barcelona . spain - European Association for the Study of the Liver

BARCELONA . SPAIN
132 POSTGRADUATE COURSE SYLLABUS ALCOHOLIC LIVER DISEASE 133
APRIL 18 - 19/2012 THE INTERNATIONAL LIVER CONGRESS TM 2012
(Figure 2) 7 . Moreover, alcohol is also recognized as a cofactor in the development of HCC. An additive
effect has been found between alcohol intake of ≥60 g/day and each hepatitis-virus infection. This effect
is more pronounced for HCV infection and has been confirmed in different studies. In patients with nonalcoholic
steatohepatitis (NASH), alcohol consumption has been recognized as the most significant factor
associated with the risk of developing HCC.
The proportion of HCCs attributable to alcohol consumption varies across the world, from 45% in Europe
and the USA, to 20% in Japan, and 0% in Africa and Asia 8 . These variations are possibly due to geographical
variations in alcohol consumption (Figure 1) but may be also due to under estimations of alcohol-related
HCC in countries where HCV or HBV incidence is still high. In fact, it is surprising to note that Japan and
the USA have the same per capita alcohol consumption (8 and 9.4, respectively) but only 20% of Japanese
HCCs are reported to be related to alcohol compared to 45% in the USA 8 . Moreover, South Korea and
France have the same per capita alcohol consumption (14.8 and 13.7, respectively), but alcohol-related
HCC is reported as 0% in South Korea compared to 70% in France. However, it is impossible to rule out
a genetic susceptibility to alcohol-related HCC, which could be higher in USA and French populations
compared to Asian. Patterns of alcohol consumption are also important because we know that binge
drinking is associated with an elevated incidence of sudden death.
GENETIC SUSCEPTIBILITY
As yet, only case-control studies or candidate-gene series have identified genetic-risk factors for HCC
in alcoholic patients. These studies have mainly identified polymorphisms in superoxide dismutase
(SOD), liver-iron overload, human hemochromatosis (HFE), glutathione peroxidase, myeloperoxidase,
methylenetetrahydrofolate reductase (MTHFR) and, more recently, patatin-like phospholipase A3 (PLPLA3).
Among these risk factors, only the first four seem to be specific to alcohol-related HCC; the others are also
associated with HCV- or HBV-related HCC. Despite the incontestable usefulness of these studies, we now
need global studies to improve our understanding of the associations between HCC and alcohol.
No genome-wide association study (GWAS) has been published yet that independently identifies the
genetic risk factors for HCC in alcoholic liver disease. Although GWAS data that evaluate the genetic risks
for HCC are available, they only provide information for chronic HBV or HCV infection.
MECHANISMS OF CARCINOGENESIS
In the past, alcohol was considered more as a co-carcinogen than a direct carcinogen, but recent studies
have clearly identified ethanol as a liver carcinogen by itself. General mechanisms of alcohol carcinogenesis
involve acetaldehyde formation, oxidative-stress induction, formation of liver cirrhosis, retinoid depletion,
and DNA-methylation deficiency (reviewed in Seitz et al. 9 and summarized in Figure 3). In the liver,
acetaldehyde is quickly metabolized and does not play a major role in HCC.
a.Oxidative stress
Oxidative stress is an important factor in hepatocarcinogenesis, whatever the etiology of liver disease.
Oxidative stress leads to the generation of reactive oxygen species (ROS). Ethanol-mediated ROS
formation may be caused by several factors:
• Chronic ethanol consumption causes a 10–20-fold increase in hepatic cytochrome
P450 2E1 (CYP2E1) in animals and humans, resulting in increased ROS production by
the CYP2E1-dependent microsomal mono-oxygenase system.
•TNFα production can increase ceramide levels, leading to increased mitochondrial ROS production.
•Hepatic iron overload (increased by chronic alcohol consumption) increases ROS.
ROS induces the generation of lipid-peroxidation products, such as malondialdehyde and 4-hydroxynonenal
(4-HNE), which can react with DNA bases to form exocyclic DNA adducts that are highly mutagenic.
b. Cirrhosis
Among the pre-neoplastic lesions found in the liver are enzyme-altered foci and dysplastic nodules. After
long-term administration of ethanol to rodents, these enzyme-altered foci are observed to contain quiescent
hepatic progenitor cells. In ALD, TNFα and TGFβ1 are markedly up-regulated. TNFα is an important
stimulator of oval-cell proliferation: it triggers JNK1 but also activates NFκB, which then activates cellsurvival
machinery. In addition, TGFβ1 is involved in the epithelial–mesenchymal transition (EMT), which
plays a role in carcinogenesis.
c.Retinoid depletion
Chronic alcohol consumption decreases vitamin A and retinoid-acid concentrations in the liver. In addition,
there is a strong inverse relationship between vitamin A concentration and later development of HCC.
The main reason for decreased retinoid acid is increased catabolism by ethanol-induced CYP2E1. This
decrease is associated with increased expression of the AP1 transcriptional complex (JUN and FOS),
resulting in hepatic hyperproliferation and decreased apoptosis, which may contribute to carcinogenesis.
The retinoid pathway has been shown to play an important role in HBV- and HCV-related HCC, and a direct
interaction between viral proteins has been shown. Moreover, it has been proposed that HBx may induce
promoter methylation of RAR-beta via up-regulation of DNA methyltransferases 1 and 3a.
d.DNA-methylation deficiency
Methylation of genes is an important tool in controlling gene expression: hypermethylation induces a
silencing effect and hypomethylation results in increased gene expression. Alcohol consumption interferes
with these methylation processes, mainly by inhibiting S-adenosyl-L-methionine (SAMe) synthesis, which
is the universal methyl-group donor and enzyme activator in methyl-transfer reactions.
It has recently been shown that, depending on the HCC risk factors (HBV, HCV, or alcohol related), the
DNA methylation profile differs 10 . Moreover, the methylation profile of a set of eight genes has been shown
to distinguish the clinical features of HCC, such as gender, age, HBV-positive tumors, pathological grade,
and alcohol intake when methylation of the MGMT gene significantly higher.
e. Genotoxicity
Few data on genome-wide HCC analyses show the influence of etiology on the underlying liver disease.
However, as mentioned above, it is possible that alcohol consumption is under-recorded in series that
have a high incidence of viral-related HCC. It has been reported that specific deletion of the 4q34-35
region is associated with heavy alcohol intake, even if the patient is infected by HCV. However conflicting
results have been published. Unsupervised global transcriptome analysis of HCC did not identify significant
differences between the underlying liver pathology. However, direct comparison of transcriptome analysis
between HCV and alcohol-related HCC has revealed some differences 11 .
f. Other mechanisms
Alcohol-induced CYP2E1 may interfere with the metabolism of certain pro-carcinogens, such as
nitrosamines, polycyclic hydrocarbons, and aflatoxins. This interaction increases the carcinogenic effect of
these pro-carcinogens (except for nitrosamines simultaneously administered with ethanol).
PROGNOSIS AND TREATMENT OF ALCOHOL-RELATED HCC
None of the frequently used HCC classification systems take into account the etiology of liver disease. In
fact, no significant differences have been found in liver function between ALD and viral hepatitis 12 . Moreover,
even though the etiology of liver disease has been defined as a prognostic factor in some studies, the
power of this finding is small.
A patient’s prognosis depends on access to treatment. Compared to HBV-, HCV-, or NASH-related cirrhosis,
ALD is considered to be self-inflected, and the general public and even some professionals question the
degree of priority given to these patients. In a study on adherence to AASLD criteria for referral for liver
transplantation, it has been shown that patients with ALD have a significantly poorer odds ratio of receiving