Treatment of chronic hepatitis B: Evolution over two decades


  • Conflict of interest
    MF Yuen acted as consultant and received research grants from Glaxo Smith Kline, Gristol Myers Squibb and Novertis. CL Lai has acted as consultant for Bristol Myers Squibb and Gilead Sciences.

Professor Man-Fung Yuen and Ching-Lung Lai, Department of Medicine, The University of Hong Kong, Queen Mary Hospital, Pokfulam Road, Hong Kong, China. Email: or


There has been a recent paradigm shift in the indications and endpoints of treatment for chronic hepatitis B (CHB). Hepatitis B e antigen (HBeAg)-negative disease is being increasingly recognized. Antiviral treatment for both HBeAg-positive and HBeAg-negative patients should aim at long-term suppression of HBV DNA, with the ultimate ideal endpoint of hepatitis B surface antigen (HBsAg) seroconversion. Conventional interferon alpha (IFN-α), the only agent licensed in 1991, has been superseded by pegylated IFN-α. HBeAg seroconversion using pegylated IFN-α is 33%, with only 25% of HBeAg-positive patients achieving undetectable HBV DNA by polymerase chain reaction (PCR) assay. Five nucleoside/nucleotide analogues have been licensed since 1998. Lamivudine, an L-nucleoside, is limited by the development of resistance in 76% of patients after 5 years of therapy. Telbivudine, another L-nucleoside, is more potent than lamivudine but resistance still develops in 25% of HBeAg-positive and 11% HBeAg-negative patients after 2 years. Adefovir, an acyclic phosphonate, is relatively weak, but is effective against lamivudine- and telbivudine- resistant mutations, for which it should be used in combination (add-on therapy) rather than substituted. Resistance to adefovir develops slowly, rising to 29% for HBeAg-negative patients by year 5, but more rapidly when used alone for lamivudine-resistant HBV. Currently the two first line nucleoside/nucleotides are entecavir and tenofovir. Entecavir, a cyclopentane (D-nucleoside), is very potent, with 94% of patients having undetectable HBV DNA after 5 years. Resistance develops in only 1.2% of treatment-naïve patients. Tenofovir, another acyclic nucleotide, is more potent with less renal toxicity compared to adefovir. It is effective against lamivudine-resistant mutations when used alone. No resistance to tenofovir has been described after its use for 3 years or longer, often for patients with human immunodeficiency virus/HBV co-infection. With these current, potent antiviral agents associated with very low rates of resistance, long-term HBV DNA suppression and possibly even reversal of cirrhosis can now be achieved in a proportion of patients. In addition, long-term treatment with these antiviral agents is associated with a reduced risk of development of hepatocellular carcinoma.


Chronic hepatitis B (CHB) infection is an important global disease. It is estimated that more than 400 million people have been chronically infected with the hepatitis B virus (HBV).1 It is of even more importance to the Asia Pacific region since 75% of those infected with HBV are Asians.1 Up to 40% of these CHB patients will develop the long-term, devastating complications of liver failure (due to liver decompensation from cirrhosis or severe acute exacerbations of CHB2 and hepatocellular carcinoma (HCC).3,4 These complications substantially reduce life expectancy and quality of life, as well as imposing great demands on healthcare resources. Although, the beneficial effects of universal HBV vaccination in reducing the number of CHB patients is already apparent in the younger population,5 the incidence of HCC remains high in most Asian countries.3 Therefore, effective treatment of CHB is still urgently needed. Fortunately, there has been an acceleration of drug development for CHB with seven agents being approved over the last 20 years (Fig. 1).

Figure 1.

Approved therapy for chronic hepatitis B over time.

Conventional interferon alpha for the treatment of chronic hepatitis B

In 1991, conventional interferon-alpha (IFN-α) 2b was the first agent approved for the treatment of CHB. Its major mechanism of action is immune modulation although there is also a weak anti-viral effect.6 The success of treatment is assessed by HBV DNA reduction, normalization of alanine aminotransferase (ALT) level and hepatitis B e antigen (HBeAg) seroconversion (for HBeAg positive patients, i.e. loss of HBeAg with sustained suppression of HBV DNA and usually gain of anti-HBe).7 It is now difficult to assess efficacy of this agent because of then-limitations of the technology of measuring viral activity. In those days, HBV DNA levels were measured in a semi-quantitative manner by dot-blot hybridization assays which had a lower limit of detection of 100 000 copies/mL (20 000 IU/mL). As discussed later, this is at least one (possibly two) order(s) of magnitude higher than the “target” for current HBV suppression. In addition, with a lack of well-documented studies of the natural history of CHB, it was difficult to decide who and when to start treatment as well as when to stop treatment. As a result, the treatment was arbitrarily given for 16–24 weeks, and “response” was defined by HBeAg seroconversion and normalization of ALT (although HBV DNA levels were also assessed, they were not used for the measurement of response because of assay insensitivity). The indication for treatment was mainly based on the likelihood of HBeAg seroconversion, that is, patients with ALT levels over twice the upper limit of normal (ULN). The endpoint of treatment was not based on any evidence of reduction of long-term complications and mortality. More recently there are long-term follow-up studies examining the effect of IFN-α on the incidence of long-term complications and HCC, and the results of such studies are conflicting.8–13

Paradigm shift in the indication and endpoints of treatment

With newer data from studies on the natural history of CHB, the paradigm of treatment of CHB has shifted over the past 10 years. In particular, a large cohort follow-up study of 3233 Asian patients showed that those with an ALT level 1–2 times of upper limit of normal (ULN) has a significantly higher risk of development of long-term complications compared to patients with ALT levels higher than 2–6 x ULN.14 In addition, patients with ALT levels at the upper half of the normal range (defined as 53 U/L for male; 31 U/L for female) are already having a higher risk of development of complications. Another study has shown that the mean ALT levels for patients with HCC during three follow-ups before the development of HCC was 51, 47 and 57 U/L, that is, just above the ULN.15 Thus, guidelines for the indication for antiviral therapy of CHB that are based on ALT > twice the upper limit of normal (x 2ULN) are no longer rational.

The importance of serum HBV DNA levels has also changed now that more sensitive and accurate assays with broad specific dynamic range are available for HBV DNA quantification. Most of these are now based on polymerase chain reaction (PCR) technology. It has been shown convincingly that patients with high serum HBV DNA levels (> 2000 IU/mL) have a significantly higher risk of development of both cirrhosis and HCC.16–18 Furthermore, there is not a threshold of HBV DNA levels below which long-term complications do not occur.19,20 In fact, 15% of patients who developed HCC have HBV DNA < 200 IU/mL at the time of diagnosis of HCC.20 What may be important, however, is level of HBV DNA during a likely “incubation time” for development of cirrhosis and/or HCC, possibly during the antecedent 2–3 decades before clinical onset of these complications.

In addition, it has been found that over two-thirds of patients with long-term complications are anti-HBe positive.14,19,20 The importance of HBeAg-negative disease is now widely recognized, but recognition of the fallacy of regarding HBeAg seroconversion as the sole endpoint for stopping therapy has been slow. Even now, this is not universally accepted.21,22 The permanent suppression of HBV DNA is a more preferable endpoint (albeit a stricter definition of HBeAg seroconversion as indicated earlier might include such suppression), with HBsAg seroconversion being the “ideal” endpoint. However, it is noted that HBsAg seroconversion is only achieved in a relatively small proportion of patients with any kind of therapy, and among those with cirrhosis, even loss of HBsAg does not convey complete protection against risk of subsequent HCC.23

With these new findings, patients should be considered for treatment when serum HBV DNA levels are higher than 2000 IU/mL and ALT levels are persistently elevated, irrespective of the HBeAg/ anti-HBe status. They are now the treatment criteria suggested by the recent European Association for the Study of the Liver (EASL) guidelines,24 although the HBV DNA and ALT are set at higher levels for HBeAg-positive patients (> 20 000 IU/mL and > 2 X ULN, respectively) according to both American Association for the Study of the Liver Diseases (AASLD) and Asia Pacific Association for the Study of the Liver (APASL) guidelines.25,26 For HBeAg-negative patients these two guidelines require the patients to have HBV DNA and ALT levels > 20 000 IU/mL and > 2 X ULN (AASLD) and > 2000 IU/mL and > 2 X ULN (APASL) before they should be considered for treatment.

According to another algorithm, treatment should be initiated in patients with elevated ALT levels and HBV DNA > 20 000 IU/mL for HBeAg-positive patients and HBV DNA > 2000 IU/mL for HBeAg-negative patients.27 In spite of these discrepancies, it has been generally accepted that the treatment target should be the permanent suppression of viral replication (preferably with HBV DNA level undetectable by PCR assay) to reduce the long-term complications of CHB.

Special considerations for starting treatment regardless of the ALT levels may be required for: (i) patients who are already older than 40 years with other evidence of advanced disease, e.g. advanced fibrosis; and (ii) patients who are already cirrhotic.

The first two nucleoside and nucleotide analogues (Lamivudine and Adefovir Dipivoxil)

It was not until the first approval of a nucleoside analogue (NA), lamivudine (chemically an L-nucleoside) in 1998 that a reduction in rate of transition to cirrhosis and risk of HCC could be achieved with some success. The arrival of lamivudine, with the convenience of oral therapy and minimal adverse effects, triggered a rapid evolution in the treatment of CHB, but there was an important down-side. Compared to placebo, use of lamivudine in CHB is associated with significantly better chance of achieving HBV DNA suppression, ALT normalization and HBeAg seroconversion after one year of therapy.28 However, viral resistance emerged in 16% of cases after just one year of therapy. Being the only approved NA at that time, lamivudine has been widely used around the world, and occasionally for inappropriately short courses. It has been shown that lamivudine resistance continues to accumulate with increasing duration of treatment (up to 76% after 5 years of lamivudine treatment).29,30 In spite of this problem, several long-term studies have shown that long-term lamivudine treatment is able to reduce the disease progression in terms of development of cirrhosis and HCC.29,31–33 These benefits are observed in both cirrhotic and non-cirrhotic patients.30,32 Patients with lamivudine-resistant HBV have a blunted response but still have a lower rate of development of long-term complications compared to untreated patients. Currently the use of lamivudine is largely limited by the occurrence of resistance.

In 2002, the second NA, adefovir dipivoxil was approved for the treatment of CHB. The arrival of this acyclic phosphonate agent also provides more new insights into the treatment of CHB. Firstly, in addition to anti-viral potency, the intrinsic stereoscopic structure is a very important factor for the emergence of viral resistance. Although adefovir is less effective in HBV DNA suppression compared to lamivudine,34 the chance of drug resistance is lower compared to the latter (29% after 5 years of adefovir treatment).35 This slower rate of development of resistance is possibly related to the minimal flexible acyclic structure of adefovir that subverts resistance due to steric hindrance.36 Secondly, with the experience of the use of adefovir in lamivudine-resistant disease, it was shown unambiguously that combination of a nucleoside with a nucleotide with complimentary resistance profiles is superior to switching from one agent to another. Thus, adding adefovir to patients with lamivudine-resistant HBV is associated with a significantly lower chance of development of adefovir resistance compared to switching patients from lamivudine to adefovir.37,38 In addition, earlier addition of adefovir at the time of virologic breakthroughs when the HBV DNA levels are low (before the development of biochemical breakthroughs) is associated with a significantly better long term outcome in term of HBV DNA suppression, ALT normalization and development of adefovir-resistant HBV.39,40 The use of adefovir as a first line agent for treatment naïve CHB patients is limited by its modest antiviral suppressive effect and its potential renal toxicity. It has mainly been used in lamivudine-resistant disease.

Pegylated interferon alpha

While waiting for more promising NA for treatment approval for CHB, a new formulation of IFN-α, pegylated IFN-α2a was approved in 2005. (It had been approved for the treatment of chronic hepatitis C in January 2001.) Since then, conventional IFN-α has been gradually replaced by pegylated IFN-α2a. Although there are more updated data on the determinants of the development of long-term CHB complications, the criteria of starting pegylated IFN-α are based on the findings from studies using conventional IFN-α, i.e. ALT > 2 ULN. The duration of pegylated IFN-α therapy is again arbitrarily chosen, this time as 48 weeks rather than the 16–24 weeks for conventional IFN-α. Even with the extension of the duration of treatment to 48 weeks, it has shown that the HBeAg seroconversion rate (33%) is almost identical to that of conventional IFN-α as determined in a meta-analysis (32%). In addition, after 3 years of follow-up for HBeAg-negative patients with lower baseline HBV DNA levels, the rate of undetectable HBV DNA by PCR assay, is only 18%.41 Though 8% of these patients also have HBsAg seroconversion, data from entecavir and tenofovir give similar rates of HBsAg seroconversion in comparable (largely European) cohorts.


Lamivudine as the first line agent for treatment naïve CHB patients, and additional adefovir for those with lamivudine-resistant disease, were the main treatment strategies during the period between 1998 and 2004. In 2005, entecavir came in the arena of CHB treatment. It has two outstanding characteristics. It is a nucleoside belonging to a new subgroup, cyclopentane, and has an extremely high anti-HBV suppressive effect, rendering 67% of HBeAg-positive and 90% of HBeAg-negative patients to have unquantifiable HBV DNA (by PCR assays) after one year of therapy.42,43 A recent study showed that over 91% of patients have unquantifiable HBV DNA (< 12 IU/mL) after three years of entecavir treatment.44 This high rate of undetectable HBV DNA levels persists after five years of continuous entecavir therapy.45 The potent antiviral effect is probably related to its rapid intracellular phosphorylation to the active triphosphate derivative, as well as its triple action in the inhibition of HBV DNA synthesis, starting with the priming of the HBV DNA polymerase.46 This potent viral suppression has now been shown to be effective in not only reducing necroinflammation, but also reversing fibrosis and cirrhosis in 57 patients on continuous entecavir treatment with a third liver biopsy (45 of the third biopsies at year 6 of therapy).

Another outstanding feature of entecavir is the very low chance of emergence of resistance (only 1.2% after five years among treatment-naïve subjects).47 The low resistance rate is related to both the profound viral suppression as well as the requirement of at least three sites of genetic mutations in order to confer entecavir resistance. (Two of these three sites overlap with lamivudine resistance, both lamivudine and entecavir being nucleoside analogues.) This characteristic is referred to as “high genetic barrier”. Because of these merits, entecavir is now the first line agent for treatment-naïve CHB patients. However, it is not a drug of choice for patients with lamivudine-resistant disease because of the common sharing of two out of the three required mutations between entecavir and lamivudine resulting in a high rate of development of entecavir resistant mutations.48,49 The chance of emergence of entecavir-resistant HBV is as high as 51% in patients with pre-existing lamivudine resistant mutations after five years of entecavir treatment.47 Because of this limitation, patients with lamivudine resistant HBV should be preferably treated by tenofovir which will be mentioned below, or adefovir if tenofovir is not widely available.

HBsAg seroconversion occurs in 5.1% of patients after 96 weeks of entecavir.50 In patients who continue to receive entecavir, a further 1.4% have HBsAg seroconversion by year 5.45


While better treatment for lamivudine-resistant disease was still under investigation, telbivudine, another NA belonging to the L-nucleoside subgroup was approved for treatment for CHB in 2006. Telbivudine is more potent than lamivudine in reducing the HBV DNA levels by an addition of l log copies/mL after one year of therapy.51 The HBV DNA undetectable rates are 60% vs 40% for HBeAg-positive and 88% vs 71% for HBeAg-negative patients, respectively. Therefore the chance of drug resistance compared to lamivudine-treated patients is lower in telbivudine-treated patients, although they share the same genetic mutation sites, and like lamivudine, a single mutation can cause resistance. However, the emergence of viral resistance to telbivudine (25% for HBeAg-positive patients and 11% for HBeAg-negative patients after two years)52 is still higher than adefovir and entecavir.

The use of lamivudine and telbivudine has shown the importance of selecting patients who achieve early potent HBV DNA suppression as a criterion for continuing therapy with these agents. Yuen et al. first demonstrate that HBV DNA levels after 24 weeks of lamivudine therapy is a reliable marker for predicting the chance of lamivudine resistance on continuous treatment.53 This concept of CHB treatment has also been proven in the GLOBE trial of telbivudine vs lamivudine.51,52 In addition to the measurement of HBV DNA treatment response at week 24, baseline HBV DNA levels and ALT levels are also important in selecting patients to be treated with these two agents. According to Zeuzem et al., by selecting HBeAg-positive patients with baseline HBV DNA levels < 9 logs copies/ml (2 × 108 IU/mL) and ALT ≥ 2 X ULN and HBeAg-negative patients by baseline HBV DNA levels < 7 logs copies/mL plus undetectable HBV DNA at week 24, the chance of telbivudine resistance at year 2 becomes low (1.8% for HBeAg-positive and 2.3% for HBeAg-negative cases).54 These finding have been confirmed by another study in lamivudine-treated patients.55

Therefore, though entecavir is the preferred option, in countries where cost is a major concern, the L-nucleoside analogues lamivudine and telbivudine can still be used by selecting patients with favorable baseline HBV DNA and ALT levels plus the on-treatment HBV DNA response at 24 weeks.

Tenofovir disoproxil fumarate

Rescue therapy for patients with viral resistance to the L-nucleoside analogues was dependent on the use of adefovir until 2008 when the second acyclic phosphonate nucleotide analogue (same subgroup of adefovir) was approved for the treatment of CHB. This latest approved agent is tenofovir disoproxil fumarate. It has three outstanding features. It causes very profound HBV DNA suppression (6 logs copies/mL), a magnitude of reduction very similar to entecavir and telbivudine. Secondly, it is very effective for the treatment of lamivudine-resistant HBV, even more effective than adefovir. And of great importance, it is similar to entecavir in having a very low chance of drug resistance (to date, no such cases have been observed after four years of therapy).56 It is therefore the ideal agent for patients with lamivudine- or telbivudine-resistant diseases, but also for treatment-naïve patients. After 96 weeks of tenofovir, loss of HBsAg and HBsAg seroconversion occurs in 7.0% and 5.6% of HBeAg-negative patients, and in 3.8% and 1.9% of HBeAg-positive patients, respectively. A study showed that tenofovir is very effective in NA treatment-experienced populations, with 79% of patients achieving unquantifiable HBV DNA by the assay used (< 80 IU/mL) after a mean treatment period of 23 months.57 Renal toxicity has been reported in a small proportion of HIV-infected patients treated with tenofovir, but no renal toxicity has been reported in immuno-competent CHB patients.

Because of the excellent features mentioned above, tenofovir is also recommended as a first line agent for treatment-naïve CHB patients.

Agent under investigation

LB80380 is a nucleotide belonging to the same group as adefovir and tenofovir (acyclic phosphonate). Two phase I and II trials demonstrate that this agent has profound viral suppressive effects in both treatment-naïve patients and patients with lamivudine-resistant disease.58,59 As an acyclic phosphonate, the resistance rate is expected to be low. Currently phase II studies in treatment-naïve patients are being conducted.


The JGH landmark article about four year lamivudine therapy published six years ago came during a “watershed” in more than 20 years of drug development for CHB. Lamivudine offered ease of antiviral efficacy, but its limitation was suboptimal efficacy and an unacceptably high rate of drug resistance. Since 2005, better control of this disease through more profound suppression of viral replication is now achievable in more than 90% of both HBeAg-positive and HBeAg-negative cases. Apart from this revolutionary improvement of patient outcome, NA treatment has also provided invaluable information on the viral dynamics of HBV. These include the way HBV adapts to antiviral therapy by developing resistant mutations with restoration of viral replication, a varying genetic profile of resistant viruses to different groups of NA, differences in the replication competency between wild-type HBV and different drug-resistant mutants, the importance of rapid control of the viral replication to prevent emergence of resistant viruses, and effective treatment of drug resistant HBV by the early addition of another, appropriately chosen NA.60 Patients with CHB should now be treated once they reach the threshold for treatment as indicated by the various guidelines or with special characteristics, namely advanced age with advanced histology or clinical evidence of cirrhosis. Treatment for both HBeAg-positive and HBeAg-negative patients should be on a long-term basis, possibly until HBsAg seroconversion. Permanent suppression of HBV replication with reversal of fibrosis and cirrhosis is achievable. The hope that this will also substantially reduce the risk of HCC, as suggested by the experience with lamivudine4,29,31,61 awaits longer follow-up of patients on continuous effective HBV antiviral therapy.