Chronic hepatitis B virus (HBV) infection is a serious clinical problem because of its worldwide distribution and potential adverse sequelae. Despite the implementation of effective universal vaccination programmes in >160 countries, there are still >350 million people with chronic HBV infection worldwide: of these, 75% reside in the Asia-Pacific region (1), where the infection is usually acquired perinatally or in early childhood (2). These people are at risk of developing hepatic decompensation, liver cirrhosis or hepatocellular carcinoma (HCC). In the past decades, tremendous advances in the understanding of the virus, the natural history and the immunopathogenesis of chronic HBV infection have been achieved. Furthermore, the advent of effective antiviral agents with different mechanisms of action, and the accumulation of substantial experience in the use of these drugs have led to better therapeutic strategies for chronic HBV infection (3). Evidence has shown that the long-term outcomes of chronic HBV infection may improve under therapy.
Chronic hepatitis B virus (HBV) infection is a dynamic state of interactions among HBV, the hepatocytes and the immune system of the patient. Perinatally or early childhood-acquired chronic HBV infection has a long ‘immune tolerant phase’, when patients are young, and HBeAg seropositive with a high viral load but with no significant liver disease. Persistent or episodic liver injuries during the ‘immune clearance phase’ may lead to decompensation, fibrosis progression or cirrhosis development in some patients, but may eventually lead to HBV-DNA seroclearance with HBeAg seroconversion and entry into the ‘inactive phase’ with remission. Hepatitis may relapse, because of reactivation of HBV with precore or basal core promptor mutations, and develop ‘HBeAg-negative chronic hepatitis’, in some patients. In contrast, HBsAg seroclearance may occur in those with sustained remission. During the course, HBV replication is the key driver of disease progression including development of cirrhosis and hepatocellular carcinoma (HCC). Among the currently available anti-HBV drugs, the most extensive and longest experience has been gained with conventional interferon (IFN)-α and lamivudine. A finite course of IFN therapy has long-term benefit in achieving a cumulative response, increasing HBsAg seroclearance and reducing cirrhosis and/or HCC. Maintained virological response to lamivudine therapy has a similar long-term benefit in reducing disease progression. Pegylated IFN and newer nucleos(t)ide analogues may have even better long-term outcomes because of better therapeutic efficacy and/or a low risk of drug resistances. The treatment outcomes are still far from satisfactory. The development of safe and affordable anti-HBV agents/strategies is needed to further improve outcomes.
Natural history of chronic hepatitis B virus infection
Hepatitis B virus is not usually cytopathogenic by itself. Liver injuries in chronic hepatitis B are considered to be the result of the host's immune responses against HBV, such as an HLA-class I antigen-restricted, cytotoxic T lymphocyte-mediated response against HBV antigen(s) expressed on hepatocytes with resultant apoptosis and necrosis (4). Accordingly, chronic HBV infection is a dynamic state of interactions among HBV, the hepatocytes and the immune system of the patient; therefore, the natural course of chronic HBV infection can be divided into different phases (5). Adult-acquired chronic HBV infection usually has no or a very short ‘immune tolerant phase’. In contrast, the perinatally or early childhood-acquired chronic HBV infection has a long ‘immune tolerant phase’ (2, 3, 5). Patients in the ‘immune tolerant phase’ are young, HBeAg seropositive with high viral loads (>2 × 106−7 IU/ml or >107−8 copies/ml), but with a normal serum alanine aminotransferase (ALT) and near-normal liver histology (5–7).
The ‘immune clearance phase’ usually develops during adolescence or adulthood. This phase is characterized by continuing hepatitis activity or episodic acute flares with serum ALT levels over five times the upper limit of normal, sometimes complicated by hepatic decompensation (4). These events may lead to fibrosis progression or cirrhosis development in some patients during the HBeAg-positive phase, but may also result in a declining serum HBV-DNA level and may eventually lead to HBV-DNA seroclearance and HBeAg seroconversion to its antibody (anti-HBe) in most patients. The estimated annual incidence of spontaneous HBeAg seroconversion was reported to be 2–15%, depending on factors such as age, ALT level and HBV genotype (4, 8). In general, patients with genotype A and B HBV infection have HBeAg seroconversion at a younger age, as compared with those infected with genotype D and C HBV respectively (8–11).
Following HBeAg seroconversion, most of the patients enter a ‘residual inactive phase’ with sustained normal serum ALT, low serum HBV-DNA (<2000 IU/ml) and no or minimal necro-inflammatory histological changes, although some of them may have developed advanced fibrosis or cirrhosis (5, 12–14). Spontaneous HBsAg seroclearance may occur several years after HBeAg seroconversion. Short-term (3–5 years) follow-up studies showed that adult-acquired chronic HBV infection appeared to have a higher HBsAg seroclearance rate (2). A recent long-term (median 23 years) follow-up study following HBeAg seroconversion in 40 Italian patients showed an annual incidence of 2.1% (14). A large long-term (mean 11 years) follow-up study in 1965 asymptomatic anti-HBe-positive subjects [16–76 (median 34) years of age] in Taiwan showed an overall annual HBsAg seroclearance rate of 1.2%, which increased to 1.8% in those over age 50 at entry, and a cumulative incidence of HBsAg seroclearance of 8% at 10 years, which increased disproportionally to 25% at 20 years and 45% at 25 years of follow-up (15).
However, active hepatitis may relapse because of reactivation of HBV with either HBeAg seroreversion (HBeAg-positive chronic hepatitis) or with precore or basal core prompter (BCP) mutations that abolish or downregulate the production of HBeAg (HBeAg-negative chronic hepatitis) (12, 14). The estimated annual incidence of hepatitis relapse following HBeAg seroconversion was 2.2–3.3% (8, 12), being higher in males, genotype C-infected patients and those in whom HBeAg seroconversion occurred after age 40 (8, 16). In a follow-up study in ‘inactive carriers’ in whom onset of HBeAg seroconversion was unknown, the incidence was 1.5%/year and significantly lower (0.9%/year) in patients <30 years of age (17). All these findings suggest that earlier HBeAg seroconversion or a shorter HBeAg-positive phase is associated with a higher chance of sustained remission. The incidence of hepatitis relapse was higher in European patients as demonstrated in a Greek study involving 85 ‘inactive carriers’ during a median follow-up of 36 (12–48) months; the cumulative incidence of HBeAg-negative hepatitis was 24% at the end of the 4th year (18). Asymptomatic HBeAg-negative subjects with HBV-DNA >2000 IU/ml and normal ALT may develop hepatitis flares and disease progression, similar to their HBeAg-positive counterparts (8, 12, 19–21). Because the immunopathogenesis of HBeAg-negative hepatitis is similar to that of HBeAg-positive hepatitis, this reactive phase can be viewed as a variant of the immune clearance phase (3). The characteristics of the different phases of chronic HBV infection are shown in Table 1.
|Age (years)*||< 20–25||20–40||>35–40||>35–40|
|HBV-DNA (IU/ml)||>2 × 106−7||>2 × 104−5||<2 × 103||>2 × 103−4|
|Fibrosis score (Ishak)||0–1||2–6||0–6||2–6|
|Precore/core promoter||Wild type||Wild type>mutant||Mutant>wild type||Mutant≫wild type|
Disease progression during chronic hepatitis B virus infection
Patients in the immune tolerant phase have no or minimal disease progression as long as serum ALT remains normal (6). In contrast, an earlier 3-year clinical study in patients with chronic hepatitis B, or patients in the immune clearance phase, showed that cirrhosis developed at an estimated annual incidence of 2.1%, being higher in those seropositive for HBeAg at entry (2.4%/year) (22). A recent long-term (1.1–16.5, median 6.8 years) follow-up study further revealed that chronic hepatitis B patients with persistent HBeAg seropositivity developed cirrhosis at a higher incidence of 3.5%/year (23). Among patients who had undergone HBeAg seroconversion, the incidence of cirrhosis development was highest in those with HBeAg reversion, followed by those who developed HBeAg-negative hepatitis (2.9%/year during 1–18 years), and almost zero (1 of 184) in those with sustained remission (12). In agreement with this trend, a 7-year prospective study in 1495 asymptomatic male HBsAg carriers (80% seronegative for HBeAg) showed a low annual incidence of 0.7%, and that age, HBeAg seropositivity and ALT elevation >6 months were independent factors for cirrhosis development (24). Studies in asymptomatic ‘inactive carriers’ of mainly adult-acquired chronic HBV infection also showed that the risk of liver cirrhosis was very low (20, 25). In addition to age and HBeAg serostatus, the frequency of flares, the extent, the severity and the duration of hepatic lobular alterations were also found to be factors for disease progression and outcomes (22). Consistent with these observations, a recent Korean long-term (mean: 10 years) follow-up study involving 188 patients (52 HBeAg-negative patients) has shown that age and persistent ALT elevation are independent factors for the development of cirrhosis, decompensation and HCC (26).
About 50% of the patients are still seropositive for HBeAg and/or HBV-DNA at their onset of cirrhosis. Further disease progression after the development of cirrhosis may continue to occur in these patients (27, 28). HCC develops at an annual incidence of 3–6% in patients with cirrhosis and far less frequently in noncirrhotic patients (12, 27, 28). HBsAg seroclearance usually confers an excellent prognosis but HCC may still occur although at a very low rate and usually in patients in whom cirrhosis or superinfection with other virus(es) had already developed before HBsAg seroclearance (29, 30). Conceivably, factors for cirrhosis development are also factors for the development of HCC. Large community-based studies have confirmed that age, sex, HBeAg serostatus and ALT levels are factors for the development of liver cirrhosis and HCC. These studies further indicate that serum HBV-DNA level is associated with cirrhosis and HCC development in a dose-dependent manner starting from serum HBV-DNA level >2000 IU/ml (31–35). These findings suggest that HBV replication, with subsequent immune-mediated liver injuries, is the primary driving force for liver disease progression (36).
Several studies in Asian patients have shown that genotype B HBV infection, compared with genotype C, is associated with spontaneous HBeAg seroconversion at a younger age, less active liver disease, slower progression to cirrhosis and less frequent development of HCC (8–11, 37–39). In general, genotype D is associated with a less favourable prognosis than genotype A (39). A study from India indicated that genotype D is more often associated with HBeAg-negative chronic hepatitis B, more severe liver diseases and may predict the occurrence of HCC in young patients (40). Patients with genotype C HBV infection have a higher HBV-DNA level, higher frequency of pre-S deletions, a higher prevalence of BCP A1762T and/or G1768/A mutations and A1762T/G1764A double mutations than patients infected with genotype B HBV and have a significantly higher chance of developing HCC (9, 31, 35, 38, 39, 41, 42). A recent study revealed that a complex mutation pattern rather than a single mutation was associated with disease progression (42). These kind of data are not available in populations other than Asians. The role of these naturally occurring HBV mutations in the pathogenesis of liver disease progression requires further studies.
Besides viral factors (viral load, genotype and genomic mutations) and host factors (age, sex and immune status), other factors may also contribute to the progression of liver disease. These include habitual alcohol consumption, cigarette smoking exposure to aflatoxin and other viral superinfections (43), as shown in Figure 1.
Antiviral therapy for chronic hepatitis B virus infection
The observations on the natural history of chronic HBV infection suggest that active HBV replication is the key to the subsequent HBV-related immune clearance events that determine the clinical outcomes (36). HBV elimination or permanent HBV suppression, spontaneously or by antiviral therapy, should theoretically be able to reduce the risk of or slow the progression of liver disease.
Currently, conventional interferon (IFN)-α, pegylated interferon (peg IFN), lamivudine, adefovir, entecavir, telbivudine and tenofovir have been approved for the treatment of chronic HBV infection. The most extensive and longest experience with the use of these agents in controlling disease progression has been gained with IFN and lamivudine. Short-term studies have shown that IFN-based therapy is modestly efficacious in inducing HBeAg loss or seroconversion (30–40%) in HBeAg-positive patients (44–47) and in producing sustained HBV DNA suppression (20–30%) in HBeAg-negative patients (48–51). Therapy with direct antiviral agents may suppress HBV replication and contribute to significant improvement in liver disease, including reversion of fibrosis (52–56). A critical review of 26 therapeutic trials involving 3428 patients showed that treatment-induced HBV DNA reduction correlated significantly and consistently with histological, biochemical and serological responses, especially in studies using direct antiviral agents and in HBeAg-positive patients (57).
Long-term outcomes under therapy
Long-term follow-up studies after a 4–6-month course of IFN therapy in HBeAg-positive patients showed a reduction in fibrosis progression and the risk of cirrhosis in sustained HBeAg seroconverters only. These studies failed to show significant reduction of cirrhosis as compared with untreated patients (44, 45), probably because of a small sample size or the small increase in the HBeAg seroconversion rate in the IFN-treated patients (46). Nevertheless, the development of cirrhosis and its complications were significantly reduced in HBeAg-seroconverted patients as compared with patients who had persistent HBeAg seropositivity (44–46). Furthermore, a recent large study comparing 233 IFN-treated HBeAg-positive patients with 233 well-matched (age, sex, ALT, HBeAg status, histology and length of follow-up) untreated patients did show a reduced cumulative incidence of cirrhosis (17.8 vs 33.7% in the controls; P=0.041) and HCC (2.7 vs 12.5% in the controls; P=0.011) during a median follow-up period of 6.8 (1.1–15.5) years (23). Studies also showed that sustained elimination of HBeAg was associated with a significant increase in survival (23, 44–46) and a reduction in the occurrence of severe cirrhotic complications and the need for liver transplantation (44).
In HBeAg-negative European patients treated with IFN for 6–24 months, sustained responders also showed decreased progression of the Ishak fibrosis score (48, 49) or a decreased risk of cirrhosis (58). Sustained responders also had significantly improved long-term outcomes, including less severe cirrhosis-related complications, reduced incidence of HCC (1.8 vs 10.5% in relapsers; P=0.027 and 7.7% in untreated; P=0.048), less need for liver transplantation and lower mortality, although the sustained response rate in HBeAg-negative patients was usually <30% (48, 49, 58). HBsAg seroconversion occurred more frequently in IFN-treated patients with a sustained response (46, 48–50) and was associated with a lower rate of hepatic decompensation and with longer survival (50).
In patients with cirrhosis, a retrospective cohort study of European patients (82 IFN-treated vs 196 untreated) showed that the treated patients had a higher likelihood of HBsAg loss (5-year probability, 16 vs 4%; P=0.0021), which was associated with a lower incidence of HCC and longer survival (50). A meta-analysis of seven nonrandomized-controlled trials in cirrhosis patients (1505 patients with 122 observed HCC cases) suggested that IFN-α therapy decreased the incidence of HCC, but the difference was significant in only three trials; the pooled estimate, however, was significantly in favour of the preventive effect of IFN-α therapy. The inconsistency among the trials was a major problem in reaching a firm conclusion (59). Recently, it has been shown that IFN therapy in compensated cirrhotic patients is safe and even more effective than noncirrhotic patients (27). This finding suggests that the benefit of reducing HCC in IFN-treated cirrhotic patients might be evident upon a longer follow-up. A subgroup analysis in a recent long-term follow-up study did show that HCC incidence was reduced significantly in IFN-treated cirrhotic patients (23).
Long-term data on the effect of direct antiviral agents on the risk of developing cirrhosis and further disease progression are limited. Long-term (>3 years) lamivudine or adefovir therapy showed reversal of advanced fibrosis (53, 54). A long-term lamivudine (median 89.9 months; range 26.5–128.3 months) therapy involving 142 HBeAg-positive, noncirrhotic patients from Hong Kong demonstrated a significantly lower cumulative rate of cirrhosis and/or HCC development (P=0.005) as compared with 124 HBeAg-positive untreated controls (60). A double-blind, randomized-controlled trial demonstrated that maintenance lamivudine therapy for a median of 32.4 months in 436 patients with cirrhosis or advanced fibrosis (Ishak fibrosis score ≥ 4) significantly reduced overall disease progression, as compared with 215 untreated controls (61). Long-term lamivudine therapy in 303 HBeAg-negative patients with cirrhosis also showed that patients who maintained a virological response were less likely than those with viral breakthrough to show worsening of liver disease (62).
The incidence of HCC was also significantly reduced (3.9 vs 7.4% in the placebo group, P=0.048) in the lamivudine-treated patients with advanced fibrosis or cirrhosis (61). A study of 377 patients (51% HBeAg-negative, 17% with cirrhosis) treated with lamivudine for up to 96 months (23.119.0 months) in Japan also showed a marked reduction in the incidence of HCC as compared with a historical control group matched for age, sex, hepatic fibrosis score, albumin level and platelet count (0.4 vs 2.5%/year; P<0.001) (63). A retrospective multicentre study involving 656 HBeAg-negative patients (353 with chronic hepatitis, 303 with liver cirrhosis) treated with lamivudine for 1–66 months (median, 22 months) also showed that HBV suppression reduced HCC development, even in patients with liver cirrhosis (62). However, long-term lamivudine therapy was associated with a high rate of drug-resistant mutations (60–62). These patients were more likely to experience disease progression and to die because of reasons related to the worsening of liver function (61). In HBeAg-negative patients, the chance of developing HCC was significantly greater in patients experiencing a virological breakthrough than those who maintained viral suppression (62).
These results have proven the concept that suppression of HBV reduces the risk of or slows the progression of liver disease, and that resumption of HBV replication may restore the potential for disease progression. The adverse effect of lamivudine mutations can now be overcome by the timely use of rescue therapy (64–66). This was well demonstrated in a trial of long-term therapy started with lamivudine and rescued by adefovir (67). It is anticipated that long-term therapy with telbivudine or adefovir or entecavir or tenofovir would have similar or even better long-term outcomes because of their higher genetic barrier to drug resistance.
Summary and perspectives
Clinical and epidemiological studies have shown that HBV replication, particularly at HBV DNA levels ≥ 104−5 copies/ml, is the key driver of hepatitis activity and subsequent disease progression in patients with chronic HBV infection. Overall long-term data show that IFN-α therapy in HBeAg-positive patients results in cumulative HBeAg seroconversion, increasing HBsAg seroclearance and reducing cirrhosis and/or HCC development, especially in those demonstrating a sustained response. However, such long-term efficacy is less conclusive in HBeAg-negative chronic hepatitis B patients, especially those with cirrhosis, although sustained responders or those lost HBsAg have better outcomes. Because peg IFN is more effective than conventional IFN in the treatment of chronic hepatitis B (68) and may result in a complete response with HBsAg seroconversion (47, 51), long-term follow-up studies will likely prove that peg IFNs have similar or even better effects.
Data also suggest that maintained suppression of HBV replication using nucleos(t)ide analogue(s) may reduce the worsening of liver fibrosis, reverse advanced fibrosis, reduce the cirrhosis development and prevent further disease progression including HCC in patients with advanced fibrosis or cirrhosis. The main problem is the emergence of drug resistance and resumption of HBV replication, which may negate the therapeutic benefits. This problem can now be overcome by the timely use of rescue drug(s) (69). Long-term studies will undoubtedly confirm that the effects of adefovir or entecavir or tenofovir in reducing disease progression are similar to or better than those of lamivudine because of a much lower risk of drug resistance.
However, current therapy can only reduce the risk of or slow the disease progression, not prevent all adverse sequelae. Surveillance of HCC using ultrasonography supplemented with an α-fetoprotein assay is required to improve outcomes by increasing the rate of early detection and the chance of curative treatment (70). The development of safe and affordable agents and the establishment of management strategies capable of producing a better sustained or maintained HBV suppression without inducing drug resistance would be the ultimate goal in the management of chronic HBV infection.
The authors are grateful for the long-term grant support provided by Chang Gung Medical Research Fund and the Prosperous Foundation, Taipei, Taiwan, and the excellent assistance of Ms Su-Chiung Chu. The author has acted as a consultant for BMS, GSK, Novartis, Roche, and SciClone and has received grant/research support from BMS, Novartis, Roche, SciClone, and Gilead.
Conflicts of interest
The author has not declared conflicts of interest.