Errata: ERRATA Volume 31, Issue 5, 748, Article first published online: 22 March 2011
Correspondence George V. Papatheodoridis, 2nd Department of Internal Medicine, Athens University Medical School, Hippokration General Hospital of Athens, 114 Vas. Sophias ave., 115 27 Athens, Greece Tel: +30 210 777 4742 Fax: +30 210 770 6871 e-mail: firstname.lastname@example.org
Current agents used in the treatment of chronic hepatitis B (CHB) can be classified into interferons-α (IFN-α: standard or pegylated) and nucleos(t)ide analogues (NUCs). NUCs are now used in most CHB patients for several reasons. They can be given to all CHB patients, even those with contraindications to IFN-α. NUCs are more convenient to use (one oral tablet daily) than IFN-α (subcutaneous injections) and are well tolerated with a good safety profile, while IFN-α has frequent and potentially severe side effects and worsens the patient's quality of life. All NUCs are potent anti-hepatitis B virus agents (all except adefovir are more potent than IFN-α) with entecavir and tenofovir being the most potent. Most importantly NUCs all have minimal risk of resistance during long-term monotherapy. Prolongation of entecavir or tenofovir monotherapy maintains and slightly increases the initially high virological remission rates (67–76% of HBeAg-positive and 90–93% of HBeAg-negative patients) and this is expected to result in improved long-term outcomes. The need for long-term, perhaps indefinite, treatment is the main limitation of NUCs and the finite duration (48 weeks) the main advantage of IFN-α. However, only a minority of IFN-α-treated patients achieve durable sustained off-treatment responses (HBeAg-positive: 30–35%, HBeAg-negative: 20–25%), while NUCs may be safely discontinued in HBeAg-positive patients with stable HBeAg seroconversion. Because there will always be concerns for safety and family planning issues with long-term therapy, NUCs should be used judiciously and should not be prescribed in young CHB patients with mild liver disease.
Treatment against hepatitis B virus (HBV) has dramatically improved over the last 10 years and it can now offer benefits in the majority of chronic HBV patients improving their prognosis and long-term outcomes (1–3). In parallel with the increasing available therapeutic options, both the complexity and the efficacy of anti-HBV treatment have increased (1). However, HBV eradication still remains an ideal but rather rarely attainable therapeutic target and therefore only patients who are at risk of progression to advanced liver disease are usually considered for treatment (1–5). Thus, treatment is widely recommended in patients with HBeAg-positive or HBeAg-negative chronic hepatitis B (CHB) (high serum HBV DNA levels, elevated aminotransferases and active necroinflammation with fibrosis), while follow-up is suggested for HBeAg-positive immunotolerant patients (high viraemia, normal aminotransferases and minimal histological lesions) and HBeAg-negative inactive chronic carriers (undetectable or low viraemia, normal aminotransferases and minimal histological lesions) (1–3, 5). Treatment is also recommended for patients with HBV-related decompensated cirrhosis and HBV transplant cases, while pre-emptive anti-HBV therapy should be provided to any chronic HBV patient receiving immunosuppressive therapy in order to prevent the HBV reactivation that often follows the discontinuation or dose reduction of immunosuppression (2, 3, 5).
Currently, there are seven drugs that are widely licensed for the treatment of CHB: interferon-α (IFN-α), pegylated IFN-α-2a (PEG-IFN-α-2a), lamivudine (LAM), adefovir dipivoxil (ADV), entecavir (ETV), telbivudine (TBV) and tenofovir disoproxil fumarate (TDF). These drugs may be classified into IFN-α (standard or pegylated), which have both antiviral and immunomodulatory activities and are administered subcutaneously, and the oral pure antiviral agents, which are all nucleoside (LAM, ETV, TBV) or nucleotide (ADV, TDF) analogues (NUCs) belonging to the same class of agents (HBV polymerase inhibitors) (1).
Several clinical trials have evaluated the efficacy and safety of these agents. Comparing them directly is not always feasible, as the conclusions reached at each trial highly depend on the study design and endpoints, while there is heterogeneity of patient populations. On the other hand, some conclusions on the advantages and disadvantages as well as on the limitations and adverse events of the anti-HBV agents can be drawn from the existing data. At the end, it is widely known that NUCs are now used in the treatment of the majority of CHB patients (6). This review summarises the reasons for the very common use of NUCs in the treatment of CHB (Table 1).
Table 1. Pros and cons of nucleos(t)ide analogue therapy for chronic hepatitis B
Improvement of the overall long-term outcome and survival even in patients with cirrhosis of any severity
Possibility of prevention of HBV exacerbation after immunosuppression
Oral administration (once daily)
Excellent tolerance, good safety
High potency (ETV/TBV/TDF>LAM>ADV) – high 1-year response rates
Probability of treatment discontinuation in HBeAg-positive patients with stable HBeAg seroconversion
Increasing HBeAg seroconversion rates with prolongation of any NUCs
Maintenance and slight increase of 1-year virological remission rates with prolongation of ETV/TDF monotherapy
Unknown (perhaps indefinite) duration of treatment
Rare HBsAg loss
Not durable sustained off-treatment responses in the majority of patients
Progressively increasing risk of viral resistance (LAM>ADV/ TBV>ETV/TDF) (minimal risk with ETV/TDF)
Safety concerns with long-term duration of treatment
Concerns in patients of reproductive age – Unclear safety in pregnancy
Advantages of nucleos(t)ide analogues
Wide therapeutic indications – absence of contraindications
Treatment with NUCs has negligible contraindications offering possibilities for therapeutic intervention to all chronic HBV patients, even those with advanced liver disease (decompensated cirrhosis or post-transplant) (1–3). In contrast, PEG-IFN-α is contraindicated in patients who mostly need treatment, such as cases with decompensated liver disease and/or haematological signs of portal hypertension (low platelet or neutrophil counts) and it is usually avoided in the transplant setting. In addition, pre-emptive therapy with a NUCs is the recommended approach for the prevention of HBV exacerbation in chronic HBV patients receiving immunosuppressive agents, while PEG-IFN-α is not recommended and is not used for this indication (2, 3).
Convenient use, excellent tolerance, good safety
It is obvious that the use of NUCs, which are administered as one oral tablet per day, is much more convenient for the patients than the use of IFN-α (standard or pegylated), which is given as subcutaneous injections. The easier administration of the currently used PEG-IFN-α (one instead of at least three subcutaneous injections per week) may have somewhat weakened but has not completely abolished the disadvantage of the parenteral administration of IFN-α.
A very important advantage of NUCs over PEG-IFN-α are their excellent tolerance and the absence of serious side effects, as it has been shown in many phase III trials (7–14). In contrast, PEG-IFN-α therapy frequently worsens the patient's quality of life and is often associated with a wide spectrum of side effects (15–18). The majority of IFN-α-treated patients develop flu-like symptoms and many (20–40%) patients may experience fatigue, headache, myalgias, local reactions at injection sites, weight loss, hair loss, irritability and/or bone marrow suppression (mainly thrombocytopaenia or neutropaenia), which requires monitoring with full blood counts at least monthly during therapy. In addition, some IFN-α-treated patients may develop more severe adverse events, such as autoimmune disorders, depression and rarely suicidal tendency, polyneuropathy, retinopathy and optic neuritis, hearing loss or seizures (15–18).
Because NUCs are often given for years, or even indefinitely, safety concerns may be raised with such a long-term use. So far, all licensed NUCs have been proven to be rather safe and well tolerated in all published trials, in which they were given for 1 to 5 years (19). Of course, the most extensive results on safety, which is similar to placebo, come from LAM which was the first licensed oral anti-HBV agent and has been used in millions of CHB patients in clinical practice over the last 12 years (19, 20). ADV, the second licensed agent, has a small risk of nephrotoxicity with 3% of CHB patients experiencing serum creatinine increases of ≥0.5 mg/dl above baseline values at 5 years of ADV monotherapy (21). TBV has been associated with creatinine phosphokinase increases in up to 12% of treated patients (13, 22), but symptomatic myopathy has been extremely rare and no specific laboratory monitoring is recommended during TBV therapy (19). ETV seems to be a fairly safe agent, while there are some concerns about nephrotoxicity during long-term TDF therapy (19). It should be noted, however, that there is no significant increase of serum creatinine or decrease of creatinine clearance so far, at 3 years of TDF therapy, in CHB patients (23, 24). Because all NUCs are excreted through the kidneys, dosing adjustments are necessary in patients with creatinine clearance <50 ml/min (2). In addition, monitoring of serum creatinine and phosphate levels is recommended during treatment with ADV or TDF (19). Given that patients and physicians adhere to the few recommendations of follow-up, the long-term use of NUCs appears to be particularly safe without any significant adverse event in the vast majority of treated patients.
High initial antiviral potency
All NUCs are potent anti-HBV agents, although their potency is not the same. However, all NUCs, except ADV, are more potent than PEG-IFN-α (Fig. 1). ADV has the lowest potency achieving HBV DNA undetectability at year-1 in 13–21% of HBeAg-positive (7, 8) and 70–75% of HBeAg-negative NUCs naive CHB patients (8, 9), while LAM has intermediate potency achieving HBV DNA undetectability at year-1 in 40–44% of HÂeAg-positive (10–13) and 60–75% of HBeAg-negative NUCs naive CHB patients (13, 14, 25, 26). ETV, TBV and TDF are the newest and most potent anti-HBV agents achieving HBV DNA undetectability at year-1 in 60–76% of HBeAg-positive (8, 12, 13) and 88–93% of HBeAg-negative NUCs naive CHB patients (8, 13, 14).
High on-therapy remission and low resistance rates under long-term therapy with newer nucleos(t)ide analogues
Despite their high initial potency, 1-year therapy with NUCs does not usually lead to sustained off-therapy responses and therefore NUCs are usually given for several years or even indefinitely (1–3, 27). Thus, the long-term on-therapy remission rates, which are strictly linked to the risk of viral resistance, are of paramount importance. It is reasonable that there will be a progressively increasing risk of viral resistance with prolongation of any antiviral therapy due to the possibility of selection of treatment-resistant mutant viral strains (28, 29), but the probability of viral resistance is not the same for all anti-HBV agents.
Lamivudine has the worst resistance profile among all currently available NUCs. HBV resistance starts to develop within the first year of LAM monotherapy and increases over time at an annual rate of 15–25% in parallel with decreasing rates of virological remission (20, 26, 30–32). The probability of resistance is higher in LAM-treated patients with higher baseline viraemia levels (20, 26, 30) and in those who fail to achieve rapid (within the first 6 months) and profound HBV DNA suppression (33, 34). Because of the high probability of HBV resistance and the negative impact of LAM resistance to other NUCs (35), LAM monotherapy is not currently considered as an optimal first-line long-term treatment for CHB (2, 3). However, the agent is still widely used in several countries mostly because of its low cost.
Good long-term ADV data are available only in NUCs naive patients with HBeAg-negative CHB, in whom 5-year ADV monotherapy has been shown to maintain virological and biochemical remission in 67–69% of cases (21). However, ADV resistance may emerge during the second year reaching cumulative rates of 3, 6, 18 and 29% at the end of the 2, 3, 4 and 5 years of therapy respectively (21). The strongest predictor of subsequent ADV resistance is the suboptimal initial suppression of viraemia at 12 months of therapy (21).
Continuation of TBV monotherapy for 2 years has been reported to maintain virological responses in the majority of patients, but the response rates seem to decrease at the end of second year compared with those achieved during the first year of therapy in both HBeAg-positive (HBV DNA undetectability by polymerase chain reaction (PCR), year-1: 60% and year-2: 54%) and HBeAg-negative CHB patients (HBV DNA undetectability by PCR, year-1: 88% and year-2: 79%) (13, 22). Cumulative rates of TBV resistance also increase over time, being 4.4 and 2.7% at the end of first year and 21.6 and 8.6% at the end of the second year of therapy in HBeAg-positive and HBeAg-negative CHB respectively (13, 22). The absence of residual viraemia (HBV DNA <400 copies/ml) at 6 months of TBV monotherapy is associated with a rather low probability of HBV resistance at 2 years (36).
Entecavir and tenofovir
The concerns of HBV resistance have minimised with the introduction of the newer NUCs, ETV and TDF. In NUCs naive HBeAg-positive or HBeAg-negative CHB patients, the cumulative rate of resistance has been reported to be only 1.2% at ≥5 years for ETV and 0% at 3 years for TDF (23, 24, 37). In agreement with the high potency and the low risk of resistance, ETV and TDF appear to offer maintenance and even an increase of the initially high virological response rates with the prolongation of treatment. In particular, continuation of ETV monotherapy in partial virological responders at 48 weeks (HBV DNA <0.7 MEq/ml with no HBeAg loss for HBeAg-positive or with alanine transaminase >1.25 × ULN for HBeAg-negative CHB patients) has been found to offer some additional virological responses during the second and third year in HBeAg-positive (38, 39) and during the second year in HBeAg-negative CHB (40) and to maintain high virological response rates for at least 5 years of ETV monotherapy (37). Similarly, continuation of TDF therapy for up to 3 years, with or without the addition of emtricitabine at ≥72 weeks, has been shown to maintain and slightly increase the initially high virological and biochemical response rates in both HBeAg-positive and HBeAg-negative CHB patients (23, 24).
The high potency and the low risk of resistance during long-term ETV and TDF monotherapy have made ETV and TDF the preferred agents for the treatment of CHB in recent clinical guidelines (3). Moreover, the efficacy of these two agents has favoured the use of monotherapy and has minimised the debates on the potential benefits of using ab initio combination of NUCs in the treatment of CHB, which were reasonable in the era of weaker anti-HBV agents like LAM.
Possibility for management of viral resistance
The management of viral resistance was a major concern in the late 1990s, when LAM was the only available NUCs, but the concerns have now significantly subsided after the availability of several NUCs. ADV was the agent of choice for the treatment of patients with LAM resistance with the early ‘add-on’ strategy (ADV added to ongoing LAM) proved to be the optimal approach (41–44). TDF, which is a more potent anti-HBV agent compared with ADV and active against LAM resistant HBV strains (35, 45, 46), has now become the treatment of choice for the management of patients with LAM resistance. Whether switching to TDF or addition of TDF to ongoing LAM is the optimal and the more cost-effective approach for using TDF in patients with LAM resistance is still debatable (47), but it seems safer for now to adopt the ‘add-on’ strategy according to the ADV experience in this setting. ETV has also been licensed for the treatment of patients with LAM resistance, but its cumulative long-term resistance rate resistance is rather high (>50% at 5 years) in such patients because the pre-existing LAM resistant HBV strains increase dramatically the probability of ETV resistance (37). The latter observations have made ETV a less attractive long-term therapeutic option for patients with LAM resistance (1–3). ADV-resistant mutants are usually susceptible to nucleoside analogues (LAM, ETV, TBV), but they may also be sensitive to TDF depending on the mutation pattern, while TBV or the rare ETV resistance strains are usually sensitive to nucleotide analogues (ADV, TDF) and should be preferentially treated with TDF (48, 49).
Limitations of nucleos(t)ide analogues
Need for long-term, perhaps indefinite, duration of treatment
The need for long-term, perhaps indefinite, duration of treatment in the majority of CHB patients represents the main limitation of NUCs therapy raising two main concerns: (a) the maintenance of virological on-therapy remission and the risk of viral resistance, (b) the long-term safety of these agents. As discussed previously, however, the risk of viral resistance has minimised during long-term therapy with ETV or TDF, which maintain and even increase the high initial response rates, while the safety profile of these two agents as well as of any NUCs seems to be acceptable even with their long-term use.
In current clinical practice, the most common limitation of long-term treatment with NUCs is the concern on the safety in case of pregnancy when NUCs are given in young adults CHB patients of reproductive age. All licensed oral anti-HBV agents have warnings for potential risk of adverse fetal outcomes with use immediately before and during pregnancy (19). Therefore, it is recommended that both men and women of child-bearing age treated with NUCs should use contraception for their sexual contacts. However, women sometimes conceive while they or their male partners are under NUCs and then several practical questions arise mainly on the risk of adverse fetal outcomes. In such cases, the use of agents classified to FDA Category B (TDF, TBV) over Category C (LAM, ADV, ETV) has a reasonable advantage (19), particularly during the first trimester of pregnancy when organogenesis is occurring. In addition, Category B agents and LAM, for which there is accumulating clinical experience, may be better used during the last trimester of pregnancy in order to decrease the risk of vertical transmission in pregnant women with very high viraemia levels. Most specialists suggest avoid breast feeding if the mother should be treated with NUCs (19), while TDF may have an advantage for use during breast feeding as the active substance, tenofovir, has poor oral availability (50).
Low sustained off-therapy response rates with treatment of finite duration
The predetermined finite duration (usually 48 weeks) is considered to be the main advantage of IFN-α-based therapy, particularly for relatively young patients. However, only a minority of PEG-IFN-α-treated CHB patients achieve sustained off-therapy responses, which are usually durable, may lead to HBsAg loss and improve the patients' long-term outcome (51–56). In particular, 48-week courses with PEG-IFN-α have been shown to achieve sustained off-treatment responses (HBeAg seroconversion and/or virological-biochemical remission) in approximately one-third of patients with HBeAg-positive CHB (17, 57) and one-fourth of patients with HBeAg-negative CHB (18, 56). A disappointing feature of PEG-IFN-α therapy is that there is no accurate and widely accepted early on-therapy predictive marker of response and therefore patients usually continue an unpleasant treatment for the predetermined duration (48 weeks), although the majority of them do not eventually respond. Recently, there have been several attempts to identify accurate on-therapy markers with high negative predictive values, such as HBeAg or HBsAg levels or decrease of serum HBV DNA levels (58–61), but none of them has been adequately validated and accepted in international guidelines and clinical practice (2, 3).
Although PEG-IFN-α is always given for a finite duration, NUCs may be also given for a finite duration in some CHB patients, In fact, NUCs given for more than 2 years may offer sustained off-therapy response rates similar to or even higher than those observed after 1-year PEG-IFN-α therapy. In particular in HBeAg-positive CHB, all current guidelines recommend that NUCs can be stopped after 6–12 months of consolidation therapy in patients who achieve stable HBeAg seroconversion and HBV DNA undetectability (2, 3, 5). Of course, compared with PEG-IFN-α, HBeAg seroconversion rates are lower at the end of year-1 without substantial differences among NUCs despite their differences in antiviral potency. In NUCs naive HBeAg-positive CHB patients, HBeAg seroconversion rates at year-1 have been reported to be 12–17% with ADV (7, 8), 16–18% with LAM (10–13) and 21–22% with ETV, TBV and TDF (8, 12, 13). HBeAg seroconversion rates tend to increase with prolongation of any NUCs therapy approaching and sometimes exceeding cumulative rates of 30% at 3 years with some patients achieving even HBsAg loss (20, 22, 23, 38, 62–64). HBeAg seroreversion may be observed in a proportion of cases after treatment discontinuation with rates of seroreversion ranging widely among studies (65–69), but some patients will remain in the HBeAg-negative inactive carrier state for years without any therapy. Because there are no directly comparative studies including HBeAg-positive CHB patients, the comparisons in the sustained off-therapy response rates after 1-year therapy with PEG-IFN-α and ≥2-year therapy with a NUCs are rather uncertain.
In HBeAg-negative CHB, there is no clear stopping rule for NUCs therapy, but there are some promising preliminary data with long-term effective ADV monotherapy. It has been reported that ADV may be safely discontinued after successful inhibition of HBV replication for at least 4–5 years leading patients to the inactive carrier state and often to HBsAg loss and anti-HBs development (70). These findings should be confirmed in other studies with the newer, more potent agents that are now widely used, such as ETV and TDF.
Current NUCs are well tolerated, safe and highly potent having minimal-negligible risk of long-term resistance (ETV, TDF). The availability of such NUCs have increased their advantages and substantially decreased their disadvantages, compared with PEG-IFN-α, for the treatment of CHB. Thus, it is easily understood why NUCs currently represent the treatment approach used by the majority of CHB patients (1). Firstly, NUCs can be used in chronic HBV patients who have contraindications to PEG-IFN-α or need prevention from HBV exacerbation after immunosuppressive therapy. Secondly, NUCs are the preferred therapeutic option even in patients who could be treated with PEG-IFN-α, because of the high probability of no sustained response, the parenteral administration and the tolerability and safety issues with PEG-IFN-α therapy. Sometimes, even the best candidates for PEG-IFN-α therapy (young patients with favourable predictors of response) may prefer to be treated with NUCs because of the fear of IFN-α-related adverse events and its potential negative effect on the quality of their life. In addition, if a proportion of treatment-naive CHB may be initially treated with PEG-IFN-α, most of them will eventually receive NUCs, as they will not achieve sustained responses after PEG-IFN-α therapy (1).
Although NUCs are used by the vast majority of CHB patients and they have been shown to modify the disease and patient long-term outcome, they do not represent the optimal therapeutic approach. Suppression of HBV replication can be initially achieved in the majority of CHB patients, but all patients under long-term NUCs therapy should be monitored carefully for virological response and possible virological breakthroughs with serum HBV DNA determinations at baseline and at least every 6 months (2, 3, 71). Because there is always a risk for emergence of viral resistance during long-term NUCs therapy, which limits their effectiveness and usually worsens liver disease, NUCs with an optimal antiviral potency and a good resistance profile should be preferentially used. It should be also noted that treatment with an NUCs improves the long-term outcome of CHB (31, 72), but it does not eliminate the risk of hepatocellular carcinoma, particularly in patients with pre-existing cirrhosis who should remain under surveillance (73). The need for long-term, perhaps indefinite, treatment represents the major limitation of current NUCs therapy raising safety and family planning issues. Therefore, NUCs should be used cautiously particularly in young CHB patients with mild liver disease.
In the future, novel therapeutic targets or creative combination therapies are required in order to increase the rates of sustained responses after treatment discontinuation. In addition, new research efforts are needed to identify therapeutic approaches that would be able to achieve HBV eradication or at least HBsAg loss and seroconversion in a sizeable proportion of patients. Such approaches would expand the current treatment indications offering chances for successful therapy in several subgroups of patients who are now excluded from treatment having a chronic viral infection, sometimes with very high levels of replication, but not enough evidence of biochemical and histological activity.
Conflicts of interest
George V. Papatheodoridis is on the advisory board and/or a speaker for Bristol-Myers Squibb, Gilead, Novartis Pharmaceuticals, Roche, Schering-Plough; Research grants from Bristol-Myers Squibb, Gilead, Roche.