducing the complications of cirrhosis, including decompensation and HCC, have been shown in two long-term studies with precirrhotic/cirrhotic and non-cirrhotic patients.20,21 However, lamivudine is associated with high rates of viral resistance of approximately 50% after 3 years. After 8 years' treatment with lamivudine, the prevalence of genotypic resistance was 76%.21

Resistance to lamivudine occurs as a result of rtM204V or rtM204I mutations with or without concomitant rtL180M mutation of the HBV polymerase gene.22 It has been shown that the initial benefits conferred by lamivudine are reduced in patients who develop lamivudine-resistant mutations during long-term follow up.23 However, even among those with drug resistance, the outcome remains better than that for untreated patients.20,21 Caution must be taken when stopping therapy without replacement with another effective antiviral agent. There is the potential for hepatitis flares to occur after stopping treatment due to rapid rebound replication of wild-type virus. Both adefovir and entecavir are effective against lamivudine-resistant CHB; the factors that need to be considered in making the choice of second agent are discussed later.


Adefovir dipivoxil has been shown to be effective in both HBeAg-positive and HBeAg-negative CHB, as well as lamivudine-resistant mutations.24–26 Although relatively slow in action, the long-term efficacy of adefovir in maintaining viral suppression, and biochemical and histological responses has been shown in patients treated for 5 years.27 With newer and more potent antiviral agents now available, adefovir is mainly used in patients who have developed resistance to lamivudine or telbivudine. Some studies have shown that adefovir monotherapy in lamivudine-resistant patients is as effective for suppressing HBV–DNA as combination therapy with lamivudine.28–30 Despite this, several studies, including a large randomized controlled trial,31 report a substantially lower rate of resistance to adefovir when treatment is continued in combination lamivudine versus “switching” to adefovir in patients with lamivudine resistance.31–34 In general, the recommendation for lamivudine-resistant patients would be adding adefovir as soon as genotypic resistance is detected. This strategy will achieve the best outcome in terms of minimizing adefovir resistance and hence maintaining HBV–DNA suppression in the long term, thereby providing the sought-after surrogate condition for preventing adverse disease outcomes.

Adefovir has a higher genetic barrier than lamivudine, meaning that the rates of resistance in treatment-naïve patients are lower. Thus after 5 years of follow up, 29% of HBeAg-negative patients developed genotypic resistance to adefovir.35 Mutations at rtA181V/T and rtN236T of the HBV polymerase gene are responsible for adefovir resistance.36 Adefovir-resistant HBV is sensitive to both entecavir and lamivudine.37


Entecavir is a carboxylic analog of guanosine, and the third oral antiviral agent approved for CHB treatment. In phase III trials for both HBeAg-positive and HBeAg-negative CHB, entecavir was superior to lamivudine in HBV–DNA reduction and in inducing histological improvement.38–40 Furthermore, no virological breakthrough from entecavir resistance has been observed after 2 years of treatment in treatment-naïve HBeAg-positive patients.41 In a dose-finding phase II trial, entecavir has been shown to be effective against tyrosine-methionine-aspartate-aspartate (YMDD) mutant strains of HBV, albeit at the higher daily dose of 1 mg instead of the recommended 0.5 mg daily dose for treatment-naïve patients.42 A subsequent phase III trial has shown significantly better histological, virological, and biochemical outcomes with entecavir compared to lamivudine in lamivudine-refractory patients.43

Resistance to entecavir in treatment-naïve patients is rare, occurring in 1.1% of patients after 4 years of therapy.44 However, in patients with pre-existing lamivudine-resistant mutations, there is a lower viral response rate, and a correspondingly higher rate of developing entecavir resistance, 39% after 4 years.45 The reason for the higher rate of resistance is because the rtM204V and rtL180M mutations that characterize lamivudine resistance are a prerequisite for subsequent (third or fourth) mutations that lead to the emergence of entecavir resistance. Thus rtM204V and rtL180M mutations are not sufficient by themselves to confer resistance to entecavir unless an extra mutation occurs at rtT184G, rtS202I, or rtM250V.45,46 As the corollary of this sequence of mutations, the pre-existence of lamivudine resistance predisposes patients to develop subsequent resistance to entecavir. For this reason, entecavir-switching therapy may be less optimal than adefovir add-on therapy for CHB associated with lamivudine resistance. However, studies of direct comparisons are required to establish this.


Telbivudine has been shown to be more potent than lamivudine against HBV.47 An open-label trial comparing telbivudine and adefovir also showed greater HBV–DNA suppression in patients treated with telbivudine than adefovir after 52 weeks.48 Despite its superior antiviral efficacy and lower resistance rate compared to lamivudine, telbivudine is still associated with higher resistance rates than adefovir or entecavir.49

Resistance to telbivudine occurs at the same mutation site responsible for lamivudine resistance. The rate of genotypic resistance after 2 years of telbivudine treatment is 22% and 8.6% among HBeAg-positive and HBeAg-negative patients, respectively.49 Compared with lamivudine, the lower resistance rate of telbivudine is partly because of the greater antiviral potency of telbivudine, but possibly also because only the M204I mutant is observed and not the M204V/L180M.47

Comparison of available treatments

There are few head-to-head comparisons of currently-available antiviral agents. A cross-study analysis of results at 1 year from 28 trials involving lamivudine, adefovir, and entecavir concluded that the antiviral efficacy of entecavir was superior to lamivudine, which in turn was superior to adefovir in nucleoside-naïve patients.50

HBeAg seroconversion

In HBeAg-positive patients, 12–27% of patients treated with HBV antiviral agents will seroconvert after 1 year of treatment, as shown in Figure 1.7,24,39,47 IFN therapy has been shown to have the highest rate of HBeAg seroconversion after 1 year of treatment. However, patients who do not seroconvert, that is, the majority of patients (approximately 70%), will require an alternative form of long-term antiviral therapy. After 2 years of therapy, HBeAg seroconversion rates are similar for all the currently-available antiviral agents, despite differences in their antiviral efficacy.

Figure 1.

Summary of hepatitis B e-antigen (HBeAg) seroconversion at 1 (a) and 2 years (b) in patients receiving different treatment. For pegylated interferon (peg-IFN)-treated patients, the HBeAg seroconversion rate at 2 years refers to 1 year of therapy followed by 1 year of follow up. Data are not head-to-head comparisons. Adv, adefovir; Ent, entecavir; Lam, lamivudine; Tel, telbivudine.

Normalization of ALT

In both HBeAg-positive and HBeAg-negative patients, the rate of ALT normalization at 1 year is approximately 38–39% for IFN and 48–78% for oral antiviral therapy, as shown in Figure 2.7,24,38,39,47,51

Figure 2.

Summary of alanine aminotransferase (ALT) normalization at 1 year in hepatitis B e-antigen (HBeAg)-positive (a) and HBeAg-negative (b) patients receiving different treatments. Data are not head-to-head comparisons. Adv, adefovir; Ent, entecavir; Lam, lamivudine; Peg-IFN, pegylated interferon; Tel, telbivudine.

Reduction in viral load

The 1-year rates of viral suppression between the different antiviral agents for HBeAg-positive and HBeAg-negative CHB are summarized in Figure 3. Approximately 21–67% of patients will have undetectable HBV–DNA after 1 year of treatment in HBeAg-positive patients, and 51–90% in HBeAg-negative patients.7,24,38,39,52

Figure 3.

Summary of HBV–DNA suppression at 1 year in hepatitis B e-antigen (HBeAg)-positive (a) and HBeAg-negative (b) patients receiving different treatments. Data are not head-to-head comparisons. Adv, adefovir; Ent, entecavir; Lam, lamivudine; Peg-IFN, pegylated interferon; Tel, telbivudine.

Viral resistance

The major setback in the treatment of CHB is the development of drug resistance. This is particularly important as the majority of CHB patients will require long-term therapy. Flares of hepatitis, liver decompensation, and death have been reported to occur in patients who develop viral resistance.53 However, there is no direct comparison of the risk of hepatitis flare, decompensation, and death between patients with drug-resistant HBV and those not receiving antiviral therapy. The rates of drug resistance have been described in earlier sections and are summarized in Figure 4.

Figure 4.

Summary of genotypic resistance rates of lamivudine, adefovir, telbivudine, and entecavir in treatment-naïve patients. Data are not head-to-head comparisons. Rebound due to resistance.

The development of drug resistance is a challenging problem because of its impact on further treatment. It has been shown that patients who develop lamivudine resistant mutations will have a higher rate of developing subsequent adefovir resistant mutations compared to those patients without lamivudine resistant mutations.54 Patients who have lamivudine-resistant HBV will also have a higher rate of developing subsequent entecavir resistance.42

Because of the adverse impact of drug-resistant HBV on the clinical outcome and on subsequent antiviral therapy, the risk of developing resistance should be considered prior to starting antiviral therapy. The high resistance rate associated with lamivudine limits its use as a first-line option with the availability of newer and more potent antiviral agents. However, lamivudine remains the least expensive oral antiviral agent with the longest and largest profile of safety data. Close monitoring of patients for early response to treatment may select those patients who will respond more favorably in the long term with a lower rate of drug resistance. This will be detailed in a later section. The recent development of a relational database combined with a HBV genome sequence analysis program may potentially allow physicians to individualize patient care based on the resistance profile.55

Treatment end-points

As complete eradication of HBV is currently not possible with available therapy, various end-points have been adopted as surrogate markers of successful treatment. These include HBeAg seroconversion, normalization of ALT, and HBV–DNA suppression. With better understanding of the natural history of CHB and the factors associated with disease progression, the treatment end-points are constantly evolving and being redefined.

HBeAg seroconversion

The occurrence of HBeAg seroconversion has been used as a marker of treatment success in many clinical trials of antiviral therapy, and is now incorporated in treatment guidelines for HBeAg-positive patients.3,4,56,57 This is based on an earlier understanding that HBeAg seroconversion resulted in low HBV–DNA levels, and normalization of ALT with the resolution of necro-inflammatory activity within the liver, the so-called residual phase of CHB infection. These patients have been defined as “healthy carriers”.58,59 This concept appears to be supported by the results of a study of 11 893 men in Taiwan, which found relative risks for developing HCC of 60 and 9.6 in HBeAg-positive and HBeAg-negative patients, respectively, compared to those not infected with HBV.60 However, HBeAg was determined only at the time of enrollment in this study. It is likely that a significant proportion of these patients had HBeAg seroconversion during the follow-up period.

More recent studies have shown that over 70% of CHB patients are HBeAg negative at the time of developing HCC.61,62 The median age of HBeAg seroconversion was 35 years in a recent study of 3233 Asian patients. The median age of patients with complications of cirrhosis was 57 years, and over 70% were HBeAg negative.62

Another consideration comes from previous studies which have shown high rates of reactivation and seroreversion with the cessation of antiviral therapy after treatment-induced HBeAg seroconversion, particular in older Asian patients. In one early study of 34 Korean patients, the cumulative relapse rate was close to 50% after 2 years of stopping lamivudine therapy.63 Another study of Indian patients reported relapse in 35% of patients during a median time of 6 months after stopping lamivudine.64 The HBeAg response appears to be more durable in Western (European) populations. Among 39 predominantly European patients in the USA who underwent lamivudine-induced HBeAg seroconversion, there was a durable response in 77% after a median follow-up period of 37 months.23 A recent study showed that the long-term outcome of treatment-induced HBeAg seroconversion is inferior to that of spontaneous HBeAg seroconversion, with significantly higher and faster rates of CHB reactivation.65

The evidence thus far shows that disease progression can occur after spontaneous HBeAg seroconversion in patients who acquired HBV at birth or during early childhood. Furthermore, there is a high chance of HBeAg seroreversion, and virological and biochemical relapses after HBeAg seroconversion associated with therapy among those with a childhood acquisition of HBV. As we have proposed recently,66,67 HBeAg seroconversion should be viewed as part of the natural history/progression of chronic HBV infection rather than as the sole treatment end-point.

HBV–DNA level

The serum HBV–DNA level has been shown to be important in both the development of liver cirrhosis and HCC, with or without underlying cirrhosis. Recent studies have shown that higher levels of HBV–DNA at study entry (and specifically at 30 years of age in the REVEAL study) are associated with the development of HCC independent of the HBeAg status and ALT levels.61,68 This finding holds true even in HBeAg-negative patients with normal ALT levels. It may be important to identify thresholds of the HBV–DNA level at various ages whereby patients may be deemed safe from the progression of disease and development of HCC. This strategy is likely to identify those at-risk patients who would benefit most from antiviral therapy.

Despite these considerations, there is no current level of HBV–DNA which is considered “safe” from disease progression or the development of HCC.61 In HBeAg-negative Chinese patients with active hepatitis, 45% may have HBV–DNA below 20 000 IU/mL.69 In the REVEAL study, 3653 CHB patients aged 30 years or older at onset were followed up for a median of 11.4 years, evaluating the relationship between viral load at study enrollment and HCC. The incidence of HCC increased with HBV–DNA levels in a dose–response relationship, rising from 108/100 000 person–years in patients with HBV–DNA <60 IU/mL to 1152/100 000 person–years in patients with HBV–DNA ≥200 000 IU/mL. The cut-off level of >2000 IU/mL was shown to be a strong risk predictor of HCC independent of HBeAg status, serum ALT, and underlying cirrhosis.68 In an update from the same study, patients with HBV–DNA <2000 IU/mL were still shown to be at a significantly higher risk of developing HCC when compared to uninfected patients.70 Even lower HBV–DNA levels have been associated with the development of HCC. In a study of 92 HCC patients, 15% had HBV–DNA levels less than 200 IU/mL.61

The evidence to date indicates that there is unlikely to be a safe threshold whereby cirrhosis and HCC do not occur. The association between occult HBV infection (as defined by negative HBsAg with positive HBV–DNA) and HCC would also suggest that there is no safe level, albeit the risk is much lower when the viral load is very low.71–75 Given the absence of a safe threshold, the optimal treatment goal should therefore be to suppress HBV–DNA to the lowest possible level.

ALT level

There has been an increasing focus on ALT levels in recent years, particularly with the definition of what should be defined as normal. The currently-used cut-off levels have been derived from apparently healthy populations which may have included patients with unknown underlying liver disease, particularly fatty liver. There is mounting evidence from modeling studies to indicate that the “safe” range of ALT is lower than the currently-defined levels. In a large prospective study of 142 055 people without known liver disease, the liver-related mortality was higher in patients with ALT levels of 0.5 × upper limit of normal (ULN) to the ULN compared to those persons with ALT levels below 0.5 × ULN. The optimal ALT cut-off for the prediction of liver disease was 30 U/L and 19 U/L in males and females, respectively.76,77 In Asian CHB patients, it has been shown that patients with ALT below half the ULN have the lowest risk of complications compared to those with 0.5 × ULN to the ULN.62

Patients who have undergone HBeAg seroconversion with subsequent normal ALT have been traditionally regarded as “healthy carriers” with no or minimal disease progression. However, in a follow up of 240 patients with normal ALT after HBeAg seroconversion, the cumulative probability of developing cirrhosis after 17 years was 13%. Both the age of HBeAg seroconversion and the presence of hepatitis flares were independent risk factors for the development of cirrhosis.78 A recent study of CHB patients showed that 37% of those with persistently normal ALT had significant fibrosis or inflammation on liver histology, and the majority of cases with fibrosis occurred in those with high–normal ALT.79 Even in HBeAg-negative CHB patients with persistently normal ALT, those with high–normal ALT (0.5 × ULN − ULN) had characteristics associated with adverse long-term outcomes compared to those with ALT <0.5 × ULN.80 In this particular study, the other adverse factors include male sex, higher HBV–DNA levels, and increasing age.80

The accumulated evidence from all the above studies indicates that the upper normal threshold for ALT is likely to be lower than the values that are currently used in clinical chemistry laboratories. Patients with ALT in the ULN are at risk of developing cirrhosis and therefore warrant ongoing disease monitoring. Furthermore, treating only those patients with ALT >2 × ULN, as indicated by most current guidelines, would exclude a significant proportion of patients who would be at risk of having or developing significant liver disease (cirrhosis and HCC) and who would therefore benefit from antiviral therapy.4

Candidates for therapy

With increasing availability of newer antiviral agents, there is a need to select patients who would benefit most from antiviral therapy. In HBeAg-positive patients, the ideal candidates for treatment would be those with a prolonged phase of immune clearance. A recent study of immunotolerant patients revealed minimal disease progression in these patients.81 Therefore, immunotolerant patients can be monitored, and treatment may not be required until active hepatitis occurs during the onset of loss of immunotolerance. Unfortunately, there are no specific criteria to define or identify the time at which immunotolerance ends and immune clearance starts. An elevated ALT level is used as a surrogate marker of inflammation and histological activity and as a marker for loss of immunotolerance. Whether this is an adequate marker remains to be determined. Nonetheless, it seems reasonable that patients with ALT levels between 1 and 2 × ULN should probably be treated. In patients with clinicopathological evidence of cirrhosis and residual HBV–DNA >2000 IU/mL, they should receive antiviral treatment regardless of the ALT level.

As discussed previously, a significant proportion of patients who become HBeAg negative with positive anti-HBe will have elevated HBV–DNA levels.82 Although the median HBV–DNA levels are lower than for HBeAg-positive patients, these patients tend to be older and have more advanced underlying liver disease. It has also been shown that the majority of complications of cirrhosis, including HCC, occur in HBeAg-negative patients.62

Therefore it is clear that even after HBeAg seroconversion a significant proportion of patients will still have ongoing active liver disease and would benefit from antiviral therapy. The difficulty lies in the identification of such patients who have ongoing disease activity after HBeAg seroconversion. There is currently no consensus as to the definition of “HBeAg-negative, active CHB”, and hence the available guidelines differ in their recommendations about the selection of patients for treatment. In the updated AASLD guidelines, indications for treatment include patients with ALT ≥2 × ULN and HBV–DNA levels ≥20 000 IU/mL, and no treatment for those with ALT <1 × ULN, irrespective of HBV–DNA levels.4 This would exclude a substantial amount of patients who would be at high risk of disease progression, including some who already have cirrhosis. In an alternative treatment algorithm from the USA, indications for treatment included ALT >1 × ULN and HBV–DNA >2000 IU/mL, whereas those with HBV–DNA <2000 IU/mL and normal ALT would not be treated.3 Both guidelines suggest liver biopsy to determine disease activity and stage of fibrosis in patients who have ALT and HBV–DNA levels in the range outside the definite treatment or non-treatment levels.3,4

Unfortunately, liver biopsy is associated with morbidity and with low patient acceptance. This means that some patients may decline liver biopsy and therefore would be excluded from treatment by the above paradigm, despite many having significant underlying liver disease. Although newer modalities are available for the non-invasive assessment of liver fibrosis, including transient elastography and the use of serum biomarkers, there is no consensus on their utility in this clinical setting. Further studies are required before they can be incorporated into routine use or in treatment algorithms.

The evolution of treatment strategies over the past decade has reflected our increasing knowledge about the natural history of CHB and the risks associated with unchecked progression of liver disease, along with improved sensitivity and accuracy of laboratory assays to measure viral load. The indications for treatment have expanded to include patients who had been thought safe from disease progression previously. Ongoing review of new evidence as it becomes available is important to optimize the selection of patients for treatment, as well as to ensure that eligible patients do not forego therapy. Patients at a higher risk can be identified according to their biochemical and serological status, in addition to the HBV–DNA level. This may ultimately obviate the need for a liver biopsy, especially as newer non-invasive modalities become validated and available in clinical practice.

Duration of therapy

The goal of preventing cirrhosis and the development of HCC is likely to be achieved by the prolonged suppression of HBV replication. In HBeAg-positive patients, the current AASLD guidelines suggest stopping treatment 6 months after HBeAg seroconversion (without mentioning the HBV–DNA or the ALT levels), and retreat if relapse should occur.4 An alternative treatment algorithm is that treatment can be stopped 6–12 months after HBeAg seroconversion, providing that HBV–DNA is undetectable by polymerase chain reaction (PCR).3 The latter approach, advocated by Keefe and colleagues, would seem more appropriate given the inadequacy of HBeAg seroconversion alone as a treatment end-point, and also the high rate of relapse after discontinuation of therapy. Close monitoring of patients after the cessation of therapy is mandatory. We recommend checking the HBV–DNA levels 1 month after stopping therapy, and every 3 months thereafter. Antiviral therapy should be restarted in those with evidence of reactivation. In HBeAg-negative patients, both guidelines suggest that long-term therapy is required. The Keefe et al. treatment algorithm and AASLD guidelines are summarized in Figures 5 and 6, respectively.

Figure 5.

Treatment of hepatitis B e-antigen-positive patients. Treatment algorithm described by Keefe et al.3 and the AASLD4 guidelines. ALT, alanine aminotransferase; ULN, upper limit of normal.

Figure 6.

Treatment of hepatitis B e-antigen-negative patients. Treatment algorithm described by Keefe et al.3 and the AASLD4 guidelines. ALT, alanine aminotransferase; ULN, upper limit of normal.

With prolonged antiviral therapy, there is a concern about the development of drug-resistant mutations and the potential for drug toxicity. Despite the high rate of drug resistance with prolonged lamivudine therapy, patients with drug-resistant HBV still benefit from treatment when compared to patients with no treatment.20,83 Furthermore, newer antiviral drugs with higher antiviral potency and higher genetic barrier (to develop drug resistance), such as entecavir, have much lower resistance rates.

In addition, there is concern about drug toxicity with prolonged therapy. The safety and long-term efficacy of antiviral drugs are paramount, particularly for HBeAg-negative patients who may be committing to life-long therapy unless HBsAg seroconversion occurs (and this has rarely been observed to date). Older agents, such as lamivudine and adefovir, have long-term safety data, but newer agents are currently lacking in long-term data both for efficacy and safety. Despite this, concerns about drug toxicity with long-term therapy appear largely unfounded given the preclinical safety results of the currently-licensed nucleotide/nucleotide analogs. In general, all the available oral nucleoside analogs are well tolerated. The documented nephrotoxic effect of adefovir occurs rarely at the dose used for HBV treatment, although renal function should be monitored regularly while patients remain on treatment.84

The advantage of IFN-based therapy over oral antiviral therapy is that the duration of therapy is more clearly defined. However, the optimal length of IFN therapy remains to be determined. A study has shown that treatment for 48 weeks with peg-IFN appeared to be superior to 24 weeks of treatment in achieving sustained viral response, although different formulations of IFN were used.85 More recently, 24 weeks of therapy has been shown to be similar to 48 weeks of treatment with peg-IFN-α2a in HBeAg-positive patients with good predictors of response.86 However, the advantage of a defined treatment length is offset by the high proportion of patients who will not respond to IFN therapy; these patients will require further therapy with oral nucleoside analogs. The side-effects of IFN therapy have been well described.87

Monitoring therapy

None of the published guidelines provide specific criteria for on-treatment monitoring of patients. During antiviral therapy, the degree of viral suppression has been shown to be the most important determinant of therapeutic outcomes.88 More specifically, the importance of effective early viral suppression in determining the long-term treatment outcome has been shown in several studies. In patients treated with lamivudine, the reduction of serum HBV–DNA to less than 200 IU/mL after 24 weeks of treatment results in a lower rate of YMDD mutations during subsequent prolonged therapy.89 Similarly, in patients treated with adefovir, a higher HBV–DNA at week 48 is predictive of the emergence of subsequent adefovir resistance.90

For patients treated with 48 weeks of peg-IFN-α2a, reduction of HBV–DNA below 80 IU/mL at week 12 was associated with a sustained response at week 72.91 Using telbivudine, patients with serum HBV–DNA <200 IU/mL at 24 weeks had higher HBeAg seroconversion rates (38% vs 8%, respectively, P < 0.0001), higher rates of HBV–DNA negativity by PCR assay in HBeAg-positive and HBeAg-negative patients (85% vs 14% and 86% vs 28%, respectively, P < 0.0001), and 8–61-fold reduced odds of viral breakthrough (P < 0.0001) at week 52 when compared with those patients with higher viral loads at week 24.52

A more recent study of lamivudine treatment has shown that HBV–DNA levels of less than 2000 IU/mL as early as week 4 can be used to predict accurately favorable responses at 5 years. In this study, a favorable treatment response was defined as HBeAg seroconversion with ALT normalization and HBV–DNA levels <2000 IU/mL without the emergence of YMDD mutations.92

A recently-published treatment algorithm has incorporated the early monitoring of viral suppression as part of the “roadmap concept” in managing CHB.93 This concept is summarized in Figure 7. The authors propose that the assessment of primary non-response at week 12, followed by the assessment of early predictors of efficacy at week 24 of treatment should be used to chart subsequent treatment choices. For example, they suggested the addition of a more potent antiviral drug for patients with primary treatment failure (<1 log reduction from baseline) at week 12, and in patients with inadequate virological response (≥2000 IU/mL) at week 24. In patients with partial virological responses (≥60 to <2000 IU/mL), the addition of a more potent drug would depend on whether the initial drug had low genetic barrier, high genetic barrier, or suboptimal antiviral potency.93

Figure 7.

Roadmap concept in the treatment of chronic hepatitis B93. PCR, polymerase chain reaction; Rx, treatment.

With the increasing number of antiviral drugs against HBV becoming available, the roadmap concept optimizes antiviral therapy adjustment, especially when initiating therapy with a drug that has a low genetic barrier or confers suboptimal antiviral efficacy. However, this roadmap concept may not be applicable to more potent drugs with high genetic barriers, such as entecavir, for treatment-naïve patients. It is also not known whether it would be useful in the treatment of patients with pre-existing lamivudine resistance, where subsequent treatment with entecavir is associated with the development of entecavir resistance, whereas adefovir add-on therapy is not (at least for 3 years).31 However, it seems likely that the roadmap concept can eventually be modified to encompass those patients with pre-existing drug-resistant mutations. Recent evidence shows that early viral suppression with adefovir in treating lamivudine-resistant HBV is associated with more favorable outcomes.94–96


With a better understanding of the natural history of chronic HBV infection and the resultant liver disease of CHB, the treatment paradigm is continuing to evolve. The primary and ultimate goal remains the eradication of HBV, but this does not appear possible with the currently-available antiviral agents used for medium-term therapy (2–5 years). On a practical level, the long-term suppression of HBV replication, as reflected by low-serum HBV–DNA levels, should be optimal aim. However, in practice, this has to be achieved through the current constraints of available antiviral agents. Although peg-IFN therapy offers a finite duration of therapy, only patients with high ALT levels are suitable, and even among those, a high proportion of patients have no response in terms of HBeAg seroconversion and HBV–DNA suppression to below PCR detectability. Furthermore, the long-term outcome of those who are apparently successfully treated is not yet available. Conversely, the long-term (or at least medium-term) outcome is more promising for oral nucleoside/nucleotide analogs. With them, however, the benefits of treatment are lessened with the development of drug-resistant mutations.

The availability of entecavir is a significant improvement in the current antiviral strategy; data from at least 2 years of treatment surveillance indicate that it may achieve the twin goals of sustained viral suppression with minimal risk of drug resistance. However, longer-term data are required to confirm this. Entecavir is also a very expensive drug for long-term use.

The optimal choice of an antiviral agent in treatment-naïve patients should be a drug with high antiviral potency coupled with a high genetic barrier to reduce the risk of resistance. If a drug with a lower genetic barrier is used, identifying early factors, such as viral load to predict long-term outcome is important. This would select patients most likely to achieve successful long-term viral suppression with their current regimens, and an alternative agent can be offered to those with suboptimal early viral response. By using newer, more potent antiviral therapy and adopting the use of early viral response as a predictor of long-term response, this is likely to further reduce the development of drug resistant mutations.

Finally, as our knowledge of CHB is changing, and with newer agents becoming available in the future, the paradigms for optimal CHB treatment are constantly evolving. Future treatment models are likely to be more tailored towards individual patients, taking into account the type of antiviral agent used, host factors, viral factors, and other predictors of treatment response.