Natural history of chronic hepatitis B in co-infected patients
Journal of Hepatology
Volume 44, Issue (Supplement 1), Pages S65-S70 (2006)
Massimo Puotia, Carlo Tortia, Raffaele Brunob, Gaetano Filiceb, Giampiero Carosia
a Clinica di Malattie Infettive e Tropicali, AO Spedali Civili, Universita di Brescia, P.zzle Spedali Civili 1, I 25123 Brescia, Italy
b Divisione di Malattie Infettive e Tropicali, IRCCS Policlinico S. Matteo, Universita di Pavia, Italy
HIV co-infection influences the course and natural history of hepatitis B virus (HBV) infection by impairing the quantity and quality of the innate and adaptive immune response. The rates of spontaneous resolution after acute infection and spontaneous anti-HBe and anti-HBs seroconversions are decreased, and levels of HBV replication are increased in HIV-infected patients. A more rapid progression of liver fibrosis and a higher rate of cirrhosis decompensation (but not hepatocellular carcinoma) have been demonstrated in co-infected patients. The risk of HBV-associated end-stage liver disease and liver-related mortality may be increased by HIV co-infection. Antiretroviral therapy may trigger spontaneous anti-HBe and anti-HBs seroconversion and/or a better immune control of HBV replication by restoring adaptive immunity, but can also increase hepatitis flares. Reactivation of chronic hepatitis B has been observed after suspension of anti-retrovirals with anti-HBV activity or after occurrence of HBV resistance to lamivudine. Future research should focus on: the impact of HIV-induced changes in innate and adaptive immune response and modifications induced by anti-retroviral therapy that may impact on progression of advanced chronic hepatitis B; the association between HBV genotype and clinical course of disease; and the role of occult HBV infection as a co-factor with other causes of liver injury.
- 1. HIV-HBV interactions
- 2. Natural history of HBV-related disease in HIV-infected patients before HAART
- 3. Natural history of HBV-related disease in HIV-infected patients during the HAART era
- 4. Occult HBV in co-infected patients
- 5. Predictors and co-factors of disease severity
- 6. Future research needs
1. HIV-HBV interactions
The outcome of HBV infection depends on the kinetics of virus-host interactions and in particular on the strength and quality of the innate and adaptive humoral and cellular immune response . In persistently infected patients who fail to downregulate or clear HBV, the continued expression of inflammatory cytokines and recruitment of activated lymphomononuclear cells to the liver ultimately results in fibrosis, cirrhosis and hepatocellular carcinoma .
HIV establishes a chronic and latent infection that induces extensive damage of the immune system through virus-related and indirect pathogenic mechanisms . HIV-infected individuals show a quantitative depletion of CD4+ T cells but also an overall immune dysfunction that includes:
- a complex dysregulation of the cytokine network,
- a decrease of the functional capability of CD8+ CTL and, at late stages, a significant reduction in their number,
- and an aberrant activation of cells of the immune system that are severely damaged and show signs of phenotypic and functional alterations.
Fibrosing cholestatic hepatitis, a severe and rapidly progressive form of HBV infection thought to be due to a direct cytopathic effect of the virus, has anecdotally been described in HIV positive patients  and in other groups of immunosuppressed patients. However, in most cases with an absent host defence, HBV replication in the liver is not cytotoxic and non-cytopathic . In HIV-infected patients, host defences are not 'absent' even in the most advanced stages and are often aberrantly activated . In such cases, the quality of HIV-induced dysregulation of a still active anti-HBV immune response may be crucial in determining the quality of necro-inflammation and of the consequent 'quantity' of fibrogenesis.
It is conceivable that HIV co-infection may influence the course and natural history of HBV infection through its heavy modification of the immune response, but studies comparing anti-HBV innate and adaptive immune responses of HIV-infected or non-infected patients at various stages of the disease and during anti-retroviral therapy are still lacking. Most data on the impact of HIV on HBV came from longitudinal and cross-sectional case-control studies that did not include a sophisticated study of the anti-HBV adaptive and innate immune response.
2. Natural history of HBV-related disease in HIV-infected patients before HAART
There is evidence that co-infection with HIV significantly modifies the natural history of HBV infection (Fig. 1). In patients with HBV infection, HIV co-infection is associated with:
- Higher chronicity rate of acute hepatitis B [4-8].
- Higher levels of HBV replication, even in the presence of hepatitis delta virus (HDV) super-infection [4-14], related to HIV-induced CD4 depletion .
- A lower rate of spontaneous loss of HBeAg and/or HBsAg and seroconversion to anti-HBe and anti-HBs [4,5] with a high rate of seroreversion after CD4+ cell depletion in those with spontaneous or interferon-induced anti-HBe seroconversion .
There are contradictory data on the activity of inflammatory liver disease in co-infected patients. Cohort studies from northern Europe and USA of men who have sex with men (MSM) showed a significantly less severe necro-inflammatory activity in HIV co-infected patients [7,10,16,17]. In contrast, some studies performed in Californian and French cohorts that included injection drug users showed increased necro-inflammatory activity in HIV seropositive subjects [18,19]. There are also contradictory data on the impact of HIV on hepatitis B progression towards cirrhosis: some studies from northern Europe and USA performed in the pre-HAART era did not reveal an unfavourable impact of HIV co-infection on hepatitis B evolution [6,7,9,10,20] while three French studies [11,15,19] identified a more rapid progression towards cirrhosis in HIV co-infected individuals. These discrepancies could be related to differences in the prevalence of infecting HBV genotypes and of mutant HBeAg defective HBV strains or the degree of immune suppression, or to the different prevalence of co-factors associated with liver injury (alcohol, HCV, HDV) in the various cohorts.
A higher rate of decompensation has been observed in HIV/HBV co-infected individuals with cirrhosis [11,15,21]. Furthermore, there is evidence in vitro and in animal models that suggests a pivotal effect of HIV on hepatitis virus-related hepatocarcinogenesis, which is mediated by Tat protein [22-24]. However, although there is a more severe presentation of hepatocellular carcinoma and a lower survival in HBsAg carriers with HIV infection , there are no data supporting an accelerated progression towards hepatic cancer in HIV/HBV co-infected people.
A recent, large case-control study by Thio et al. , measured liver-related mortality in 5293 MSM classified according to their HIV antibody and HBsAg status. Individuals co-infected with HIV and HBV were at a greater risk of liver-related death than those infected by HIV or HBV alone (or not infected with either virus). Risk of liver death in individuals with HIV was associated with HBsAg reactivity independent of age, history of drug use and number of male sexual partners . These results have been confirmed by another cohort study from the USA, which demonstrated a high liver-related mortality in HIV-infected people with chronic hepatitis B .
HIV co-infection also leads to a worsening of hepatitis B disease with an increase of liver-related mortality.
Note from Jules Levin: several recently published studies suggest HAART can slow HCV progression to a rate similar as that in monoinfected patients. One of these studies by Norbert Brau in the VA system found that only undetectable HIV RNA was associated with slowing HCV disease progression to a rate similar to that in monoinfection. Conversely, there remain numerous studies finding a significant proportion of HIV+ individuals experience accelerated HCV progression despite the use of HAART. At this point, I think we should assume that HIV accelerates HCV and treat accordingly. Two recent studies from an Italian research group and from Mark sulkowski at Johns Hopkins found paired liver biopsies performed 3-4 years apart found 28-33% of patients had accelerated HCV progression in this short period of time, of 2 stages of liver disease or more.
3. Natural history of HBV-related disease in HIV-infected patients during the HAART era
Highly active anti-retroviral therapy (HAART) may either prevent or treat opportunistic infections and some opportunistic tumours, thus decreasing the incidence of AIDS and increasing life expectancy of HIV-infected patients . The prolonged survival of HIV-infected individuals allows a longer time for cirrhosis to develop and, therefore, since the introduction of HAART, the relative proportion of deaths attributable to liver disease has been rising. This increase is due partly to an absolute decrease of classical opportunistic infections and tumours, but possibly also to an absolute increase in liver-related deaths .
HAART may have a major impact on HBV co-infection because of restoration of specific and non-specific immune responses and decrease of aberrant activation and dysregulation of the immune system [2,28]. In addition, at least three anti-retrovirals (lamivudine, tenofovir and emtricitabine) are potent inhibitors of HBV replication . HAART may therefore influence the natural history of chronic hepatitis B by:
- preventing acute infection in patients using anti-retrovirals with dual activity,
- inhibiting HBV replication after the introduction of anti-retrovirals with dual activity,
- reactivating HBV replication after withdrawal of anti-retrovirals with dual activity or emergence of HBV resistant strains,
- restoring adaptive HBV specific immune response and innate non-specific immune response,
- and reducing aberrant activation of the immune system.
Note from Jules Levin: entecavir was approved for HBV treatment in 2005. Entecavir does not have anti-HIV activity. Entecavir has potent anti-HBV activity showing median 6.98 log HBV DNA reductions in the phase III study. At the Nov 2005 AASLD meeting BMS reported after two years no resistance was observed in patients in phase III studies in treatment-naive patients. Another therapy for HBV is peginterferon. In 2005 Pegasys was approved for HBV treatment. Peginterferon is associated with side effects but does not cause drug resistance and is associated in the short term with higher rates of HBsAg clearance than orally administered drugs. Combination HBV drug therapy has been shown to perhaps reduce the incidence of drug resistance but has yet to show synergy in reducing HBV DNA, Truvada is the 2-drug combination of tenofovir+FTC and is a popular HIV treatment. It is used by HIV doctors in HAARt to treat HBV/HIV coinfected patients. And some doctors use Truvada in HBV monoinfected patients.
Prevention of acute HBV infection by administration of lamivudine has been demonstrated in a persistently exposed HIV-infected woman , but use of lamivudine as an anti-retroviral was not associated with a decreased incidence of acute hepatitis B in a large cohort of HIV-infected subjects . Tenofovir, lamivudine and emtricitabine are able to suppress HBV replication depending on immune status , despite the frequent persistence of residual replication . The prolonged suppression of HBV replication leads to histological improvement, and induces a significant decrease or normalization of aminotransferase levels and prevents disease progression to cirrhosis and end-stage liver disease [28,29]. However, anti-HBe and anti-HBs seroconversion have been observed in a minority of patients starting HAART regimens that include one or two of these drugs . In contrast, withdrawal of these drugs has been associated with reactivation of HBV infection and with flares of liver enzyme elevations and hepatic decompen-sation in patients with advanced liver disease [34,35]. The same phenomenon has been observed after the loss of lamivudine inhibition of HBV replication due to the occurrence of HBV genome mutations that are associated with resistance to lamivudine [34-40], which have also been associated with rapidly progressive disease [36,37].
Independent of the use of lamivudine, HAART may induce reconstitution of the adaptive immune response and this is a 'double-edged sword' in patients infected with HBV: on one side, anecdotal reports suggest the induction of HBV seroconversion and therefore a better control of HBV replication with or without flares of necro-inflammatory activity [41-44]; on the other side, flares of necro-inflammatory activity have been reported without modification of HBV markers and of HBV replication levels [45-48]. However, the incidence of HAART-induced HBsAg seroconversion has not been accurately determined, but occurred in one out of 24 HBsAg-positive patients prospectively followed during HAART in one of the largest published studies . Additionally, the rapid phase inhibition of HIV replication that occurs when HAART is started has been associated with a transient increase of HBV replication . This association may be related to cessation of HIV-induced secretion of cytokines with anti-viral activity and may trigger flares of necro-inflammatory activity in the very early phases of HAART.
Given the complex interactions between HAART and HBV co-infection it is not so easy to establish if HAART has a beneficial or a negative effect on chronic hepatitis B. Thio and co-workers observed an increase of liver-related mortality in HBsAg positive people living with HIV in the HAART era . They failed to observe a significant association between lamivudine use and a lower liver-related mortality because of lack of information in most patients . However, several direct and indirect data support a positive effect of HAART on the natural history of hepatitis B. In a recent analysis of data from the Italian ICONA cohort, the use of lamivudine as a part of the first HAART regimen in naive patients was independently associated with lower liver decompensation related morbidity . To complicate this matter further, low CD4+ T cell count nadir value was associated with higher liver-related mortality in HBsAg positive patients in many studies [15,21,25,26,51] while the majority of most severe cases of reactivation of hepatitis B related to immune restoration have been observed in immune suppressed subjects [32-50]. Therefore, an earlier use of HAART for prevention of severe immune dysfunction and the inclusion of a drug with dual anti-HBV activity might slow down progression of chronic hepatitis B in HIV-infected patients.
4. Occult HBV in co-infected patients
Occult hepatitis B is defined by the detection of low levels of HBV DNA in serum samples and/or within the liver in people who are HBsAg negative . In many instances occult hepatitis B is associated with positivity for hepatitis B core antibody (HBcAb) and/or hepatitis B surface antibody (HBsAb). However, no HBcAb or HBsAb could be detected in a significant proportion of individuals . Then, the optimal method for identification of occult hepatitis B infection is sensitive PCR with two sets of primers, respectively, from C and S regions of the HBV genome both on serum and liver tissue followed by sequencing of the amplified HBV genome . This method has never been used in HIV-infected patients. The use of sensitive PCR with two sets of primers on serial serum samples has identified occult HBV infection in 89.5% of 57 HIV-infected patients with isolated anti-HBc reactivity . By using the same methods, but with less sensitive PCR, Piroth et al. detected occult hepatitis B in 35% of 37 anti-HBc positive patients . HBV DNA has been assayed cumulatively in 1244 HIV-infected HBsAg negative patients by a commercial PCR assay (Roche COBAS Amplicore) using primers from core genomic region only, and it was detected only in 12 patients [54-57]. The prevalence of occult HBV infection in HIV-infected persons is still controversial because of the lack of standardization of diagnostic methods. Anecdotal and even fatal reactivation of HBV infection following severe immune depletion and/or lamivudine withdrawal have been described in HIV co-infected patients without HBsAg and with markers of past HBV infection [58,59], however, HBsAg seroreversion occurred in only one out of 98 subjects followed prospectively, at a rate of 0.23/100 patient-year . Because of methodological issues, data on the clinical impact of occult HBV infection in HIV-1 infected subjects are still controversial. In one study, occult HBV infection was associated with chronic 'transaminitis' in anti-HCV negative patients and with increased levels of ALT activity in HCV co-infected subjects ; in another study conducted on only four patients with occult B infection, this association was not confirmed .
5. Predictors and co-factors of disease severity
Older age, necro-inflammation and fibrosis stage at histology, ongoing HBV replication, HBeAg persistence and CD4 depletion, have been identified as predictors of disease severity by three different studies [15,19,26]. Alcohol consumption higher than 40g/day, hepatitis Delta and hepatitis C co-infection have been repeatedly identified as important co-factors of disease severity [15-19,21,61-66]. Chronic HBV carriers are more prone to developing hepatotoxicity following exposure to some drugs. In HBV/HIV-co-infected patients, particular attention should be paid to anti-tuberculous agents ; however, the role of this and other potential co-factors of disease severity such as steatosis, insulin resistance, iron metabolism and toxicity of illicit drugs and of drugs other than anti-retrovirals should still be assessed in HIV-HBV co-infected patients.
6. Future research needs
There are several points in the natural history of chronic hepatitis B in HIV-infected patients which deserves future studies: the impact of HIV-induced changes in innate and adaptive immune response and of their modification by anti-retroviral therapy on the pathogenesis of HBV induced liver damage, the natural history of chronic hepatitis B and cirrhosis in HIV-positive patients receiving HAART with or without drugs with dual activity, the role of occult HBV infection in the pathogenesis of liver injury and the impact of steatosis and insulin resistance on the course of chronic hepatitis B. In addition, HBV genotypes have been related to different clinical outcomes in HIV seropositives . The prevalence of HBV genotypes in HIV-infected patients seems to differ from that of the general population . Thus, studies on the impact of HBV genotypes on the clinical course of HBV infection are probably needed.