Personal view: the management of chronic hepatitis B infection


Dr M. Sherman, Toronto General Hospital, 200 Elizabeth Street, Toronto, Ontario M5G 2C4, Canada.


There are two new nucleoside analogues available for the management of chronic hepatitis B, adefovir and entecavir, and several more in development. In addition, pegylated interferon has become available. Large-scale population studies have re-emphasized the significance of viral load in predicting a poor outcome over the longer term. These new developments have prompted a reassessment of the indications and objectives of therapy for chronic hepatitis B.

Hepatitis B virus deoxyribonucleic acid, rather than alanine aminotransferase should be the prime indication for therapy. Hepatitis B e antigen seroconversion can be achieved in 30–40% of treated patients whatever agent is used. However, it takes longer for nucleoside analogues to achieve the same seroconversion rates as interferon. In anti-HBe-positive disease long-term therapy is required for most patients because the relapse rate after withdrawal of therapy is very high, irrespective of the agent used.

Viral resistance limits the use of lamivudine, and to a lesser extent adefovir. Resistance to entecavir has so far only been described in pre-existing lamivudine resistance. Although therapy with combinations of nucleoside analogues has not been investigated to any extent, this is the only way to reduce the emergence or resistance, and studies are urgently needed.


In the last 3 years two new nucleoside analogues have been introduced for the management of chronic hepatitis B. Other potent agents are in various phases of development and will also be available in the next few years. Hepatologists and others treating patients with chronic hepatitis B will soon be in a position similar to that faced by those treating human immunodeficiency virus (HIV), i.e. there will be a plethora of agents, and it will take some skill to choose the best agent(s) for any individual patient. Thus, hepatitis B treatment is entering a new phase and it is therefore time to reassess current treatment paradigms. In this review I will revisit the indications for and end points of therapy, and try to indicate how the philosophy of managing patients with chronic hepatitis B is likely to change in the next 5–6 years.

Indications for therapy

The prime indication for starting a patient on treatment for chronic hepatitis B is an elevated alanine aminotransferase (ALT).1–3 This may be modified by various factors, such as severity of disease on biopsy, magnitude of ALT elevation, presence of hepatic decompensation, or the use of immunosuppressive agents. Nonetheless, for most patients and doctors, patients are not started on therapy until or unless the ALT is elevated. All the treatment algorithms recommended by the manufacturers of the hepatitis B anti-viral drugs, and all guidelines issued by the continental liver disease societies [American Association for the Study of Liver Disease (AASLD), EASL (European Association for Study of the Liver), Asian Pacific Association for Study of the Liver (APASL)]1–3 require an elevated ALT before considering treatment. This is because it was believed that if the short- and intermediate-term goals of normalizing ALT were achieved, the longer term goals of preventing cirrhosis, liver failure and hepatocellular carcinoma (HCC) could also be achieved. ALT therefore was thought to identify patients who had a higher risk of developing cirrhosis and HCC. Somewhat paradoxically, however, although elevated ALT is used to initiate treatment, it is not always used as a marker of successful therapy.

Recent studies have cast doubt on the validity of using an elevated ALT as the sole criterion for initiating treatment. First, when ALT is the main indication for treatment only a minority of all hepatitis B carriers will undergo treatment, far fewer than the number who will eventually die of liver-related causes. Clearly many patients who might benefit from treatment are not being treated. Secondly, there is considerable evidence that as with hepatitis C, the serum ALT does not always correspond with degree of fibrosis. Yang et al.4 have shown that up to 45% of patients with chronic hepatitis B and normal ALT have significant fibrosis (stage 2 or higher). This proportion is likely to be different in populations with different ages than the population studied by Yang et al.,4 because the higher the mean age of the population under study, the more likely it is that fibrosis will be present. Nonetheless, the presence of advanced fibrosis represents a failure of management. Thirdly, a recent large-scale population study in Taiwan has demonstrated, in a large cohort of subjects that ALT does not correlated well with outcome and does not strongly predict the development of HCC, cirrhosis or liver-related death.5 Thus, elevated ALT, although it identifies patients with active liver disease, misses some patients with significant inflammation and scarring, who are at risk for bad outcomes.

When interferon was the only drug available for hepatitis B it was a reasonable compromise to consider only short-term outcomes, given that interferon was only suitable for patients with elevated ALT concentrations. However, with the availability of several nucleoside analogues and more to come, we are now in a position to consider the possibility of long-term viral suppression so that it may no longer be appropriate to exclude patients with normal ALT from treatment.

Objectives of therapy

The objectives of therapy must be to decrease the incidence of cirrhosis, liver failure and liver cancer, if not to prevent these complications altogether. However, because these end points are difficult to study we have, to date, used surrogate end points to assess efficacy of therapy. Most prominent of these has been seroconversion from hepatitis B e antigen (HBeAg) to anti-HBe, normalization of ALT and histological improvement. Seroconversion from HBeAg to anti-HBe is associated with viral suppression in many, if not most patients in whom this occurs, but this can be achieved in only a minority of treated patients. Normalization of ALT is transient if viral load remains high, and histological improvement also depends on viral suppression.

HBeAg seroconversion

Historically HBeAg seroconversion has been regarded as an indication of remission of disease activity.6, 7 The early studies of treatment of chronic hepatitis B were conducted exclusively in HBeAg-positive patients, and used HBeAg seroconversion as an end point. The first long-term studies of the effect of interferon did in fact show that if seroconversion could be achieved the incidence of adverse outcomes were reduced.6, 7 More recent data, however, cast doubt that treatment-induced HBeAg seroconversion is a meaningful end point. First, treatment-induced seroconversion may not be very durable.8, 9 Following interferon or lamivudine therapy up to 50% of patients who seroconvert lose anti-HBe or have a reappearance of hepatitis B virus deoxyribonucleic acid (HBV-DNA). A long-term study from Hong Kong has suggested that treatment-induced seroconversion merely advances spontaneous seroconversion by 3–5 years.10 Yuen et al.10 have shown in matched interferon-treated and untreated control cohorts that although there was a higher initial seroconversion rate in the treated group, 5 years after recruitment to the study the proportion of patients who had undergone HBeAg seroconversion was similar in both groups. There was no difference in the incidence of adverse outcomes in the two groups. A similar study in children also suggested that all interferon treatment advanced seroconversion by 3 years,11 but did not increase the number of patients undergoing seroconversion. The difference in the outcomes of the earlier and later studies may be a reflection of different genotypes predominating in the populations. The earlier studies in Europe probably included many patients with genotype A, who have a better response to interferon, whereas the later studies from Asia likely included mainly genotypes B and C, which are known to respond less well to interferon.

Finally, whether patients are treated with interferon or nucleoside analogues, only a minority achieve seroconversion (30–40%), so that for the remainder, some form of ongoing therapy is necessary.12–18

Anti-HBe-positive chronic hepatitis B

In anti-HBe-positive patients remission of disease by either interferon or nucleoside analogue therapy has proved transient, so that normalization of ALT and a fall in HBV-DNA are not sustained over periods longer than 1–2 years, except for a small proportion of patients (20%).19–22 There is no known HBV-DNA concentration in these patients that ensures a sustained response. Nor is it clear that the likelihood of a relapse is greater with HBV-DNA concentrations that are, e.g. >105 or <104. Longer therapy than 1 year does not ensure permanent remission any more than shorter therapy.21 There is thus at present no defined end point of therapy for patients who are anti-HBe-positive. Many, if not all these patients will require indefinite therapy, or at least until HBsAg loss or seroconversion to anti-HB positivity occurs.

A new approach

If elevated ALT misses identifying patients who have a high risk of developing cirrhosis and HCC are there any other markers that have a better predictive value? Some factors, such as jaundice,23 either therapy-induced or spontaneous have been shown to predict short-term adverse outcomes, including death related to liver failure. Older age,22 and the presence of cirrhosis22 are also obvious predictors of adverse outcomes. Cirrhosis confers a risk for HCC and liver failure, so age is a predictor, mainly as a surrogate marker for duration of infection. But by the time a patient is cirrhotic or elderly, the opportunity to prevent HCC may largely be in the past. Ideally what is needed is a marker that predicts cirrhosis and HCC years in advance of either of these events occurring.

In the last year several studies have been presented in abstract form that show that the best predictor of the development of complications of chronic hepatitis B was the serum HBV-DNA concentration. There are three studies, one in Philadelphia, one in Haimen City China and one in Taiwan. The Philadelphia study24 recruited 3754 hepatitis B carriers of whom 37 developed HCC. In this cohort the risk of HCC was closely related to the HBV-DNA concentration at recruitment. The higher the HBV-DNA concentration the greater was the risk of HCC (Table 1). In the Haimen City cohort25 there were 2354 hepatitis B carriers aged 25–64 years with serum available for study. The cohort was followed for up to 10 years. Once again the HBV-DNA at recruitment was related to the risk of developing HCC, cirrhosis and of liver-related death. After 10 years the HCC-related mortality in the high HBV-DNA group (HBV-DNA concentration: >105 copies/mL) was almost 20% compared with the mortality in the low HBV-DNA group (HBV-DNA: <105 copies/mL) of about 4%. In the HBV-DNA-negative group the mortality from HCC was only about 2%, not statistically different from the low HBV-DNA group. Death from chronic liver disease at 10 years was about 8% in the high HBV-DNA group, compared with <1% in the low HBV-DNA and negative HBV-DNA groups. The most detailed reports come from the Taiwan cohort (REVEAL study).5, 26–28 This was a population study of 89 293 individuals aged 30–65 years recruited from seven townships in Taiwan. There were 4155 hepatitis B carriers in the population, of which 3851 formed the HCC study cohort, and 3774 formed the cirrhosis study cohort. Once again, the risk of progression to cirrhosis and mortality from HCC was related to baseline viral load. Persistence of a high viral load had a higher risk of HCC than if the viral load declined over time. Other risk factors for HCC are shown in Table 2. Thus, although an ALT greater than normal at recruitment was a risk factor for an adverse outcome, the predictive value was much less than for HBV-DNA. HBeAg was also a risk factor for the development of HCC, but this too was a minor risk factor.

Table 1.  Relationship between HBV-DNA and the risk of HCC in the Philadelphia cohort (from Ref.24)
Risk categoryHCC cases [n = 37; N (%)]Controls [n = 61; N (%)]RR (95% CI)
  1. HBV-DNA, hepatitis B virus deoxyribonucleic acid; HCC, hepatocellular carcinoma.

Undetectable HBV-DNA5 (13.5)32 (51.6)Reference
Low-titre HBV-DNA7 (18.9)11 (18.0)4.1 (1.1–15.5)
High-titre HBV-DNA25 (67.6)18 (29.5)8.4 (2.5–25.7)
Table 2.  Risk factors for HCC identified in the REVEAL study26
Risk factorAdjusted RR (95% CI)P-value
  1. HBV-DNA, hepatitis B virus deoxyribonucleic acid; HCC, hepatocellular carcinoma; HBeAg, hepatitis B e antigen.

Baseline HBV-DNA (≥105 copies/mL)6.4 (4.1–10.1)<0.001
Baseline HBV-DNA (104 to <105 copies/mL)2.5 (1.5–4.3)<0.01
HBeAg-positive status2.3 (1.6–3.3)<0.001
Male gender2.1 (1.4–3.2)<0.01
ALT ≥1 × ULN1.7 (1.2–2.6)<0.01
Alcohol use1.6 (1.1–2.3)<0.05
Older age1.10 (1.08–1.12)<0.01

Although these studies dramatically illustrate the point, we have known for years that suppression of virus replication is associated with improved outcomes, whether this occurs spontaneously in patients who undergo spontaneous HBeAg seroconversion and remission, or whether remission is medication-induced. However, the results of these studies now require a reassessment of the value of treating the virus, rather than treating the liver disease. Because the risk of HCC is higher in patients with active inflammatory disease,26, 29 and because higher HBV-DNA concentrations are more likely to be associated with active liver disease, it is intuitively reasonable that suppression of HBV-DNA would lead to a lower incidence of adverse liver-related outcomes. There is also direct evidence that suppression of DNA on therapy reduces the risk of HCC, at least in cirrhotics. Liaw et al.30 showed that lamivudine-treated cirrhotics had a lower incidence of HCC and other liver-related adverse events over about 3 years of follow-up than untreated cirrhotics. Failure of therapy due to lamivudine resistance was associated with progression of disease.

These results suggest that the prime indication for treatment should be the presence of a high HBV-DNA, rather than a high ALT. Clearly however, not all patients with a high HBV-DNA should be treated. In the natural history of chronic hepatitis B acquired in childhood the infected children pass through a period of very high HBV-DNA concentration, yet have little inflammatory activity in the liver, and have no or minimal fibrosis. These patients probably do not need treatment, or at least, with the limited treatment regimens currently available, should not be treated. It remains to be shown whether suppression of HBV-DNA replication in these young children will alter their risk of adverse outcomes, given that they are likely to be a low risk over any potential study period. The age of subjects recruited in the Hiamen city and Taiwan cohorts was between 25 and 65 years old. However, in the absence of full publication of these studies we do not know what age group was at highest risk for HCC. It is likely that HCC occurred in the older age group. More specifically, we need to know whether the risk was high enough in the youngest patients with high HBV-DNA concentrations to warrant starting therapy (in the absence of other factors indicating therapy). Patients who have any degree of hepatic fibrosis and a high HBV-DNA are indeed en route to cirrhosis, and possibly to HCC as well, and such patients may need treatment, regardless of age.

This concept has two other important corollaries to consider. These are; the HBV-DNA concentration that should trigger a consideration of treatment, and what the target HBV-DNA on therapy should be. A ‘high’ HBV-DNA concentration has not yet been adequately defined. Guidelines from AASLD and other liver disease societies1–3 suggest that the HBV-DNA concentration should be >105 copies/mL before considering treatment. In the REVEAL study referred to above HCC risk increased with increasing HBV-DNA concentration.26 At an HBV-DNA concentration of <104 copies/mL the HCC incidence over the follow-up period (12–13 years) was about 0.1–0.15%. At an HBV-DNA over >105 copies/mL the incidence was 1.1%. These are not the high incidence rates normally seen in clinic populations. Thus, although HBV-DNA concentrations are important indicators of risk, at low HBV-DNA levels there must be other indicators that should trigger treatment. These might be finding fibrosis on biopsy, or finding persistent or intermittent inflammation. Perhaps other risk factors such as male gender should also be included as a factor to be considered in deciding on therapy. Although there is no data to support this, it is likely that patients in their 30's who still have active liver disease should be treated. Many of these patients will in any case have an elevated ALT and significant fibrosis and qualify for treatment on that basis. However, an elevated ALT should no longer be a requirement. Whether patients with lower HBV-DNA concentrations should be treated or not remains uncertain. Cases have now been described in which active hepatitis has been associated with HBV-DNA concentration much lower than 105 copies/mL, and even lower than 104 copies/mL.31, 32 This, coupled with the risk of HCC, particularly in patients over the ages of 30–40 years suggest that patients with HBV-DNA concentration between 104 and 105 copies/mL should be considered for therapy if they have any level of fibrosis on biopsy, regardless of HBV-DNA. One exception might be the young HBeAg-positive patient who might be in the process of seroconverting and might spontaneously go into remission within a few months. Below an HBV-DNA of 104 the likelihood of active hepatitis B is quite low, and more often the cause of the elevated ALT will be due to something else, most commonly steatohepatitis.

So far it remains to be proven that reducing HBV-DNA over many years will reduce the incidence of cirrhosis, hepatic decompensation and HCC. However, it has been clearly shown that suppression of HBV-DNA in patients with cirrhosis does result in a decreased incidence of hepatic adverse outcomes.30 Furthermore, spontaneous declines in HBV-DNA associated with HBeAg seroconversion are also associated with a decrease in the rate of adverse outcomes.6, 7 Finally, although it probably would not satisfy the advocates of evidence-based medicine, it certainly makes sense that because high concentrations of HBV-DNA are associated with active disease, suppression of virus should cause remission of disease.

The second issue that arises out of the three cohort studies is that if a high HBV-DNA is a requirement to start treatment, how deeply must HBV-DNA be suppressed, and for how long to prevent emergence of hepatitis B complications. This is unknown at present. However, there are two important considerations. If nucleoside analogues are being used, prevention of emergence of resistance is an important on-therapy objective. We will later discuss the use of combination therapy, but it is clear that the risk of resistance is related to inadequate viral suppression. Again, this is intuitive. The less viral replication present, the lower the chance of a resistance mutation developing.33, 34 Secondly, in order to achieve a histological improvement in the majority of patients the HBV-DNA should be suppressed to below 103 copies/mL.35

Therapeutic agents

To some extent the way in which we use the therapeutic agents available to treat hepatitis B is anachronistic. The algorithms were developed in the era when it was believed that HBeAg seroconversion was desirable, and so management algorithms and studies to prove efficacy of drugs were designed around that concept. Thus, the initial studies of interferon were in HBeAg-positive individuals in whom seroconversion to anti-HBe was considered to be an appropriate end point.36, 37 We now recognize that this may no longer be the case for all patients. All the studies of interferon looked at a time-limited course of therapy. Most studies were of 6 months duration or less,36–40 although there are studies going on to 2 years.41, 42 All the studies of nucleoside analogue therapy were also designed around a time-limited course of therapy, with either histological improvement or HBeAg seroconversion as end points.14–16 These studies have dictated the way in which we use these agents. Standard interferon or pegylated interferon, whether for HBeAg-positive43, 44 or -negative disease45 are given for a limited time period. Nucleoside analogues, although widely used for periods longer than 1 year were licensed for single year use. Because HBeAg seroconversion was used in the registration studies for all these drugs, it continues to be used as an end point. ALT normalization is also used as an end point, even though, as hepatitis C has shown us, we need to redefine what a normal ALT really is, and even though we know that even with a normal ALT inflammatory activity can continue.

Standard interferon


Most guidelines indicate that HBeAg-positive patients with active disease should be treated for 16–24 weeks.1–3 The indication for starting patients on therapy in the initial studies was an ALT of more than twice normal.1–3 The expected seroconversion rate is between 30% and 35%.12 The predictors of response were high ALT, ‘low’ HBV-DNA (undefined), and active disease on biopsy.46, 47 These characteristics also identify patients with a higher likelihood of spontaneous seroconversion, as well as a higher likelihood of seroconversion with nucleoside analogue treatment.48 Other predictors of response were age younger than 40 years, and female gender. There are several studies looking at the durability of the HBeAg seroconversion, and the eventual loss of HBsAg. van Nunen et al.8 described that 2 years after therapy 20% of patients on interferon had sero-reverted to HBeAg-positive, but in patients who started off with ALT levels of five times normal had a sero-reversion rate of <10%. In contrast, Yuen et al.10 found that although the seroconversion rate on therapy was higher than in controls, the sero-reversion rate was also higher, so that 5 years after therapy the rate of seroconversion was similar in interferon-treated and untreated patients. The HBsAg seroconversion rate is about 8% of all treated patients.12

Pegylated interferon has also been used to treat HBeAg-positive hepatitis B. The first study, using pegylated interferon-α2a for 6 months showed a seroconversion rate of 37% at 90 μg/week, and 33% with 180 μg/week.49 A subsequent study using the same agent for 1 year found a similar response rate (35%), with HBV-DNA being undetectable by polymerase chain reaction (PCR) in 33%.44 Pegylated interferon-α2b-induced seroconversion in 36% of patients.43 Thus, 1 year of pegylated interferon induces approximately the same proportion of patients to seroconvert as does 4–6 months of standard interferon. The durability of the pegylated interferon-induced response is not known.


The duration of interferon therapy in anti-HBe-positive hepatitis B is 12–24 months.1–3, 20 The on-therapy response was 54%, measured by solution hybridization (and thus almost certainly lower than that if measured by PCR). However, 1 year after therapy the proportion of patients with undetectable HBV-DNA (again by solution hybridization) was no more than 20% in most reports.20, 22 Others have found that at about 1 year after treatment about 33% of responses (negative HBV-DNA by solution hybridization assay) were sustained.21 In addition between 9% and 12% of patients lost HBsAg or developed anti-HBs. As with the HBeAg-positive group, this is a negligible proportion, and this rate should not be used as a justification for using interferon.

Pegylated interferon has also been used to treat anti-HBe-positive hepatitis B.45 Twenty-four weeks after completion of therapy the biochemical response was 43%, but the virological response (HBV-DNA-negative by PCR) was only 19%. Eighty percentage of these responses were sustained for a further 6 months.50 Again, no long-term data are available.



Lamivudine has fair anti-viral activity, reducing HBV-DNA concentrations by about 4.5-logs over a year of therapy.16 In the initial registration trial lamivudine was able to induce seroconversion in about 16–18% of HBeAg-positive patients after 1 year of therapy.14 About 38% of patients achieved undetectable HBV-DNA, and 69% normalized ALT while on therapy. Over a longer course of therapy lamivudine-induced seroconversion reached 40% after 3 years of therapy.18 The seroconversion induced by 1 year of lamivudine is not as durable as that induced by interferon, with about 80% of patients who started off with an ALT less than twice normal relapsing within 2 years.8 However, with higher ALT levels the seroconversion appears to be more durable, with only about 50% of patients who started off with ALT levels >5 times the upper limit of normal relapsing at the 2-year mark. Prolonging therapy for 4–6 months postseroconversion may enhance the durability of the response.9 Most comparisons between nucleoside analogues and interferon are made on data from 1 year of nucleoside analogue use. This may be an unfair comparison, since over a longer treatment period the seroconversion rates appear to be about equal to that of interferon, at least in the lamivudine-treated patients who do not develop resistance to treatment, and it may be the overall durability is similar to that of interferon.


Anti-HBe-positive patients treated with lamivudine achieve normalization of ALT in 80–90% of patients, and about 73% achieve undetectable HBV-DNA (solution hybridization).51 These results are better than in the HBeAg-positive patients, probably because the initial HBV-DNA level at the start of therapy was 1–2-logs lower than in the HBeAg-positive group. Unfortunately, the relapse rate once therapy is stopped is high. After 1 year about 80% of patients relapse.52, 53 After 2 years only 50% remain in virological remission, although about 30% remain in biochemical remission with detectable circulating virus. Interestingly, those who remain in remission have a small chance of achieving HBsAg loss. Longer therapy (2 years) results in a higher sustained biochemical response off therapy, but the virological response is still only 25%.54

Lamivudine resistance develops quickly. Resistance starts to emerge after about 6 months of therapy and increases over time. At the end of a year about 20% of patients have lamivudine resistance. By 4 years the proportion is 70%.55 This severely limits the usefulness of lamivudine. Initial reports suggested that the YMDD mutant virus, although resistant, did not replicate as well as the wild-type virus, so that patients were better off staying on therapy despite resistance, rather than stopping therapy.56 However, there is a compensatory mutation that occurs that restores replication competency.57, 58 Once this has happened there is no advantage to continuing lamivudine.59 Stopping the lamivudine does not induce a flare of ALT activity, despite case reports to the contrary.

Adefovir dipivoxil


Adefovir dipivoxil was the second nucleot(s)ide analogue to be licensed for hepatitis B. In the registration studies for this agent the HBeAg seroconversion rate was a disappointing 12%.60 However, the advantage of adefovir was the low rates of resistance. ALT normalization occurred in 48% of patients. In contrast to lamivudine, after 4 years of adefovir only 18% of patients exhibit resistance.61 Adefovir however, at the licensed dose of 10 mg/day is not a highly potent drug. The mean reduction in viral titre in the first year was about 3.5-logs, compared with a mean reduction of about 4.5-logs using lamivudine. Longer therapy with adefovir is associated with increasing numbers of patients who normalize ALT, who become HBV-DNA-negative, and who seroconvert to anti-HBe-positive. After about 4 years of therapy the HBeAg seroconversion rate is about 40%.62


In the anti-HBe-positive studies the rate of normalization of ALT was 72%, and of HBV-DNA-negative by PCR was 51%.15 There is no information about the durability of this response.

Tenofovir disiproxil

Tenofovir disiproxil is not licensed to treat hepatitis B, but is licensed for HIV disease. Nonetheless, it is a relatively potent anti-viral for hepatitis B as well.63 There are as yet no large-scale studies showing benefit in the management of hepatitis B. However, it is a more potent agent that adefovir, and can be used in patients in whom the response to adefovir has been suboptimal. It has also been used in hepatitis B HIV-co-infected individuals, in whom adefovir is contraindicated.64 It is effective in lamivudine-resistant individuals.63 Tenofovir can achieve about 5.5-log reduction in viral load, i.e. similar to that achieved by lamivudine.63



Entecavir is a potent nucleoside analogue, which unlike the other agents, has no activity against HIV. In a randomized-controlled study of 0.5 mg/day of entecavir compared with standard doses of lamivudine in HBeAg-positive patients 21% achieved seroconversion after a year of therapy.16 The mean decline in HBV-DNA concentration was about 7-log10. This makes entecavir the most potent of the available nucleoside analogues.65, 66 At the end of a year of therapy 69% of patients had undetectable HBV-DNA by PCR compared with 38% in the lamivudine-treated group. No resistance was found in the entecavir-treated patients. There are no long-term follow-up data available yet.


In anti-HBe-positive patients entecavir was equally effective.67 The mean decline in HBV-DNA concentration was 5.2-logs. This is less of a decline than in the HBeAg-positive group, because the mean HBV-DNA at the start of the study was about 2-logs lower than for the HBeAg-positive group. Ninety-one percentage of entecavir-treated patients had undetectable HBV-DNA by PCR after 48 weeks, compared with 73% in the lamivudine-treated group. Again, no long-term data are available.

Viral resistance to nucleoside analogue treatment

As has been discussed earlier, lamivudine use is associated with a significant rate of development of resistance. After 4 years of therapy up to 70% of patients exhibit viral resistance.55 Although it was initially believed that the lamivudine-resistant HBV was less potent than the wild type and replicated more slowly, this has turned out not to be the case. There are additional mutations that are selected for that restore replication competency.57, 58 Thus, initially it was thought that patients who remained on lamivudine despite the appearance of resistance did better than those in whom lamivudine was stopped.56 In addition, there were several reports of flares of ALT on withdrawal of lamivudine in these patients. However, now it is clear that disease progresses substantially in patients who have lamivudine-resistant infection,68, 69 and it is no longer acceptable to maintain these patients on lamivudine. It is also now agreed that there is no danger in stopping the lamivudine.59, 70 Flares of ALT are uncommon.

Adefovir has been used to treat lamivudine-resistant infection.71, 72 Lamivudine-resistant mutants are fully susceptible to adefovir. There has been some discussion as to whether it was preferable to maintain the patient on lamivudine as well as introduce adefovir, or to switch to adefovir and stop the lamivudine. Although the numbers of patients studied is still small, data so far indicate that dual therapy is preferable. Adefovir resistance has been described in lamivudine-resistant patients using only adefovir, but not in patients receiving both adefovir and lamivudine.61 In nucleoside naive patients adefovir resistance is 18% at 4 years and 28% at 5 years.73

Entecavir resistance is uncommon. To date no resistance has been described in patients who were nucleoside naive prior to the introduction of entecavir. Entecavir can be used to treat lamivudine resistance.74 The dose of entecavir used is 1.0 mg/day, compared with the 0.5 mg/day used in naive patients. Entecavir can induce 75% of lamivudine-resistant individuals to normalize ALT, and 55% to have undetectable HBV-DNA (bDNA assay). The mean log drop in HBV-DNA concentration in lamivudine-resistant individuals was 5.14-log10 copies/mL. Entecavir efficacy is reduced by the YMDD mutant, but the potency is such that adequate viral suppression is achievable.75 Resistance to entecavir has only been described in viral strains that already exhibit lamivudine resistance. The YMDD mutant appears to be a prerequisite for the development of entecavir resistance.76 Three new mutations have been described that are associated with entecavir resistance, but in the absence of the YMDD mutation, in vitro none of these mutations exhibit resistance, either alone or in combination. Entecavir resistance occurred in 2% of lamivudine-resistant patients after 1 year of therapy.74 These mutants remain fully susceptible to adefovir. As lamivudine therapy predisposes to entecavir resistance, and because lamivudine has such a high resistance rate it should probably no longer be used as first-line therapy.

Newer nucleoside analoges

There are a series of new nucleoside analogues currently under testing, including telbivudine, D4T, remofovir, emtricitabine and clevudine. Some are highly potent (clevudine and telbivudine), although possibly not quite as potent as entecavir. Many are also susceptible to the YMDD mutation (all except pradefovir). The place that these agents will have in the management of chronic hepatitis B is at present uncertain.

Who to treat and how


There is probably a subset of patients who have a high likelihood of seroconversion whatever therapy is used, whether interferon for 6–12 months or nucleoside analogues for 1–4 years. Which drug should be started will be a matter of doctor and patient choice.

  • 1In the HBeAg-positive group interferon only works well in patients who have high levels of ALT, and low concentrations of HBV-DNA. These are characteristics that also identify patients who have the highest rate of spontaneous seroconversion. These characteristics also identify patients that will respond best to nucleoside analogue therapy. It takes longer with nucleoside analogues to obtain a seroconversion than with interferon (3 years vs. 4–12 months), but the proportion of patients undergoing seroconversion is similar (see Table 3).
  • 2It has been said that interferon-induced seroconversion is more durable than that induced by lamivudine or adefovir after 1 year of therapy, but it is unfair to compare interferon results with 1 year of nucleoside analogue therapy, because that is not how we use nucleoside analogues. The durability of seroconversion with longer nucleoside analogue therapy than 1 year is unknown. Just as with interferon, the durability of the HBeAg seroconversion is better in those who start with a higher ALT.
  • 3Interferon is associated with considerable side-effects, whereas the nucleoside analogues have very few side-effects.
  • 4Some have argued that there is a higher rate of ‘s’ antigen loss with interferon than with the nucleoside analogues. Again the comparison that as been made is between interferon given for a standard duration and 1 year of nucleoside analogues. However, if the comparison is between standard interferon therapy and 3–4 years of nucleoside analogue therapy, the rate of ‘s’ antigen loss is similar. Even so, ‘s’ antigen loss is a rare event, occurring in fewer than 10% of treated patients. This is not a reason to prefer one drug over another.
  • 5Whatever agent is used only about 30–40% of HBeAg-treated patients undergo seroconversion, leaving 60–70% requiring further treatment (Table 3). Seroconversion might be durable in 33–66% of those achieving this end point. Thus, over time a further 10–14% of the originally treated patients will require further treatment, i.e. a total of 70–84% of those treated will require long-term therapy.
Table 3.  Rates of HBeAg seroconversion with different drugs used to treat hepatitis B
Therapeutic agentHBe seroconversion rate (%)
  1. HBeAg, hepatitis B e antigen; Std IFN, standard interferon; Peg IFN, pegylated interferon.

Std IFN3312
Peg IFN3246
Peg IFN + lamivudine2746
Lamivudine year 117–2014
Lamivudine year 22717
Lamivudine year 34018
Adefovir year 11215
Adefovir year 44464

As long-term therapy will be required for the majority of patients the major concern is the development of viral resistance. To minimize the emergence of resistance the treatment chosen should suppress viral replication deeply. Ideally the on-treatment viral load should be undetectable or at least <104 copies/mL.77 All nucleoside analogues can achieve this in some patients, but to increase the likelihood of suppressing virus maximally, the most potent drug should be used. Therefore, entecavir should be the first line of therapy. This is a good choice as not only is it the most potent agent, but it also seems to have the least likelihood of resistance.

For patients already on lamivudine or adefovir, if viral suppression is adequate (<104 copies/mL) there is probably no need to change drugs. However, if viral suppression is not adequate consideration should be given to switching therapy.

Patients on nucleoside analogues should be treated until ‘e’ antigen seroconversion occurs. It has been suggested that ‘e’ antigen seroconversion on its own is not sufficient, and that treatment should continue for at least another 6 months to ensure durability of the seroconversion. Alternatively, if the HBV-DNA postseroconversion remains high (>104 copies/mL) it has been suggested that treatment should continue until the HBV-DNA falls below that level. There is no evidence to support that doing so results in more durable seroconversion, but at levels >104 copies/mL the risk of active liver disease remains despite seroconversion.

Patients who achieve seroconversion should remain under close observation (every 2–4 months) for several years, as there is a risk of sero-reversion. However, the longer the disease remains inactive, the more likely it is that the remission is permanent, so that after 2–3 years it may be possible to check the ALT and HBV-DNA at less frequent intervals. It must be emphasized, however, that all these patients need continued follow-up indefinitely.


All forms of therapy for anti-HBe-positive chronic hepatitis B are associated with a high relapse rate after therapy is withdrawn. Therefore, anti-HBe-positive hepatitis B requires long-term therapy. It may be possible, if there has been an initial good response with very low or undetectable viral levels, that treatment can be withdrawn at some point, provided that the patient is kept under observation, and therapy re-instituted if the viral load increases to >104 copies/mL. What proportion of patients in this situation will relapse is unknown.

As the relapse rate after standard interferon is high, once again it seems that there is no place for interferon, given that it is much more expensive than nucleoside analogues given for the same period of time, and has considerable side-effects. However, because long-term use of nucleoside analogues will be required once again, one of the objectives of therapy must be to prevent the emergence of viral resistance. Lamivudine is not a good candidate for this type of hepatitis B, because of the high rate of resistance. However, in some patients, in whom the viral load can be suppressed to undetectable levels, lamivudine is acceptable. If the viral load does not fall rapidly then the patient should be switched to another agent. Adefovir is an acceptable agent, because again, although not highly potent it does induce undetectable virus in some. Once more, however, if the viral load does not reach low levels (certainly below 104) a more potent drug should be used.

Combination therapy

Interferon and pegylated interferon have been studied in combination with lamivudine with very disappointing results. There was no advantage of the combination over interferon alone.

Combinations of nucleoside analogues have not been studied to any extent so far, so that there is no data to support their use. However, although there might not be greater efficacy, the use of two agents will likely decrease the risk of viral resistance developing. Perhaps patients with advanced liver disease should be treated with combination therapy, because emergence of resistance in this population is often associated with clinical deterioration. However, once again, there is no data to support this use.


The treatment of chronic hepatitis B has improved substantially, and will continue to improve. The availability of potent nucleoside analogues will likely broaden the indications for treatment. As with hepatitis C, those patients who need treatment are those who have predictors of bad outcomes. High viral loads must now be taken into consideration as potential indicators for therapy. However, it is achieved, suppression of the viral replication will lead to less hepatic inflammation, and in turn to less HCC and cirrhosis. Deep and permanent suppression of virus should be the objective, whether this is achieved by medication-induced durable seroconversion or by maintenance therapy, whether this is achieved using a single agent or whether combination therapy is required. Although not discussed in detail, we need to give serious thought to trials of multiple drug therapy. If HIV has taught us anything, it is that monotherapy leads to resistance. We need to start trials of combination therapy soon or risk our patients developing infections that become sequentially resistant to everything that we have.