Hepatitis flares and hepatitis B e antigen seroconversion: Implication in anti-hepatitis B virus therapy1

Authors


  • 1

    Presented as a Journal of Gastroenterology and Hepatology lecture at the World Congress of Gastroenterology in Bangkok, February 2002.

Dr Y-F Liaw, Liver Research Unit, Chang Gung University, 199, Tung Hwa North Road, Taipei, Taiwan. Email: liveryfl@so-net.net.tw

Abstract

Abstract  Hepatitis flares or acute exacerbations, defined as an abrupt elevation of serum alanine aminotransferase (ALT) over fivefold the upper limit of normal (ULN), of chronic hepatitis B virus (HBV) infection are the results of HLA-I restricted, cytotoxic T lymphocyte (CTL)-mediated immune response against HBV and its downstream mechanisms. Higher ALT levels reflect a more vigorous immune response and a more extensive hepatolysis that, in the extreme situation, may lead to decompensation and failure. In contrast, higher ALT also reflects a more robust immune clearance of HBV and, therefore, a higher chance of HBV-DNA loss and hepatitis B e antigen (HBeAg) seroconversion, both in the setting of natural course and drug therapy. Alanine aminotransferase of fivefold the ULN appears to be a significant cut-off level to categorize the patients in terms of endogenous immune response against HBV. Patients with ALT levels less than fivefold the ULN or those with a less vigorous immune response require immunomodulation to induce robust immune response to enhance HBV clearance. In contrast, those with a more vigorous immune response or those with ALT flare over fivefold the ULN should be monitored closely for spontaneous HBV clearance/HBeAg seroconversion or to start direct antiviral therapy in time to prevent the occurrence or deterioration of hepatic decompensation. In conclusion, a better understanding of the pathogenetic mechanisms and natural course of hepatitis flares, wiser selection of patients and the timing of drug therapy are crucial to achieve better treatment results.

INTRODUCTION

Chronic hepatitis B virus (HBV) infection is a serious clinical problem caused by its worldwide distribution and possible adverse sequelae, such as hepatic decompensation, cirrhosis and/or hepatocellular carcinoma.1 Chronic HBV infection is a dynamic state of interactions among HBV, hepatocytes and the immune system of the host. Accordingly, the natural course of chronic HBV infection is characterized by a series of hepatitis flares and remissions.2 Superinfection with other hepatotropic virus(es) or superimposed toxic or chemical hepatic injuries may also cause hepatitis flares in patients with chronic HBV infection,3 and may suppress HBV replication, thus leading to hepatitis B e antigen (HBeAg) and even hepatitis B surface antigen (HBsAg) loss.4 As superinfection leading to HBV clearance and HBeAg seroconversion is not an intended intervention, it will not be covered in this review. This paper will aim to examine the implication of hepatitis flares related to HBV itself (acute exacerbation) in anti-HBV therapy.

DEFINITIONS OF HEPATITIS FLARE OR EXACERBATION

In early 1980s, when chronic hepatitis was classified into chronic active hepatitis (CAH), chronic persistent hepatitis (CPH), chronic lobular hepatitis (CLH) and non-specific reactive hepatitis (NSRH), a clinicopathological study done in our unit showed that the vast majority of inactive hepatitis (CPH and NSRH) had an alanine aminotransferase (ALT) level below 200 U/L, which is fivefold the upper limit of normal (ULN), while active hepatitis (CAH and CLH) had an ALT value far over this cut-off level.5 Based on these findings, we defined acute exacerbation or hepatitis flares as ‘an abrupt elevation of ALT over 300 U/L in patients with a baseline ALT level below 200 U/L or fivefold the ULN’.6,7 Lok and Lai later arbitrarily defined acute exacerbation or hepatitis flare as ‘an abrupt elevation of serum ALT to beyond 200 U/L (over fivefold the ULN) or a greater than threefold increase in ALT, whichever is higher’.8 More recently, a definition of ‘intermittent elevations of aminotransferase activity to more than 10-fold the ULN and more than twice the baseline value’ was adopted at the National Institute of Health workshop on management of hepatitis B, 2000.9 All of the above definitions agree that an ALT elevation over 200 U/L or fivefold the ULN is the minimum criterion.

CLINICAL AND PATHOLOGICAL PRESENTATIONS OF HEPATITIS FLARE

Acute exacerbation or hepatitis flare occurs from time to time during the natural course of chronic HBV infection; particularly frequent in HBeAg-positive patients.7,10 The clinical spectrum of acute exacerbation or hepatitis flare varies from totally asymptomatic to typical symptoms of acute hepatitis with the extreme manifestation of hepatic decompensation or hepatic failure.2,7,10 In addition to ALT elevation to over fivefold the ULN, serum bilirubin may increase and the prothrombin time may be prolonged. These abnormalities, however, are generally less severe than those of acute hepatitis or acute superinfections.3

A surge of the serum HBV-DNA level usually occurs prior to the abrupt elevation of ALT, which is followed by a precipitous decline of serum HBV-DNA.11,12 In HBeAg positive patients, seroclearance of HBeAg and the seroconversion to its antibody (anti-HBe) may follow (Fig. 1). More severe acute exacerbations are associated with an elevation of serum α-fetoprotein (AFP), of which the peak level is determined by three to four weekly measurements after the onset of abrupt ALT elevation (Fig. 1). There is usually a 1–2 week time lag between the peak levels of ALT and AFP.2,13

Figure 1.

Clinical course of a hepatitis B e antigen (HBeAg)-positive patients with hepatitis flare. There is a surge of serum hepatitis B virus (HBV)-DNA level prior to the abrupt elevation of serum alanine aminotransferase (ALT) level, which is followed by a precipitous decline of serum HBV-DNA and subsequent HBeAg seroconversion to its antibody (anti-HBe). Note that the ALT flare is associated with an elevation of α-fetoprotein (AFP). There is a time lag of 1–2 weeks between the peak levels of ALT and α-fetoprotein (AFP). A liver biopsy (arrow) showed bridging hepatic necrosis (BHN). Bil, bilirubin; PT, prothrombin time prolongation. (▴) HBV-DNA, (•) ALT, (▪) AFP.

Liver biopsies obtained during acute exacerbation invariably show lobular necroinflammatory changes, which are distributed unevenly and may be so extensive that bridging hepatic necrosis (BHN) may occur.7,10 A close correlation is observed between serum AFP levels and the presence of BHN during acute exacerbation, as BHN is evident in greater than 80% of the patients with AFP > 100 ng/mL during hepatitis flares.2,13 Therefore, AFP > 100 ng/mL can be used as a marker of BHN.

Immunopathological studies during acute exacerbation have shown that the cell infiltrates, at the site of necroinflammatory reaction, are mainly cytotoxic T lymphocytes (CTL) or CD8+ T cells.14,15 Strong simultaneous expressions of class I human leukocyte antigen (HLA-I) and intercellular adhesion molecular I (ICAM-I) on the hepatocyte membrane occur at the site of necroinflammation and active cellular infiltration.16,17 Hepatitis B core antigen (HBcAg) and HBeAg are also present, predominantly in the cytoplasm or sub-membranous/membranous region of hepatocytes,18–23 while HBsAg, large protein (pre-S1) and the middle protein of the envelope antigen of HBV (pre-S2) are expressed in the membranous region.24

PATHOGENESIS OF HEPATITIS FLARE AND SPONTANEOUS HEPATITIS B E ANTIGEN SEROCONVERSION

Spontaneous HBeAg seroconversion to anti-HBe is a very important event in the natural history of chronic HBV infection because it represents a locking point after which HBeAg reversion rarely occurs (< 5%); and sustained remission with good prognosis usually follows.25‘Hepatitis B virus-DNA loss with HBeAg seroconversion’ has, therefore, been used as an end-point of the complete response to therapy in HBeAg-positive patients.

Studies have shown that temporary exacerbation of hepatitis is a frequent event before spontaneous HBeAg seroconversion,6,8,11 and that such an event represents ‘immune clearance’ of hepatocytes containing replicating HBV or those expressing HBcAg/HBeAg.6,20 In addition, T cell responses to HBcAg and HBeAg usually increase before, or coincide with, the surge of serum HBV-DNA and ALT flare-ups, and correlate with the elevation of HBcAg/HBeAg-specific precursor T cell frequencies.26,27 Together with the immunopathological findings during hepatitis flares and the findings of autologous hepatocytotoxicity assays,16 this evidence suggests that hepatitis flares are the results of HLA-I restricted, CTL-mediated immune response against HBV antigen(s) and its downstream apoptotic mechanisms.2 Furthermore, CTL also secrete cytokines including interferon (IFN)-γ and tumor necrosis factor (TNF)-α that abolish HBV gene expression and viral replication.28 Therefore, HBV-DNA clearance and HBeAg seroconversion to its anti-HBe will occur eventually.

In view of the immune mechanisms of ALT or hepatitis flare, the ALT levels may represent the magnitude of the immune response against HBV. It is then conceivable that patients presenting with higher ALT have a more vigorous endogenous immune response against HBV, and therefore have a higher chance of spontaneous HBeAg seroconversion. In fact, a study of a large number of patients from our unit has shown that those with serum ALT levels over fivefold the ULN at entry had a spontaneous HBeAg seroclearance rate over 50% at the end of a 12-month follow up, and over 60% at the end of a 18-month follow up, a sharp contrast to < 5 and < 10%, respectively, in patients with entry ALT levels less than fivefold the ULN, as summarized in Table 1.29 These findings suggest that patients with serum ALT levels over fivefold the ULN are different from those with serum ALT levels less than fivefold the ULN, at least in terms of endogeneous immune response against HBV. On a short-term scale, approximately two-thirds of spontaneous HBeAg seroconversions are preceded by hepatitis flares within 3 months.6,11 However, less than 25% of hepatitis flares were followed by spontaneous HBeAg/HBV-DNA seroclearance within this time scale unless the acute exacerbation was more severe, as reflected in an AFP elevation over 100 ng/mL and/or the presence of BHN (Table 1).7,30 These possibilities of spontaneous HBeAg seroconversion should be considered when designing the therapeutic strategy.

Table 1.  Hepatitis flare and spontaneous hepatitis B virus (HBV)-DNA/hepatitis B e antigen (HBeAg) seroclearance
 Hepatitis flareHBV-DNA/HBeAg clearance (%)
3 months12 months18 months
  1. AFP, α-fetoprotein; ALT, alanine aminotransferase; BHN, bridging hepatic necrosis; (–) no data available. *, **, ***, ****, *****P < 0.001.

 > fivefold ULN 16*52**68***
 < fivefold ULN< 2*  5**  8***
 >100 ng/mL 72***
 <100 ng/mL 18****
 Yes 67*****
 No 16*****

THERAPY-RELATED HEPATITIS FLARE AND HEPATITIS B E ANTIGEN SEROCONVERSION

Given the pathogenetic mechanisms of hepatitis flare, it is conceivable that hepatitis flare may occur in relation to drug therapy, particularly if the drug has immunomodulating effects. The best known example is the ALT flares during the second to third month of α-interferon therapy.31 A more obvious example is the ALT rebound following the abrupt withdrawal after a short course of corticosteroid therapy.32 A retrospective analysis of a randomized placebo-control trial conducted in the USA33 showed that hepatitis flares were documented in 70% of the patients receiving corticosteroid priming prior to IFN therapy, in 45–49% of the patients receiving IFN alone, and in 26% of the control patients.34 This retrospective analysis confirms an earlier observation that flares are associated with a higher likelihood of a virological response.33 It further indicates that among the independent predictors of a virological response (high baseline HBV-DNA, high pretherapy ALT, moderate and severe ALT flare), severe flare is the most powerful predictor of a sustained response.34 In thymosin-α1 therapy, an immunomodulating agent able to enhance T-helper type I (Th1) immune response, hepatitis flare followed by HBV-DNA/HBeAg seroconversion was also documented.35 Even in the use of direct antiviral therapy such as lamivudine, which has no direct immunomodulating effect, patients with pretherapy ALT levels greater than fivefold the ULN have the highest response rate (up to 65% in 1 year and 80% in 2 years), suggesting that patients with a stronger endogenous immune response against HBV have a better response to direct antiviral agents.36–38 All these findings are consistent with the natural events (hepatitis flares during the natural course of chronic HBV infection), in that an induction of a robust flare may enhance viral clearance and HBeAg seroconversion.

However, like the natural events, not all therapy-related hepatitis flares are followed by HBV clearance. One study has shown that only the ALT flares with a substantial increase of interleukin-12 are associated with the induction of Th1 cytokine response, and ultimately the virological response in IFN therapy.39 Similarly, ALT flares with a Th1 response, following a short course of corticosteroid priming, are associated with a virological response to lamivudine therapy.40 Such findings are somewhat similar to the finding that Th1 response is dominant in self-limited acute HBV infection.41 A recent study has further shown that activation of Th1 immunity, with enhancement of CTL activity, is a common immune mechanism for the successful treatment of hepatitis B, including IFN-α, thymosin-α1 and lamivudine therapy.42 It seems that hepatitis flare caused by CTL destruction of the infected hepatocytes, per se, is not enough to clear HBV. Without an enhanced TH1 response, the attempt of HBV clearance will be abortive, while the substantial induction of a Th1 cytokine response with non-cytolitic HBV suppression, which is much stronger than the cytolytic pathway, will lead to a successful HBV/HBeAg seroclerance.

LAMIVUDINE-RELATED HEPATITIS FLARES

Although lamivudine is a potent inhibitor of HBV replication, it takes a long time to achieve HBeAg seroconversion, except in patients with high pretherapy ALT levels.36–38 Hepatitis flares may occur after stopping lamivudine therapy in the majority of patients who remained HBeAg positive.43 Out of our 123 patients who remained HBeAg positive, 75% experienced such hepatitis flares within 1 year (70% of the flares occurred within 3 months, and an additional 20% between 3 and 6 months) after cessation of lamivudine therapy (YF Liaw and RN Chien, unpubl. data, 2002). Even in patients who had undergone a HBeAg seroconversion during lamivudine therapy, hepatitis flares with or without HBeAg reversion may occur if the duration of lamivudine therapy after HBeAg seroconversion is shorter than 4–6 months.44 Of our 115 patients who stopped lamivudine therapy within 3 months after HBeAg seroconversion, 40% developed HBeAg reversion with hepatitis flares and 14% developed HBeAg negative but HBV-DNA positive hepatitis flares (90% within 6 months) after stopping lamivudine therapy (Liaw and Chien, unpubl. data, 2002). Obviously, long-term lamivudine therapy is usually required in patients with chronic HBV infection.

Unfortunately, the drug-resistant mutations of HBV with amino acid substitution over tyrosin-methionine-aspartate-aspartate (YMDD)-motif of the RNA-dependent DNA polymerase start to emerge after 6–9 months of lamivudine therapy.45 Studies have shown that hepatitis flares develop frequently following the surge of mutant HBV viremia during continued lamivudine therapy.46–50 Our most recent study in 66 patients with YMDD mutants has shown that 67% of these patients experienced hepatitis flare with ALT > fivefold the ULN, 39% with ALT > 10-fold the ULN and 14% with serum bilirubin over 2 mg/dL, despite continuing lamivudine therapy.51 Such flares are associated with hepatic decompensation at a higher incidence and may appear to be more severe than those observed during the natural course of wild-type HBV chronic infection (Table 2).45,46 Although such hepatitis flares may be followed by HBeAg/HBV-DNA seroclearance and/or immunoclearance of the mutants,45 new and distinct mutant HBV may be selected and another hepatitis flare may be elicited.52 Therefore, the benefit of prolonged lamivudine therapy must be carefully weighed against the concern about YMDD mutations and the durability of therapeutic response, at least before effective drugs for YMDD mutants become available.

Table 2.   Lamivudine therapy-related hepatitis flares
 Hepatitis flare (ALT > fivefold the ULN) in 1 year
No.ALT > 10-fold ULNALT > 1000 U/LBil > 2 mg%PT >3 s
  1. Bil, bilirubin; m, mutant on continuing lamivudine therapy; PT, prothrombin time prolongation; ULN, upper limit of normal; w, wild type, after stopping lamivudine therapy. *P < 0.02.

YMDDm  44  26 (59%)  7 (16%)  9 (21%)  7 (16%)*
YMDDw310233 (75%)88 (28%)70 (23%)16 (5%)*

IMPLICATIONS AND PERSPECTIVES

In conclusion, acute exacerbations or hepatitis flares occur frequently in patients with chronic HBV infection, particularly in those seropositive for HBeAg, and such ALT/hepatitis flares are the results of HLA-I restricted, CTL-mediated immune response against HBV. As the ALT levels reflect the magnitude of the immune response, higher ALT levels represent more hepatolysis that, in the extreme situation, may lead to hepatic decompensation and failure. In contrast, higher ALT levels also reflect a more vigorous immune clearance of HBV with a greater chance of HBV-DNA loss with HBeAg seroconversion, both in the setting of natural course and drug therapy. Alanine aminotransferase levels fivefold the ULN appears to be a significant cut-off level to categorize the patients in terms of endogenous immune response against HBV. With regard to drug therapy, patients with ALT levels less than fivefold the ULN or those with a less vigorous immune response require immunomodulation, such as IFN-α, thymosin-α or corticosteroid, to induce a robust immune response to enhance HBV clearance. In contrast, those with a more vigorous immune response or those with hepatitis flare or ALT levels over fivefold the ULN should be monitored closely for spontaneous HBV clearance/HBeAg seroconversion or to start direct antiviral therapy in time to prevent the occurence or deterioration of hepatic decompensation.53 Alanine aminotransferase flares, with decreasing serum HBV-DNA levels, represent effective immune clearance of HBV, and this is likely to be followed by spontaneous HBeAg seroconversion. Those patients with stable or increasing serum HBV-DNA levels reflect the ineffective immune clearance of HBV, and they are at the risk of developing further hepatocytolysis and hepatic decompenastion, which will eventually require treatment (Fig. 2). In conclusion, a better understanding of the pathogenetic mechanisms and natural course of hepatitis flares, wiser selection of patients and the timing of drug therapy are crucial to achieve better treatment results.

Figure 2.

 Changes of serum alanine aminotransferase (ALT) levels and hepatitis B virus (HBV)-DNA in two hepatitis B e antigen (HBeAg)-positive patients. The ALT flare with decreasing HBV-DNA level represents effective immune clearance of HBV, and is therefore likely to be followed by spontaneous HBeAg seroconversion (a; patient a). In contrast, ALT flare with increasing HBV-DNA levels represents an ineffective immune clearance of HBV, and the patient is at the risk of deterioration and needs immediate fast-acting therapy (b; patient b). (bsl00003) HBeAg-positive; (bsl00004) normal range of ALT; (bsl00005) lamivudine therapy; (bsl00023) anti-HBe-positive; (□) HBeAg-negative, anti-HBe-negative; (•) ALT; (▴) HBV-DNA.

ACKNOWLEDGMENTS

The authors are grateful for the long-term grant support from the Chang Gung Medical Research Fund (CMRP800), and the excellent assistance of Ms SC Chu at this unit.

Ancillary