| |
Liver Disease in HIV: predictors of severity
|
| |
| |
Natural history and predictors of severity of chronic hepatitis C virus (HCV) and human immunodeficiency virus (HIV) co-infection
Journal of Hepatology
Volume 44, Issue (Supplement 1), Pages S28-S34 (2006)
Anais Vallet-Pichard, Stanislas Pol
Co-infection by the hepatitis C virus (HCV) is observed in up to 30% of HIV-infected individuals. In studies conducted in the 'pre-HAART era', the late consequences of HCV-related chronic liver disease were overshadowed by extra-hepatic causes of deaths, related to severe immune deficiency, and the impact of HCV infection on mortality of HIV-infected patients was low.
While the development of HAART has resulted in a significant decrease in morbidity and mortality amongst HIV-infected patients, this clear benefit allowed the expression of liver-related complications associated with HCV chronic infection. The impact of HCV on HIV remains debated but HIV infection significantly modifies the natural history of HCV infection.
HIV infection increases levels of HCV viraemia by 2- to 8-fold, resulting in a significant decrease in spontaneous recovery of acute hepatitis. HIV co-infection also worsens the histological course of HCV infection by increasing and accelerating the risk of cirrhosis or leading to rare but lethal fibrosing cholestatic hepatitis.
Liver disease is now one of the leading causes of morbidity and mortality in co-infected patients, even if HAART and especially protease inhibitors, may decrease the severity of the liver disease and the liver-related mortality.
Several non-exclusive pathogenic processes explain the increasing rate of liver complications associated with HCV-related liver disease.
Article Outline
- Abstract
- 1. Introduction
- 2. Pathogenic reciprocal interactions between HIV and HCV before HAART
- 2.1. Prevalence of HCV/HIV co-infection
- 2.2. Reciprocal consequences of the HIV/HCV co-infection
- 3. Pathogenic reciprocal interactions between HIV and HCV in the HAART era
- 4. Predictors and co-factors of disease severity
- 4.1. Drug-related hepatotoxicity
- 4.2. Other causes of liver damage in HIV/HCV co-infected patients
- 5. Morbidity and mortality
- 6. Conclusion
1. Introduction
Co-infections by hepatotropic viruses and HIV are frequent given the shared (sexual, mother-to-child and parenteral) routes of transmission. We shall mainly focus on the natural history of HCV-related disease in HIV-infected patients before the introduction of highly active antiretroviral therapies (HAART) [1-3] and in patients under HAART emphasising causes of death and predictors and co-factors of disease severity [4]. In studies conducted in the 'AIDS era' (pre-HAART), the late consequences of HCV-related chronic liver disease were overshadowed by extra-hepatic causes of deaths, related to severe immune deficiency, namely opportunistic infections, lymphomas or wasting syndrome [5-7] and the impact of HCV infection on mortality of HIV-infected patients was low [4].
The development of HAART (regimens composed of nucleoside reverse transcriptase inhibitors [NRTIs], protease inhibitors [PIs] and/or non-nucleoside reverse transcriptase inhibitors [NNRTIs]) resulted in a significant decrease in morbidity and mortality amongst HIV-infected patients [8-10]; this benefit allowed the expression of liver-related complications associated with HCV chronic infection which is mainly acquired before HIV infection, at least in haemophiliacs and intravenous drug users (IVDUs). Liver disease is nowadays one of the leading causes of morbidity and mortality in co-infected patients [11-14]. Several non-exclusive pathogenic processes that include drug-related hepatotoxicities, chronic HCV infection, other liver diseases such as steatosis or non-alcoholic steato-hepatitis (NASH) and other liver diseases that are common in the setting of alcohol or drug abuse explain the increasing rate of liver complications associated with HCV-related liver disease [11]. They account for around 10-15% of morbidity and mortality in HIV-HCV co-infected patients [11] with a rise paralleling that observed in the general population [15,16], even if it has been suggested that HAART and especially protease inhibitors may decrease the severity of the liver disease [17] and the liver-related mortality [18].
2. Pathogenic reciprocal interactions between HIV and HCV before HAART
2.1. Prevalence of HCV/HIV co-infection
Anti-HCV antibodies are frequently detected in HIV-infected patients (around 30% by third generation ELISA assays), especially in haemophiliacs and IVDUs (around 70-90%) and are usually associated with active infection as assessed by detectable HCV viraemia [1-3,19]. Given the good sensitivity and specificity of the third-generation tests, the risk of underestimation by delayed anti-HCV seroconversions [20] or seroreversions [21] is now low.
2.2. Reciprocal consequences of the HIV/HCV co-infection
Despite most early studies failing to show a change of the natural history of HIV infection associated with HCV infection [22-24] later studies have indicated that genotype 1b HCV may worsen the spontaneous evolution of HIV to both AIDS and death in haemophiliacs as well as in drug users [25-27]. Moreover, HCV infection (and active intravenous drug use) has been suggested as an important factor in the morbidity and mortality among HIV-1-infected patients, possibly through impaired CD4-cell recovery in HCV seropositive patients receiving potent antiretroviral therapy [28]. Several non-exclusive mechanisms potentially explain how HCV may act as co-factor of HIV disease progression including unspecific immune stimulation enhancing HIV replication, CD4 T-cell depletion reflecting infection of immune cells by HCV or HAART discontinuation favoured by the underlying liver disease (see below).
HIV infection significantly modifies the natural history of HCV infection [1-3]. HIV infection increases levels of HCV viraemia [29-32] and HIV seroconversion increase HCV-RNA titres [33]. This 2- to 8-fold increase results in a significantly lower rate of spontaneous recovery of acute hepatitis from 0% (personal data) to 25% [34], an increased risk of mother-to-child or sexual transmission (from a mean of 6-20% and from 0 to 3%, respectively) [35,36]. HIV co-infection worsens the histological course of HCV infection by increasing and accelerating the risk of cirrhosis [37-42] or leading to rare but lethal fibrosing cholestatic hepatitis which is clearly related to direct cytotoxicity of HCV with high viraemia leading to accumulation of viral proteins in the endoplasmic reticulum and hepatocyte death [43]. Indeed, the rate of cirrhosis is 2- to 5-fold increased in HIV/HCV co-infected patients as compared to HCV-infected patients and the mean time elapsed between contamination and cirrhosis is significantly reduced [44]. This increased rate of cirrhosis and shorter evolution is explained by a significantly increased yearly progression of the fibrosis score in co-infected subjects [45], which depends on the CD4-cell count (<200 cells/ml). The influence of chronic alcohol consumption or HBV-HCV mixed infections remains debated (38,45).
The more severe liver disease observed in all the situations of immune deficiency may reflect, at least partially, a deleterious impact of increased viraemia.
3. Pathogenic reciprocal interactions between HIV and HCV in the HAART era
HCV infection appears to have a limited impact on response to antiretroviral therapy: the early delayed CD4 count recovery [28] is not sustained and is not associated with increased HIV disease progression [46].
High-level expression of HCV proteins may lead to cellular injury: the expression of HCV proteins in cell lines diminishes the proliferative capacity of cells [47]. By contrast with the direct HBV cytotoxicity which can be controlled by the use of nucleoside or nucleotide RTIs (lamivudine, emtricitabine, adefovir dipivoxyl or tenofovir) [48-50], antiretroviral drugs do not control HCV infection. Only anecdotal cases of the disappearance of detectable HCV RNA have been described in patients receiving HAART (especially ritonavir) [51]. Significant restoration of CD4 and CD8 cell counts rarely decreases serum HCV load and there is no relevant decrease, as compared to baseline values, in HCV RNA loads after 3, 6, or 9 months of HAART in HCV/HIV co-infected individuals [52,53]. These data suggest that HAART does not clearly decrease the incidence of HCV-related liver disease associated with HCV replication.
A marked decrease in the CD4 cell count may modify the re-partition of HCV quasi-species with a putative risk of selection of HCV quasi-species with increased pathogenicity [54]. At the opposite and given the immune-mediated pathogeny of HCV, the enhancement of immunity could be deleterious (hepatitis of immune restoration) as reported after withdrawal from chemotherapy in HCV-infected people [55]. The number of liver CD8 T lymphocytes directed against specific HCV antigens correlate with disease activity and it is known that HAART restores the numbers of both CD4 and CD8 cells [56,57]. Since the introduction of HAART, several cases of liver deterioration paralleling the immune restoration in HCV/HIV co-infected patients have been reported [58,59]. This phenomenon has been previously reported in cases of HBV infection, regardless of the cause of immune suppression [60], and has been suggested in cases of HCV infection [61]. Physicians should be aware and a sensitive liver follow-up in HCV/HIV co-infected patients taking HAART is needed and calls into question the respective temporal sequence of anti-HCV and anti-HIV therapies. On the contrary, it has been suggested that HAART and especially PIs may decrease the severity of the liver disease: chronic use of PIs and maintenance of high CD4 count could have a beneficial impact on liver fibrosis progression in HIV/HCV co-infected patients [17].
4. Predictors and co-factors of disease severity
4.1. Drug-related hepatotoxicity
Each component of HAART may result in drug-related hepatotoxicity [13,41]. All classes of antiretroviral drugs have been associated with liver enzyme abnormalities; the prevalence of these cases is not well defined and varies according to the studies and their methodology. Early diagnosis of drug-associated hepatic events is difficult since most HIV-infected patients do not fulfil the semiological and chronological signs that allow a clear diagnosis of drug-related hepatotoxicity [62-65], except for the typical but rare cases of idiosyncratic hepatitis [66]. Patients are treated with polypharmacy, they may use alcohol or drugs and HBV or HCV co-infection occurs in up to 30% of cases. Hepatotoxicity related to the use of NRTIs, especially zidovudine, stavudine and didanosine [67], correspond to exceptional cases of severe microsteatosis with lactic acidosis [68]. PI-related hepatitis is observed in 2-8.5% of PI-treated patients [69-71], with an increased risk in patients co-infected with HBV or HCV [71,72]. Liver toxicity, including fulminant hepatitis, has been associated with NNRTIs and pre-existing liver disease increases the relative risk [66,73,74].
In most cases of drug-related hepatitis, apart from fulminant presentations, liver enzyme abnormalities resolved after discontinuation of the drug and did not relapse, for example, after changing the PI [75]. We speculate that drug-related hepatitis may be involved in the deteriorating liver histology, including mechanisms associated with drug-related NASH [76].
4.2. Other causes of liver damage in HIV/HCV co-infected patients
Paralleling HAART-related hepatotoxicity and HCV-chronic hepatitis, HIV-infected patients may have other risk factors for liver enzyme elevations. These mainly include alcohol consumption, drug or other medications abuse and abnormal metabolic syndromes. The impact of these three last co-factors of liver deterioration has not been clearly analyzed in co-infected patients. The deleterious consequences of alcohol in HCV-mono-infected patients include: increase in levels of HCV RNA [77], which resolve with alcohol withdrawal [78] and this is usually achieved [79]; and increase in fibrosis progression [45,80]. In any case, in co-infected as well as in mono-infected patients, alcohol withdrawal appears necessary to limit its detrimental impact and also to improve the immune situation [81]. Chronic alcohol consumption is observed in 30% of HIV-infected patients (at least ex-drug abusers) [40,45] and may induce alcoholic hepatitis leading to liver deterioration [76]. Interestingly, alcohol, may per se enhance drug-related toxicity [82,83].
Drugs, especially cocaine and methamphetamine, have their own hepatic effects and these are sometimes severe (e.g. fulminant hepatitis with renal failure and rhabdomyolysis) [84,85].
Finally, abnormal metabolic syndromes may participate in liver deterioration of co-infected patients and include NASH in relation to abnormal metabolic syndromes (such as diabetes and dyslipidaemia) and to HAART or HAART-related metabolic syndromes [86,87]. The potential involvement of NASH in liver biochemical abnormalities and pathological deterioration is suggested by the significantly higher frequency of steatosis, Mallory bodies and neutrophils infiltrates in HIV-positive than negative patients [76]. This may be due to the effects of the antiretroviral therapy on the mitochondria [65,88,89], triggering lipid peroxidation and the release of pro-inflammatory cytokines, which favour fibrosis and apoptosis. As in immunocompetent patients, steatosis may be an independent factor associated with fibrosis progression, especially those infected with genotype 3 HCV or in overweight patients [90,91].
All of these co-factors can be partially controlled by effective and less toxic antiretroviral therapies and by medical counselling including advice on reducing alcohol consumption.
5. Morbidity and mortality
The decline of HIV-related mortality after the widespread use of HAART paralleled the emergence of HCV-related liver disease as an important cause of mortality among co-infected patients- estimates range from 10 to 45% [92-94]. In a nationwide cohort of HIV-infected patients followed-up from 1995 to 2003, mortality due to end-stage liver disease (ESLD) progressively increased over time and became a leading cause of mortality [4,13,95]. Among the 20,940 HIV-infected patients followed in 2003, 215 deaths occurred (mortality rate of 1%) [95]; causes of death were AIDS, ESLD and other in 46.9, 12.6 and 40.4%, respectively. Between 1995 and 2001, although overall mortality significantly decreased from 8 to 1%, the proportion of deaths attributable to cirrhosis and/or hepatocellular carcinoma (HCC) increased from 1.5 to 14.3% [4,13], while from 2001 to 2003 both overall mortality and death due to ESLD (12.6% in 2003) were stable [95]. Among the patients who died from ESLD, the percentage of patients with HCV infection was higher in 2003 (92.0%) than in 1995 (57.1%); HCC, as a cause of death, increased between 1995 and 2001 from 4.7 to 25% and decreased in 2003 (14.8%). In 2003, mortality attributable to ESLD seems to be steady which suggests a stabilization or a slowing down which may correspond either to a temporary trend or a sustained evolution. Finally, in 2003, ESLD-related death represented 24% of all non-AIDS-related deaths and constituted the most frequent non-AIDS-related cause of death in the HIV-positive population [95]. These data confirm results obtained from other French or European studies showing the primary and increasing importance of liver disease as a cause of mortality in the global HIV-population [96-100]. Several factors may explain the progressive increase of mortality due to ESLD observed since the introduction of HAART: prolonged longevity attributable to HAART, increased prevalence of alcohol use [13], or potential hepatotoxicity of HAART. In 2003, these may be counterbalanced by the favourable effect of hepatitis C treatment [101-104]. The role of HAART in liver disease progression and in overall mortality of HCV-HIV co-infected patients is still debated. HAART may be responsible for hepatotoxicity [13,41]. In contrast, other studies have suggested that the use of protease inhibitors may be protective with respect to the progression of HCV-related liver disease [17] and that HAART reduced long-term liver-related mortality, in addition to improved overall survival [18,105].
6. Conclusion
If the effect of HCV on HIV disease progression remains unclear, it is clear that HIV infection worsens the course of HCV infection. The development of HAART has resulted in a significant decrease in AIDS-related morbidity and mortality amongst HIV-infected patients allowing the expression of liver-related complications associated with the HCV chronic infection which is nowadays one of the leading causes of morbidity and mortality in co-infected patients. The prolonged longevity attributable to HAART, and the increased prevalence of alcohol use and NASH may explain this progressive increase of mortality due to end stage liver disease. The favourable effect of hepatitis C treatment (combination of pegylated interferon and ribavirin) in this population may slow the progression of morbidity and mortality due to ESLD in the near future (Fig. 1).
|
|
| |
| |
|
|
|
