REVIEWS

Alcoholic liver disease: mechanisms of injury

and targeted treatment
Alexandre Louvet and Philippe Mathurin
Abstract | Alcoholic liver disease (ALD) is a complex process that includes a wide spectrum of hepatic
lesions, from steatosis to cirrhosis. Cell injury, inflammation, oxidative stress, regeneration and bacterial
translocation are key drivers of alcohol-induced liver injury. Alcoholic hepatitis is the most severe form of all
the alcohol-induced liver lesions. Animal models of ALD mainly involve mild liver damage (that is, steatosisand
moderate inflammation), whereas severe alcoholic hepatitis in humans occurs in the setting of cirrhosis and is
associated with severe liver failure. For this reason, translational studies using humans and human samples
are crucial for the development of new therapeutic strategies. Although multiple attempts have been made to
improve patient outcome, the treatment of alcoholic hepatitis is still based on abstinence from alcohol and
brief exposure to corticosteroids. However, nearly 40% of patients with the most severe forms ofalcoholic
hepatitis will not benefit from treatment. We suggest that future clinical trials need to focus on end points
other than mortality. This Review discusses the main pathways associated with the progression of liver
disease, as well as potential therapeutic strategies targeting these pathways.
Louvet, A. & Mathurin, P. Nat. Rev. Gastroenterol. Hepatol. 12, 231242 (2015); published online 17 March 2015;
doi:10.1038/nrgastro.2015.35

Introduction
Excess alcohol consumption is the primary cause of liver- survival (>5years) is better in patients with cirrhosis who
related mortality in Western countries. The consequences remain abstinent than in those who begin drinking again.5,6
of alcohol consumption on health vary according to the However, unlike outcomes after treatment for other liver
drinking pattern (excessive or not, acute or chronic); envi- diseases such as chronic HBV, HCV or autoimmune hepa-
ronmental and individual factors are also key drivers.1 titis, no evidence exists for the reversal of cirrhosis in alco-
Alcohol-induced liver injury includes fatty liver, fibrosis holic liver disease (ALD). Early lesions such as steatosis
and alcoholic hepatitis; these basic lesions can occur sepa- might disappear after alcohol withdrawal in heavy drinkers
rately, simultaneously or sequentially in the same patient. who do not have other risk factors for steatosis (such as
Alcoholic hepatitis is a necroinflammatory process that is insulin resistance and/or obesity),7 whereas the inflam-
associated with the fastest progression of fibrosis and leads matory lesions of alcoholic hepatitis can persist for several
to cirrhosis in 40% of cases.24 This pivotal lesion mark- months after alcohol withdrawal.8,9 It is important to bear
edly increases the risk of liver decompensation, whereas in mind that ALD is often diagnosed at very advanced
patients without alcoholic hepatitis are at lower risk of stages of the disease, unlike other types of liver disease
developing cirrhosis.4 These discrepancies in the natural such as HCV or NAFLD. This fact emphasizes the need
progression of the disease explain why a subgroup of heavy for systematic screening of fibrosis in heavy drinkers, using
drinkers with inflammatory features are candidates for noninvasive methods such as transient elastography or
treatment to reduce short-term and intermediate-term blood tests.1 A large study to assess the effect of this type of
liver injury, whereas patients without these lesions are screening programme is needed. This Review discusses the
candidates for treatment to reduce long-term morbidity relevance of certain pathways according to the type of liver
and mortality. Thus, in patients with inflammatory fea- lesion, and of potential molecules targeting these pathways
tures, new drugs should target specific pathways to restore in relation to the pattern and the severity of liver injury.
liver function, whilst patients with milder liver disease
(for example, isolated steatosis) could benefit from other Ethanol metabolism and cell injury
Service des Maladies therapeutic strategies to maintain abstinence and limit the Ethanol is a small molecule that diffuses easily across cell
de Lappareil Digestif
etUnit INSERM 995,
progression of fibrosis. membranes. Many factors influence ethanol absorption
Hpital Huriez, Generally speaking, whatever the severity of disease, and metabolism, including gender, age, ethnicity and body
RuePolonovski, recovery from alcohol-induced liver injury is dependent weight, and <10% of ethanol is directly eliminated by the
59037LilleCedex,
France (A.L., P.M.). upon cessation of alcohol consumption. Indeed, long-term lungs, kidneys and sweat in its unchanged form.10 The
main metabolic pathway involved in the biotransforma-
Correspondence to:P.M.
philippe.mathurin@ Competing interests tion of ethanol is oxidation into acetaldehyde (Figure1).
chru-lille.fr The authors declare no competing interests. This process uses NAD+ and is primarily achieved by

NATURE REVIEWS | GASTROENTEROLOGY & HEPATOLOGY VOLUME 12 | APRIL 2015 | 231

2015 Macmillan Publishers Limited. All rights reserved
REVIEWS

Key points system, based on cytochromes P450 and in particular

CYP2E1.12,13 Under normal conditions, CYP2E1 accounts
Alcoholic liver disease (ALD) is characterized by a complex spectrum of
for 10% of ethanol oxidation,14 but this system is induc-
histological lesions, ranging from steatosis to cirrhosis
Alcoholic hepatitis is a specific entity, which is associated with a fast
ible by its substrates, in particular ethanol itself. CYP2E1
progression to cirrhosis and with liver failure and a poor outcome in its most is mainly expressed in the perivenous zone (Rappaport
severe form zone3), which might partially explain why alcohol-related
ALD is characterized by oxidative stress, disturbance of hepatocyte metabolism, liver injury is more frequent in this part of the hepatic
liver inflammation, modifications in the regeneration process and translocation lobule.11,15 Regardless of the involved pathway, after oxi-
of bacterial products from the gut microbiota into the portal blood stream dation most acetaldehyde is converted into acetate by
Because animal models do not reproduce the complete spectrum of alcohol- aldehyde dehydrogenase, which is expressed in the cyto-
induced hepatic lesions, translational research based on human samples is
plasm and in mitochondria. This reaction is catalyzed by
crucial to identify new treatment options
The best-documented treatment for severe alcoholic hepatitis is
NAD+/NADH and increases the amounts of NADH in the
corticosteroids, but this treatment is not ideal and 40% of patients still die liver.13 As alcohol dehydrogenase, CYP2E1 and aldehyde
after a short period of time dehydrogenase are mainly expressed in hepatocytes, most
The future evaluation of new strategies in severe alcoholic hepatitis should of the direct cellular toxicity of ethanol affects these cells.
focus not only upon survival, but also on surrogate markers of outcome, with a Ethanol metabolism leads to accumulation of reac-
specific plan of development tive oxygen species (ROS), mainly hydrogen peroxide
(H2O2) and superoxide anion O2,16 which is exacerbated
by hypoxia, bacterial translocation and the release of pro-
inflammatory cytokines. Given their short half-life and
LPS TLR4 Kupffer cell
ROS high reactivity, these radicals quickly bind to ethanol or
Ethanol NADPH iron atoms to form reactive metabolites such as hydroxyl
oxidase radical (OH), ferrous oxide (FeO) or hydroxyethyl radical
(CH3CHOH), which are all responsible for lipid peroxi-
Endothelium TNF dation of cell membranes. Mitochondria (through their
Ethanol ROS respiratory chain), the endoplasmic reticulum (through
Hepatocyte CYP2E1) and Kupffer cells (through NADPH oxidase)17
TNFR1
are the main sources of ROS. During alcohol-induced
Ethanol HO-1 Lipid liver injury, iron is also involved in oxidative stress and
NADH
peroxidation promotes fibrosis by catalysing the formation of ROS.18,19
ADH MEOSCYP2EI The cytotoxic effects of ethanol metabolism and ROS
UPR
lead to cell death; evidence exists of apoptosis andnecrosis
ER
NAD+ Acetaldehyde stress in ALD.20 DAMPs (damage-associated molecular patterns)
UPR are released after cell death, mainly necrosis, and trigger
ALDH
Lipid ER macrophage and neutrophil activation, fibrogenesis and
peroxidation stress hepatic regeneration.20 Senescence (via natural killer
NADH Acetate IRF3
GSH [NK] cells) and autophagy are regulators of liver inflam-
mation after cell death and the release of DAMPs.21,22
Oxidative stress Although it is tempting to hypothesize that these mol-
ROS ROS
Steatosis ecules (for example, HMGB1 and miR122) have a key
role in alcohol-induced cell death, few data are available in
ALD compared with other liver diseases such as NAFLD
Inflammation
Apoptosis or paracetamol-induced liver failure.
Chronic exposure to ethanol induces glutathione deple-
tion, which makes hepatocytes more sensitive to oxidative
Figure 1 | Metabolism of ethanol and related
Nature cell injury.
Reviews Ethanol is mainly&metabolized
| Gastroenterology Hepatology stress,2326 as reduced (not oxidized) glutathione protects
by alcohol dehydrogenase and MEOS into acetaldehyde, which is responsible for cells against ROS.27,28 Heme oxygenase1 (HO1) is also a
thegeneration of ROS. ROS cause oxidative stress, ER stress andsteatosis. ROS
protective factor against oxidative stress in ALD because
arealso generated through activation of NADPH oxidase in Kupffer cells. All
thesechanges in hepatocyte metabolism lead to inflammation and apoptosis.
it blocks CYP2E1 activity, reducing generation of ROS.29
Abbreviations: ADH, alcohol dehydrogenase; ALDH, aldehyde dehydrogenase; The endoplasmic reticulum is a central organelle for the
CYP2E1, cytochrome P450 2E1; ER, endoplasmic reticulum; GSH,reduced maturation of proteins, in particular their folding process.
glutathione; HO1, heme oxygenase 1; IRF3, interferon regulatory factor 3; LPS, In response to various stimuli, incomplete maturation
lipopolysaccharide; MEOS, microsomal ethanol oxidation system; NAD, nicotinamide produces an accumulation of unfolded proteins in the
adenine dinucleotide; ROS, reactive oxygen species; TNFR1, tumour necrosis factor lumen of the endoplasmic reticulum that activate several
receptor 1; TLR4, Toll-like Receptor 4; UPR, unfolded protein response. pathways leading to endoplasmic reticulum stress.30 In
ALD, enhanced endoplasmic reticulum stress, due in
alcohol dehydrogenase, which is mainly expressed in the part to ROS and hyperhomocysteinemia,31 participates
liver and gastrointestinal tract.11 Two other systems can in inflammation through the activation of NFB and
lead to ethanol oxidation: catalase, levels of which are JNK, and in apoptosis, in particular through disruption
negligible in the liver,11 and microsomal ethanol oxidation of cellular calcium homeostasis and the activation of the

232 | APRIL 2015 | VOLUME 12  www.nature.com/nrgastro

2015 Macmillan Publishers Limited. All rights reserved
 REVIEWS

TNF

TNFR1
Ethanol
NADH HO-1 Homocysteine
UPR
ADH ER
MEOSCYP2EI Betaine stress
UPR
NAD+ ER
Acetaldehyde stress
Protective genes
against steatosis
GSH
SREBP1c
AMPK
ROS
PPAR- CPT1 ACC FAS SCD1

Fatty acid oxidation and export Lipogenesis

Steatosis

Hepatocyte

Adiponectin (released in the blood vessels)

Adipose tissue Ethanol

Proinflammatory cytokines
TNF, IL-6, CCL2

Macrophages expressing
TNF, IL-6, CCL2
Nature Reviews | Gastroenterology & Hepatology
Figure 2 | Mechanisms of alcohol-related steatosis. Chronic alcohol consumption leads to steatosis via generation of
acetaldehyde, ROS and ER stress. The consequences are blockade of PPAR and of AMPK, which is responsible for fattyacid
oxidation and export via ACC and CPT1. In addition, chronic alcohol consumption induces SREBP1c activation, which is
responsible for fatty acid synthesis through FAS and fatty acid desaturation through SCD1. TNF also leads tosteatosis by
activating SREBP1c. Betaine reduces homocysteine levels, which thus enhances ER stress. Chronic alcohol consumption
also induces adipose tissue inflammation, which decreases the release of the protective adiponectin, thus favouring
steatosis. All these mechanisms lead to disruption ofhepatic lipid metabolism by increasing lipogenesis and decreasing
fatty acid oxidation and export. Abbreviations: ACC, acetyl CoA carboxylase; AMPK, adenosine monophosphate-activated
protein kinase; CCL2, CC-chemokine ligand2; CPT1, carnitine palmitoylrransferase 1; ER, endoplasmic reticulum; FAS,fatty
acid synthetase; PPAR, peroxisome proliferator-activated receptor ; ROS, reactive oxygen species; SCD1, stearoyl-CoA
desaturase 1; SREBP1c, sterol regulatory element-binding protein 1c; TNFR1, tumour necrosis factor receptor 1.

CHOPGADD153 pathway.31,32 IRF3 links endoplasmic Mechanisms of alcohol-induced steatosis

reticulum stress to mitochondria and subsequent hepato- Chronic alcohol consumption promotes steatosis by
cyte apoptosis by activation of caspases, but also to liver disrupting hepatic lipid metabolism via SREBPs and
inflammation in hepatocytes and liver mononuclear cells PPAR (peroxisome proliferator-activating receptor),
by stimulating the interferon and the NFB pathways.33 which are directly influenced by AMPK (5' adenosine
Endoplasmic reticulum stress, amplified by elevated monophosphate-activated protein kinase) (Figure2).
homocysteine levels, contributes to alcohol tissue damage SREBP1c exerts its deleterious role by increasing fatty
and is counterbalanced by transcription genes involved acid biosynthesis through fatty acid synthase and enzymes
in the unfolded protein response, such as GRP78, GRP94 responsible for fatty acid desaturation such as stearoyl-
or Herp.34,35 Another consequence of endoplasmic reticu- CoA desaturase.39 SREBP expression is increased by
lum stress is the activation of the steatogenic pathways acetaldehyde and TNF,40 which also stimulate its matura-
in hepatocytes, in particular via SREBPs (sterol regula- tion in hepatocytes.41 SREBP might be downregulated by
tory element-binding proteins).36,37 Ethanol also gener- betaine (also known as glycine betaine) administration.37
ates endoplasmic reticulum stress in stellate cells through Conversely, PPAR prevents ethanol-induced steatosis.42
inhibition of a protective autophagy phenomenon, leading Its expression is reduced in hepatocytes by acetaldehyde43
to their activation in fibrogenic cells.38 and during chronic alcohol intake by downregulation of

NATURE REVIEWS | GASTROENTEROLOGY & HEPATOLOGY VOLUME 12 | APRIL 2015 | 233

2015 Macmillan Publishers Limited. All rights reserved
REVIEWS

RXR, which prevents its heterodimerization to PPAR After binding to TLR4 via LBP (lipopolysaccharide
and subsequently alters PPAR signalling.44 binding protein), LPS activates Kupffer cells via a signalling
As in metabolic steatosis, AMPK downregulation is cascade that includes CD14, MyD88, MD2 (also known as
central to alcohol-induced lipid accumulation by stimu- lymphocyte antigen96) and finally results in the activation
lating SREBP1c and inhibiting PPAR expression.2,45,46 of mitogen-activated protein kinases (such as ERK1, ERK2,
AMPK expression is decreased by ethanol invitro and JNK and p38), NFB76 and AP1.77 This activation might
in experimental models of alcoholic steatosis.47,48 AMPK also be MyD88-independent.78 When these pathways are
downregulation also causes a decrease in fatty acid oxi- activated, Kupffer cells release ROS, adhesion molecules
dation mediated by acetyl-coA carboxylase and carnitin (such as ICAM1 [intracellular adhesion molecule1] and
palmitoyltransferase1. Lipid metabolism disturbances VCAM1 [vascular cell adhesion protein1]), chemokines
are also worsened by endoplasmic reticulum stress (IL8 and CCL2) and proinflammatory cytokines (TNF,
and ROS. The metabolic pathways sirtuin1,49,50 can- IL1 and IL6).2,79 Targeting NADPH oxidase in Kupffer
nabinoid receptors,51,52 complement fractions5355 and cells might be an interesting therapeutic target because of
PPAR56 might also participate in alcohol-induced stea- its role in the defence against bacteria, in inflammatory
tosis. All of these mechanisms increase lipogenesis and responses, generation of ROS and in the control of hepatic
decrease fatty acid oxidation and export during chronic fibrosis.16,80,81 Finally, the intracellular labile iron fraction
alcoholconsumption. regulates activation of NFB in Kupffer cells82 and the
Insulin resistance has a key role in alcohol-induced cascade of inflammation pathways associated with the early
steatosis, as shown by the faster disease progression in stages of alcohol-induced liver injury.8385
patients who are overweight and drink to excess compared Leukocyte recruitment amplifies the inflammatory
with normal weight individuals who drink to excess.5759 response to LPS. In humans, alcoholic hepatitis is char-
However, data also suggest that modest alcohol intake acterized by the presence of large amounts of cytokines
might result in a decrease in inflammatory lesions in that are highly sensitive to PAMPs, such as RANTES
patients with NAFLD.60 In animal models, ethanol feeding (also known as CC chemokine receptor type1), IL8,
induces insulin resistance, macrophage infiltration and the Gro and CCL2,8689 leading to neutrophil recruitment
production of proinflammatory cytokines such as TNF, in the liver. Besides their role in neutrophil infiltration,
IL6 and CC-chemokine ligand2 (CCL2, also known as proinflammatory changes in Kupffer cells also seem to
MCP1) in adipose tissue.61 Adiponectin is downregulated enhance steatosis.90,91 The role of Kupffer cells is not
in those models,61 leading to a decrease in fatty acid oxi- just deleterious in ALDthese cells can also adopt an
dation, an increase in fatty acid synthesis and inhibition antiinflammatory phenotype92 and release IL10.93
of AMPK and PPAR.6264 Decreased production of adi- Although innate immunity has a central role in the
ponectin has also been observed in adipose tissue due to pathogenesis of ALD, adaptive immunity might also con-
endoplasmic reticulum stress through an accumulation of tribute to the progression of alcohol-induced liver injury.
homocysteine inadipocytes.65 Indeed, chronic alcohol consumption leads to the devel-
opment of antibodies directed against lipid peroxidation
Immunity, dysbiosis and ALD products, which activate an adaptive immune response.2
The role of bacterial translocation in the pathogenesis of More specifically, alcohol consumption stimulates splenic
ALD has been clearly established. Chronic alcohol expo- Tcells and natural killerT(NKT) cells in the liver, leading
sure in humans and animal models increases circulating to the development of a cytotoxic response against hepato-
concentrations of lipopolysaccharide (LPS) compared cytes, which might be related, at least in part, to Fas and
with controls,6668 and the severity of hepatic injury is TNF receptor1 signalling.94,95 Other cells, such as stellate
correlated to serum levels of LPS.69,70 Exogenous admin- cells, might also participate in the release of inflammatory
istration of LPS induces necroinflammatory lesions in the mediators, such as CCL2, IL8, TNF and IL1.9698
liver,66,71 whereas administration of antibiotics or probiot- When considering the effects of bacterial translocation
ics reduces the severity of hepatic lesions in ethanol-fed on the pathogenesis of ALD, the question of modifications
rats.72,73 After translocation from the gut lumen to the in gut microbiota (that is, dysbiosis) in patients with excess
liver, LPS and other pathogen-associated molecular pat- alcohol consumption has become an important issue.
terns (PAMPs) are recognized by pathogen-recognition Bacterial overgrowth99 and dysbiosis have been observed
receptors (PRRs), including TLRs (Toll-like receptors, in animal models100 with subsequent intestinal oxidative
mainly TLR4, which is responsible for LPS recognition) stress, leaky gut and intestinal inflammation.101,102 In dys-
and NOD1 and NOD2 (nucleotide oligomerization biosis, ethanol targets gut microbes by decreasing their
domain) receptors. As well as LPSTLR4, liver inflam- ability to synthesize saturated fatty acids; by contrast, sup-
mation is also mediated by other bacterial compounds, plementation of saturated long-chain fatty acids restores
which explains why the injection of other bacterial eubiosis, reduces gut leakiness and ALD.103 Compared
motifs also leads to recruitment of inflammatory cells.74 with nondrinkers, some patients who drink to excess
More specifically, the activation of TLR2 and TLR6 have been shown to have colonic dysbiosis, including a
(which are both responsible for recognition of bacte- lower percentage of Bacteroidetes and a higher percent-
rial lipopeptides), and TLR9 (which recognizes bacterial age of Proteobacteria, which are associated with higher
DNA-containing unmethylated CpG motifs), leads to an levels of endotoxins. This finding suggests that PAMP-
increase in the proinflammatory cascade in ALD.74,75 induced liver inflammation might be enhanced in these

234 | APRIL 2015 | VOLUME 12  www.nature.com/nrgastro

2015 Macmillan Publishers Limited. All rights reserved
 REVIEWS

individuals.104 Quantitative metagenomics has also shown liver injury.124 An ideal animal model of alcoholic hepatitis
that patients with cirrhosis can be identified according to should at least include inflammation, steatosis, cell damage,
clusters of cognate bacterial species compared with healthy fibrosis and the activation of regeneration. The first model
individuals,105 suggesting that gut microorganisms have a of chronic alcohol exposure was developed in rats125 and
role in the severity of liver disease. Thus, probiotics might induced the development of isolated steatosis. This model
be able to help limit the progression of ALD. Indeed, pre- was modified so that larger quantities of ethanol could
liminary studies have shown that the administration of be administered for 412weeks (the so-called Lieber-De
probiotics (for example, Lactobacillus sp.) restored gut Carli model126). Later, a model based on chronic intragas-
microbiota, reduced hepatic inflammation in animals,106 tric infusion of ethanol was developed that resulted in mild
restored the phagocytic capacities of neutrophils in inflammation and fibrosis.127,128 To mimic the inflamma-
patients with alcoholic cirrhosis107 and improved liver tion observed in humans, some researchers have suggested
enzyme levels in heavy drinkers.108 Prebiotics have also injecting bacterial compounds in chronically fed rodents at
been shown to be protective in NAFLD109 and can be the end of the ethanol diet.74,119,129 For the moment, animal
viewed as a promising strategy in ALD, although no data models with severe inflammatory lesions exist (although
support their use at present.110 these models must be confirmed by others; for example,
the chronic plus binge drinking model130 and the modified
Regeneration and ALD TsukamotoFrench model131), but there is still no model
Liver regeneration is a key target for the development of with cirrhosis and bilirubinostasis.
new treatment molecules in ALD, in particular alcoholic As data from rodents can only barely be extrapolated
hepatitis. Regeneration is necessary to restore functional to humans, data from human samples are the best way to
liver tissue after insults such as partial hepatectomy or toxic further the understanding of the pathogenesis and treat-
injury. Hepatocytes proliferate in the absence of chronic ment of ALD. For example, transcriptome analysis based
liver disease, but in ALD this mechanism is not as effective on human samples has identified the TNF superfamily
when liver progenitor cells might act as facultative cells for receptors as a potential area of development of therapeutic
regeneration (discussed later).111,112 Models of partial hepa- strategies, because their expression is upregulated in the
tectomy have identified TNF, IL6, STAT3 and NFB as liver of patients with alcoholic hepatitis compared with
the key drivers of liver regeneration.113117 In alcohol-fed healthy controls or patients with other liver diseases.132
mice, steatosis is associated with increased production of On the basis of these examples, outcomes can now be
ROS and inhibition of mature hepatocyte proliferation.118 linked to the histological133 and/or molecular signatures
Chronic ethanol exposure also impairs liver regeneration of patients with alcoholic hepatitis.87,122,134136 Another
by limiting DNA synthesis by mature hepatocytes119 and approach that has been taken with other liver diseases
impairing normal miRNA signalling during the regen- that is, confirming data from human samples in animal
eration process.120 At a cellular level, chronic alcohol con- modelscould help identify new therapeutic pathways.137
sumption induces senescent replication of hepatocytes and Because translational research based on human samples
proliferation of liver progenitors.111 An important missing has such a key role in the understanding of mechanisms
piece of the research puzzle is the lack of an experimental of alcoholic hepatitis, the collection of biospecimens in
model with altered liver regeneration. future clinical trials could help substantially in the design
Although data are lacking on the effect of the accumula- of new therapeutic strategies.
tion of liver progenitor cells in alcohol-injured livers, an
incomplete differentiation process leading to repopula- Targeted treatment of ALD
tion of the liver parenchyma by ductular cells instead of Corticosteroids
mature hepatocytes could explain why some patients with Corticosteroids were proposed in the AASLD (American
severe forms of alcoholic hepatitis fail to recover.121,122 Association for the Study of Liver Diseases)138 and EASL
Our group has shown that patients with severe alcoholic (European Association for the Study of the Liver)1 guide-
hepatitis who do not respond to medical treatment have lines to treat severe forms of alcoholic hepatitis (defined
low hepatic expression of TNF and IL6 and an aberrant by a Maddreys discriminant function 32) on the basis of
regeneration process in which hepatic progenitor cells meta-analyses showing a reduction in short-term mortal-
cannot differentiate into mature hepatocytes.123 Thus, new ity.139,140 Nevertheless, despite the positive effects of this
therapeutic strategies should target hepatocyte survival treatment on short-term survival, mortality 6months after
and replication or liver progenitor cell differentiation into the onset of disease is still ~3040%.1,3
mature hepatocytes. Because some patients do not respond to existing
corticosteroids, it is crucial to identify nonresponders
Disease models of ALD rapidly. Scoring systems for alcoholic hepatitis include
A major unmet need in the study of ALD is the lack of the Glasgow Alcoholic Hepatitis Score,141 the ABIC (age
a reliable animal model that mimics the entire spectrum bilirubinINRcreatinine) score142 and the Lille model.143
of this disease in humans. Indeed, owing to marked dif- The formula of the Lille model144 includes the effect of
ferences in alcohol metabolism between rodents and corticosteroids on bilirubin levels after 7days of treatment,
humans, most animal models with simple ethanol feeding which is highly predictive of outcome.145 The difference
lead to steatosis, mild liver inflammation and no fibrosis, in bilirubin level is combined with age, the presence of
with major differences between mice strains in terms of renal insufficiency, albumin level, prothrombin time and

NATURE REVIEWS | GASTROENTEROLOGY & HEPATOLOGY VOLUME 12 | APRIL 2015 | 235

2015 Macmillan Publishers Limited. All rights reserved
REVIEWS

Table 1 | Main ongoing studies of targeted treatment for alcoholic hepatitis and alcoholic cirrhosis
Treatment Type of molecule Mechanism of action Identifier*
Emricasan (IDN 6556) Pan-caspase inhibitor Reduces apoptosis NCT01912404187
Metadoxin Hepatoprotective agent Antioxidant NCT02161653188
Metadoxin and garlic acid Hepatoprotective agent Antioxidant NCT02019056189
Anakinra (+pentoxifylline+zinc) IL1 receptor antagonist Decreases hepatic inflammation NCT01809132190
Amoxicillin clavulanate Antibiotic Decreases bacterial overgrowth and bacterial NCT02281929191
translocation; prevents development of infection
Rifaximin Antibiotic Decreases bacterial overgrowth and bacterial NCT02116556192
translocation; prevents development of infection
Ciprofloxacin Antibiotic Decreases bacterial overgrowth and bacterial NCT02326103193
translocation; prevents development of infection
Lactobacillus rhamnosus Probiotic Decreases bacterial overgrowth; regulates gut NCT01922895194
inflammation
G-CSF Growth factor Promotes liver regeneration NCT01820208195
Obeticholic acid Biliary acid Improves cholestasis NCT02039219196
Mycophenolate mofetil and Immunosuppressor Decreases hepatic inflammation NCT01903798197
rilonacept andIL1 blocker
Zinc Oligoelement Decreases gut inflammation NCT02072746198
Stem cells Progenitors Promote liver regeneration NCT01875081199
*All trials are listed at www.clinicaltrials.gov. Four consortia have been funded by the NIH to test targeted therapies; these consortia are involved in some of
these clinical trials. More information about these consortia can be found at RePORT. 200 Abbreviation: G-CSF, granulocyte colony-stimulating factor.

bilirubin level at the beginning of treatment to generate treatment limits the ability of neutrophils and monocytes
a score that ranges from 0 to 1. The ideal cut-off of 0.45 to produce TNF in the presence of LPS.
can be used to define responders to corticosteroids (Lille Patients with alcoholic hepatitis have decreased sen-
score <0.45) and nonresponders (Lille score 0.45). The sitivity to corticosteroidsas shown by defective inhi-
therapeutic response can be refined further by defin- bition of lymphocyte proliferationand so another
ing three populations: complete responders (Lille score strategy to improve patient outcome could be to increase
0.16); partial responders (Lille score 0.160.56); and the effect of this treatment in this group.150 Interestingly,
null responders (Lille score 0.56). Corticosteroids can patients who respond to corticosteroids were shown
be withdrawn after 7days in null responders, because to have better exvivo sensitivity to this drug than non
the outcome of these patients is not different from that of responders.151 The results of this trial were confirmed in
patients treated with placebo.140 an independent cohort 152 and a suggested threshold of
In theory, corticosteroids help treat severe alcoholic 60% maximum intensity (of lymphocyte proliferation)
hepatitis because of their anti-inflammatory proper- was proposed to define resistance to corticosteroids.
ties; however, the underlying pathways of this treatment Sensitivity to corticosteroids is improved as liver dys-
have still not been clarified. In animal models, alcohol- function recovers and might be restored by theophyl-
induced liver injury was prevented by the administration line (a xanthine), potentially via enhanced recruitment
of antibodies against TNF in wild-type mice and was not of histone deacetylases, which silence the expression of
observed in mice lacking TNFreceptor1.146,147 Because proinflammatory genes. Basiliximab (a monoclonal anti-
of this evidence from animal models, the first studies body triggering the IL2 receptor [anti-CD25]) might be
focused on TNF pathways. Levels of TNF,89 its p55 soluble another therapeutic option to compensate for defective
receptors and IL8 (a strong chemoattractant responsi- sensitivity to corticosteroids.151 Another option to target
ble for neutrophil recruitment) are elevated in patients signalling pathways involved in histone acetylation might
withsevere alcoholic hepatitis148 compared with those with be decreasing acetate levels via inhibition of acetyl-coA
alcoholic cirrhosis or healthy livers. On the other hand, synthetase, leading to downregulation of proinflamma-
IL10 levels are unaffected by the presence of alcoholic tory genes (IL6, IL8 or TNF) in macrophages exposed to
hepatitis, suggesting an imbalance between proinflamma- ethanol.153 Downregulation of proinflammatory genes in
tory and anti-inflammatory cytokines. Although prelimi- macrophages and monocytes can also be obtained by stim-
nary studies showed a correlation between baseline serum ulation of GILZ (glucocorticoid-induced leucine zipper),
TNF levels and poor outcome, this relationship has not whose expression is stimulated by corticosteroids.154 New
been confirmed in further studies.149 After corticosteroid molecules that increase sensitivity to corticosteroids could
treatment, IL8 and TNFsRp55 (soluble TNF recep- be potential targets to improve patient management.
tor p55) plasma levels decrease, particularlyTNFsRp55
in survivors, whereas IL10 levels increase. Conversely, Anti-TNF antibodies
serum TNF levels are unchanged after corticosteroid The results of pilot studies and the first randomized
therapy, although exvivo experiments have shown that this controlled trial155157 evaluating anti-TNF antibodies

236 | APRIL 2015 | VOLUME 12  www.nature.com/nrgastro

2015 Macmillan Publishers Limited. All rights reserved
 REVIEWS

Liver
effective163 or more effective than corticosteroids in such
Antioxidants Growth factors
patients.164 Experimental data have demonstrated the anti-
Oxidative stress Regeneration
Stem cells inflammatoryeffects of pentoxifylline,165,166 although its
benefit in severe alcoholic hepatitis seems to be related
Anti-caspase Antifibrotic
Steatosis drugs to the prevention of hepatorenal syndrome.162 In the sub-
Stellate
group of patients not responding to corticosteroids, pent
cell oxifylline was an ineffective rescue strategy.167 Although
Cell
death pentoxifylline and corticosteroids might have a synergis-
tic effect as suggested by the different pattern of effects,
a large randomized controlled trial did not show any
Kupffer benefit in survival with the combined administration of
cell
PMN
Anti-inflammatory pentoxifylline and prednisolone, compared to predniso-
TLR molecules
lone alone;168 a lower incidence of hepatorenal syndrome
TLR
antagonists
was noted in the combined treatment group, but did not
result in improved survival compared to the prednisolone
PAMPs (LPS)
alone group. A head-to-head randomized study showed
Bacterial Portal vein that survival was not as good with pentoxifylline as with
translocation
prednisolone.169 In the large randomized controlled trial
STOPAH, which included >1,000 patients, results with
Intestinal pentoxifylline were not better than placebo for short-term
Antibiotics inflammation
Gut mortality. On the other hand, corticosteroids improved
Probiotics
Prebiotics 28-day survival.170 One important limitation of this study
Zinc
was the absence of biopsy-proven alcoholic hepatitis and
Figure 3 | Potential targeted therapies in ALD. There are many potential targets of previously validated noninvasive criteria for alcoholic
Nature liver
totreat ALD, including the gut microbiota, Reviews | Gastroenterology
inflammation, & Hepatology
regeneration, cell hepatitis. As a secondary end point, 1year survival was
death and oxidative stress. Several of them are being studied (see Table1). only affected by alcohol relapse and not by treatment allo-
Abbreviations: ALD, alcoholic liver disease; LPS, lipopolysaccharide; PAMPs, cation.170 In another study, pentoxifylline was not better
pathogen-associated molecular patterns; PMN, polymorphonuclear neutrophil; than placebo in patients with decompensated cirrhosis,
TLR,Toll-like receptor. although it seemed to decrease the risk of liver decom-
pensation.171 In summary, little evidence exists to support
in patients with alcoholic hepatitis were encouraging. using pentoxifylline in ALD.
Administration of TNF antagonists was associated with
improved portal haemodynamics and liver function,155,158 Other treatments
reduction of plasma IL8 and IL6 levels157 and 1month Oxidative stress is one of the key mechanisms that drives
survival in >80% of patients. The first randomized study early ALD, in particular steatosis. As the precursors of
included 20 patients with severe alcoholic hepatitis, and glutathione have been shown to improve steatosis in
demonstrated a greater decrease in Maddrey function rodents,2,3,172 it is tempting to evaluate these molecules
in the infliximab plus prednisolone group compared to in humans. Sadenosylmethionine could be an interest-
the placebo plus prednisolone group.156 However, the ing therapeutic option for ALD because this molecule
results of two large randomized controlled trials testing restores mitochondria levels of glutathione, whereas
infliximab159 or etanercept (soluble TNF receptor)160 were Nacetylcysteine does not.173,174 Betaine supplementation
disappointing, and reported higher rates of infection and reduces steatosis in ethanol-fed mice by decreasing endo-
mortality in the patients treated with anti-TNF agents. plasmic reticulum stress by conversion of homocysteine
Major discrepancies exist in the administration proto- to methionine.37 However, no data yet support the use of
col between the preliminary studies and the two large these two molecules in humans in the early stages of ALD.
studies; the latter used sequential infusions whereas the Antioxidant molecules alone are not effective in severe
first preliminary study used a single infusion of anti-TNF forms of alcoholic hepatitis.175 Although experimental
molecules. The sequential infusions might have increased studies suggested that Nacetylcysteine was the best can-
the risk of infection. Besides the deleterious effect of anti- didate, this drug did not improve liver injury or patient
TNF antibodies on the risk of infection in humans, these outcome in randomized studies.176,177 These molecules
agents might also inhibit hepatic regeneration via the could be considered preventive treatment to reduce the
blockade of TNFR1 and promote apoptosis as suggested risk of disease progression, although prospective data are
by experimentaldata.161 needed. Conversely, the combination of Nacetylcysteine
with prednisolone was associated with a greater early
Pentoxifylline improvement in liver function and a lower incidence of
In a randomized controlled trial in severe alcoholic hepatorenal syndrome than prednisolone alone, associ-
hepatitis, treatment with pentoxifylline (a xanthine) ated with a trend towards improved short-term survival.178
improved survival compared to placebo.162 After pub- The mechanisms underlying this synergistic effect have
lication of this initial study, two trials were published in not yet been elucidated. Further studies investigating this
small populations showing that pentoxifylline might be as combination are needed.

NATURE REVIEWS | GASTROENTEROLOGY & HEPATOLOGY VOLUME 12 | APRIL 2015 | 237

2015 Macmillan Publishers Limited. All rights reserved
REVIEWS

Potential new treatments mortality in the control groups ranging from 0100%.140,183
Infection is a major driver of outcome in patients with The development of the Maddrey discriminant score was
decompensated cirrhosis179 and severe alcoholic hepati- a major step in accurately calculating an adequate sample
tis.180 The incidence of infection in patients with severe size based on a robust hypothesis of mortality in untreated
alcoholic hepatitis is ~25% at admission, and another 25% patients.184 Some experts do not recommend liver biopsy
of patients will develop an infectious event in the 2months to diagnose alcoholic hepatitis and define disease sever-
after the initiation of corticosteroids.180 The main mecha- ity. They suggest using combined clinical and biological
nisms of this increased susceptibility to infections include criteria that have not yet been validated as diagnostic tools.
deficiencies in complement, decreased phagocytic abili- Thus, studies that are not based on a histological diag-
ties of macrophages, neutrophil dysfunction, increased nosis of disease have a 1050% risk of including patients
bacterial translocation and impaired activation of immu- without disease.1,185,186 These false positive inclusions sub-
nity.179,181 Owing to this increased susceptibility to infec- stantially bias the statistical hypothesis by increasing the
tions, it is tempting to evaluate the therapeutic role of risk of typeI and typeII errors, because, by definition, the
antibiotics in ALD. Several studies in this field are ongoing outcome of patients without alcoholic hepatitis should not
(Table1), including absorbable antibiotics (amoxicillin be affected by molecules targeting the specific pathways
clavulanate and quinolones) and nonabsorbable antibiot- driving disease progression.
ics (rifaximin). On the basis of the deleterious effect of Finally, experts and health agencies urgently need to
TLR4 in severe infections and its role in the pathophysiol- agree on the different primary end points for phaseI, II
ogy of ALD, TLR4 antagonists seem to be good potential and III studies, so that pharmaceutical companies and
candidates for testing in alcoholic cirrhosis and hepati- scientific societies can plan on the development of future
tis.182 These molecules are now available and being tested molecules. For example, phaseIII studies evaluating
in many indications including severe infection. safety and the influence of molecules on liver injury must
New drugs should be tested on the basis of the main include patients with a minimal competitive risk of mor-
mechanisms involved in the type of liver injury, according tality to ensure sufficient exposure time. PhaseIII studies
to novel findings from translational research and animal should focus on proof-of-efficacy on the basis of validated
models. For example, in patients with alcoholic hepatitis, primary end points. Mortality alone should not be used
an interesting potential area for investigation, the most as an end point, but to develop and propose surrogate
promising options include molecules targeting inflamma- markers strongly associated with outcome, such as early
tion such as TLR4 antagonists or IL1 receptor antagonists improvement of liver function.143,145,168,180
(such as anakinra), or those targeting bacterial translo-
cation, apoptosis (for example, emricasan) and/or liver Conclusions
regeneration, as well as antibiotics. In patients in the early Scientific studies have made major progress in understand-
stages of ALD, probiotics, prebiotics, molecules limiting ing the pathogenesis of and the effect of treatment in ALD.
oxidative stress and alleviating steatosis and/or reducing This Review discusses the need to build a bridge from the
the progression of fibrosis should be developed, in close bench to the bedside as well as to adapt the design of future
association with the management of alcohol dependence studies to the histological lesions of the disease and to mor-
(Figure3). The main ongoing studies of new targeted tality. For example, the sample size in studies evaluating
treatments in ALD can be found in Table1. molecules to reduce disease progression in patients with
alcoholic steatosis must be extremely large with an end
Requirements for study designs point that has been accepted by health agencies, such as
As in other fields of medical research, an urgent need the risk of developing cirrhosis. The study populations in
exists to create the optimal scientific conditions for the these trials will have to be as large as those in cardiovas-
development of new drugs in the area of ALD. These cular studies and will have to use noninvasive markers to
conditions include a clear definition of disease, a rational assess progression to cirrhosis. On the other hand, studies
evaluation of severity to calculate sample size, the choice on alcoholic hepatitis must focus on short-term outcome
and validation of end points, the development and correct and the development of surrogate markers to identify and
assessment of surrogate markers of outcome, as well as validate new therapeutic pathways.
study designs for phaseI to phaseIII trials. This scien-
tific approach is necessary to correctly assess the effect
Review criteria
of tested strategies on chosen end points. For example, a
study design in heavy drinkers with steatosis alone should We searched PubMed using the following terms:
focus on the long-term progression of disease and use the ethanol; liver regeneration; animal models;
steatosis; pathogenesis of alcoholic liver disease;
development of cirrhosis as an end point. When consider-
oxidative stress; pentoxifylline; corticosteroids;
ing the expected risk, a very large sample size is needed; liver biopsy; N-acetylcysteine; tumor necrosis
thus noninvasive markers of fibrosis must be used because factor; anti-TNF antibodies; endoplasmic reticulum
liver biopsy cannot be practiced in this case. stress; dysbiosis; probiotics; translational
Most initial studies testing molecules in the field of research; cytochrome P450; and alcoholic liver
alcoholic hepatitis underestimated the importance of this disease, alcoholic hepatitis. All selected papers were
scientific approach. In particular there was no precise full-text in English. We searched the reference lists of
identified papers for further relevant papers.
definition of the severity of liver disease, as shown by

238 | APRIL 2015 | VOLUME 12  www.nature.com/nrgastro

2015 Macmillan Publishers Limited. All rights reserved
 REVIEWS

1. European Association for the Study of the Autophagyin alcohol-induced liver diseases. element binding protein-1c (SREBP-1c). Exp. Biol.
Liver.EASL clinical practical guidelines: Alcohol. Clin. Exp. Res. 36, 13011308 (2012). Med (Maywood) 232, 614621 (2007).
management of alcoholic liver disease. 23. Fernandez-Checa, J.C., Garcia-Ruiz, C., 41. Lawler, J.F. Jr, Yin, M., Diehl, A.M., Roberts, E.
J.Hepatol. 57, 399420 (2012). Ookhtens, M. & Kaplowitz, N. Impaired uptake of &Chatterjee, S. Tumor necrosis factor-
2. Gao, B. & Bataller, R. Alcoholic liver disease: glutathione by hepatic mitochondria from chronic stimulates the maturation of sterol regulatory
pathogenesis and new therapeutic targets. ethanol-fed rats. Tracer kinetic studies invitro element binding protein1 in human hepatocytes
Gastroenterology 141, 15721585 (2011). and invivo and susceptibility to oxidant stress. through the action of neutral sphingomyelinase.
3. Lucey, M.R., Mathurin, P. & Morgan, T.R. Alcoholic J.Clin. Invest. 87, 397405 (1991). J. Biol. Chem. 273, 50535059 (1998).
hepatitis. N. Engl. J. Med. 360, 27582769 24. Fernandez-Checa, J.C., Ookhtens, M. 42. Nakajima, T. etal. Peroxisome proliferator-
(2009). &Kaplowitz, N. Effects of chronic ethanol activated receptor alpha protects against
4. Mathurin, P. etal. Fibrosis progression occurs feeding on rat hepatocytic glutathione. alcohol-induced liver damage. Hepatology 40,
ina subgroup of heavy drinkers with typical Relationship of cytosolic glutathione to efflux 972980 (2004).
histological features. Aliment. Pharmacol. Ther. and mitochondrial sequestration. J. Clin. Invest. 43. Galli, A., Pinaire, J., Fischer, M., Dorris, R.
25, 10471054 (2007). 83, 12471252 (1989). &Crabb, D.W. The transcriptional and DNA
5. Pessione, F. etal. Five-year survival predictive 25. Hirano, T., Kaplowitz, N., Tsukamoto, H., binding activity of peroxisome proliferator-
factors in patients with excessive alcohol intake Kamimura, S. & Fernandez-Checa, J.C. activated receptor is inhibited by ethanol
and cirrhosis. Effect of alcoholic hepatitis, Hepaticmitochondrial glutathione depletion metabolism. A novel mechanism for the
smoking and abstinence. Liver Int. 23, 4553 andprogression of experimental alcoholic liver development of ethanol-induced fatty liver. J. Biol.
(2003). disease in rats. Hepatology 16, 14231427 Chem. 276, 6875 (2001).
6. Verrill, C., Markham, H., Templeton, A., Carr, N.J. (1992). 44. Fischer, M., You, M., Matsumoto, M. &Crabb,D.W.
& Sheron, N. Alcohol-related cirrhosisearly 26. Wheeler, M.D. etal. Overexpression of Peroxisome proliferator-activatedreceptor alpha
abstinence is a key factor in prognosis, even in manganese superoxide dismutase prevents (PPARalpha) agonist treatment reverses
the most severe cases. Addiction 104, 768774 alcohol-induced liver injury in the rat. J. Biol. PPARalpha dysfunction and abnormalities in
(2009). Chem. 276, 3666436672 (2001). hepatic lipid metabolism in ethanol-fed mice.
7. Mendenhall, C.L. Anabolic steroid therapy as 27. Fernandez-Checa, J.C., Kaplowitz, N., J.Biol. Chem. 278, 2799728004 (2003).
anadjunct to diet in alcoholic hepatic steatosis. GarciaRuiz, C. & Colell, A. Mitochondrial 45. Molina, P.E. Alcoholintoxicating roadblocks
Am. J. Dig. Dis. 13, 783791 (1968). glutathione: importance and transport. andbottlenecks in hepatic protein and lipid
8. Galambos, J.T. Natural history of alcoholic Semin.Liver Dis. 18, 389401 (1998). metabolism. Am. J. Physiol. Endocrinol. Metab.
hepatitis. 3. Histological changes. 28. Fernandez-Checa, J.C., Hirano, T., Tsukamoto, H. 295, E1E2 (2008).
Gastroenterology 63, 10261035 (1972). & Kaplowitz, N. Mitochondrial glutathione 46. Zhou, G. etal. Role of AMP-activated protein
9. Pares, A., Caballeria, J., Bruguera, M., Torres, M. depletion in alcoholic liver disease. Alcohol 10, kinase in mechanism of metformin action. J. Clin.
& Rodes, J. Histological course of alcoholic 469475 (1993). Invest. 108, 11671174 (2001).
hepatitis. Influence of abstinence, sex and extent 29. Gong, P., Cederbaum, A.I. & Nieto, N. 47. Garcia-Villafranca, J., Guillen, A. & Castro, J.
of hepatic damage. J. Hepatol. 2, 3342 (1986). Hemeoxygenase1 protects HepG2 cells against Ethanol consumption impairs regulation of fatty
10. Norberg, A., Jones, A.W., Hahn, R.G. cytochrome P450 2E1-dependent toxicity. acid metabolism by decreasing the activity
&Gabrielsson, J.L. Role of variability in FreeRadic. Biol. Med. 36, 307318 (2004). ofAMP-activated protein kinase in rat liver.
explaining ethanol pharmacokinetics: research 30. Imrie, D. & Sadler, K.C. Stress management: Biochimie 90, 460466 (2008).
and forensic applications. Clin. Pharmacokinet. how the unfolded protein response impacts fatty 48. You, M. & Crabb, D.W. Recent advances in
42, 131 (2003). liver disease. J. Hepatol. 57, 11471151 (2012). alcoholic liver disease.I. Minireview:
11. Cederbaum, A.I. Alcohol metabolism. 31. Malhi, H. & Kaufman, R.J. Endoplasmic molecularmechanisms of alcoholic fatty liver.
Clin.LiverDis. 16, 667685 (2012). reticulum stress in liver disease. J. Hepatol. 54, Am. J. Physiol. 287, G1G6 (2004).
12. Lieber, C.S. ALCOHOL: its metabolism and 795809 (2011). 49. Yin, H. etal. Deletion of SIRT1 from
interaction with nutrients. Ann. Rev. Nutr. 20, 32. Ji, C., Deng, Q. & Kaplowitz, N. Role of TNF- in hepatocytesin mice disrupts lipin1 signaling
395430 (2000). ethanol-induced hyperhomocysteinemia and and aggravates alcoholic fatty liver.
13. Lieber, C.S. Alcoholic fatty liver: its pathogenesis murine alcoholic liver injury. Hepatology 40, Gastroenterology 146, 801811 (2014).
and mechanism of progression to inflammation 442451 (2004). 50. You, M., Liang, X., Ajmo, J.M. & Ness, G.C.
and fibrosis. Alcohol 34, 919 (2004). 33. Petrasek, J. etal. STINGIRF3 pathway links Involvement of mammalian sirtuin 1 in the
14. Lu, Y. & Cederbaum, A.I. CYP2E1 and oxidative endoplasmic reticulum stress with hepatocyte action of ethanol in the liver. Am. J. Physiol. 294,
liver injury by alcohol. Free Radic. Biol. Med. 44, apoptosis in early alcoholic liver disease. Proc. G892G898 (2008).
723738 (2008). Natl Acad. Sci. USA 110, 1654416549 (2013). 51. Jeong, W.I. etal. Paracrine activation of hepatic
15. Lieber, C.S. Microsomal ethanol-oxidizing 34. Esfandiari, F., Villanueva, J.A., Wong, D.H., CB1 receptors by stellate cell-derived
system (MEOS): the first 30years (1968 French, S.W. & Halsted, C.H. Chronic ethanol endocannabinoids mediates alcoholic fatty liver.
1998)a review. Alcohol. Clin. Exp. Res. 23, feeding and folate deficiency activate hepatic Cell. Metab. 7, 227235 (2008).
9911007 (1999). endoplasmic reticulum stress pathway in 52. Louvet, A. etal. Cannabinoid CB2 receptors
16. Cederbaum, A.I. Cytochrome P450 2E1- micropigs. Am. J. Physiol. 289, G54G63 (2005). protect against alcoholic liver disease by
dependent oxidant stress and upregulation of anti- 35. Ji, C. Dissection of endoplasmic reticulum stress regulating Kupffer cell polarization in mice.
oxidant defense in liver cells. J. Gastroenterol. signaling in alcoholic and non-alcoholic liver Hepatology 54, 12171226 (2011).
Hepatol. 21 (Suppl. 3), S22S25 (2006). injury. J. Gastroenterol. Hepatol. 23 (Suppl. 1), 53. Bykov, I. etal. Hepatic gene expression and
17. Wheeler, M.D. etal. The role of Kupffer cell S16S24 (2008). lipidparameters in complement C3/ mice
oxidant production in early ethanol-induced liver 36. Crabb, D.W. & Liangpunsakul, S. Alcohol and thatdo not develop ethanol-induced steatosis.
disease. Free Radic. Biol. Med. 31, 15441549 lipid metabolism. J. Gastroenterol. Hepatol. 21 J.Hepatol. 46, 907914 (2007).
(2001). (Suppl. 3), S56S60 (2006). 54. Cohen, J.I., Roychowdhury, S., McMullen, M.R.,
18. Philippe, M.A., Ruddell, R.G. & Ramm, G.A. 37. Ji, C. & Kaplowitz, N. Betaine decreases Stavitsky, A.B. & Nagy, L.E. Complement
Role of iron in hepatic fibrosis: one piece in the hyperhomocysteinemia, endoplasmic reticulum andalcoholic liver disease: role of C1q in the
puzzle. World J. Gastroenterol. 13, 47464754 stress, and liver injury in alcohol-fed mice. pathogenesis of ethanol-induced liver injury in
(2007). Gastroenterology 124, 14881499 (2003). mice. Gastroenterology 139, 664674. (2010).
19. Tsukamoto, H. etal. Experimental liver cirrhosis 38. Hernandez-Gea, V. etal. Endoplasmic reticulum 55. Pritchard, M.T. etal. Differential contributions of
induced by alcohol and iron. J. Clin. Invest. 96, stress induces fibrogenic activity in hepatic C3, C5, and decay-accelerating factor to ethanol-
620630 (1995). stellate cells through autophagy. J. Hepatol. 59, induced fatty liver in mice. Gastroenterology 132,
20. Luedde, T., Kaplowitz, N. & Schwabe, R.F. 98104 (2013). 11171126 (2007).
Celldeath and cell death responses in liver 39. You, M., Fischer, M., Deeg, M.A. & Crabb, D.W. 56. Tomita, K. etal. Pioglitazone prevents alcohol-
disease: mechanisms and clinical relevance. Ethanol induces fatty acid synthesis pathways by induced fatty liver in rats through up-regulation
Gastroenterology 147, 765783 (2014). activation of sterol regulatory element-binding of cMet. Gastroenterology 126, 873885 (2004).
21. Brenner, C., Galluzzi, L., Kepp, O. & Kroemer, G. protein (SREBP). J. Biol. Chem. 277, 2934229347 57. Naveau, S. etal. Harmful effect of adipose
Decoding cell death signals in liver (2002). tissue on liver lesions in patients with alcoholic
inflammation. J. Hepatol. 59, 583594 (2013). 40. Endo, M., Masaki, T., Seike, M. & Yoshimatsu, H. liver disease. J. Hepatol. 52, 895902 (2010).
22. Dolganiuc, A., Thomes, P.G., Ding, W.X., TNF- induces hepatic steatosis in mice by 58. Naveau, S. Excess weight risk factor for alcoholic
Lemasters, J.J. & Donohue, T.M. Jr. enhancing gene expression of sterol regulatory liver disease. Hepatology 25, 108111 (1997).

NATURE REVIEWS | GASTROENTEROLOGY & HEPATOLOGY VOLUME 12 | APRIL 2015 | 239

2015 Macmillan Publishers Limited. All rights reserved
REVIEWS

59. Raynard, B. etal. Risk factors of fibrosis in 77. Wheeler, M.D. & Thurman, R.G. Up-regulation of 95. Szabo, G. & Mandrekar, P. A recent perspective
alcohol-induced liver disease. Hepatology 35, CD14 in liver caused by acute ethanol involves on alcohol, immunity, and host defense.
635638 (2002). oxidant-dependent AP1 pathway. J. Biol. Chem. Alcohol.Clin. Exp. Res. 33, 220232 (2009).
60. Dunn, W. etal. Modest alcohol consumption 278, 84358441 (2003). 96. Brun, P., Castagliuolo, I., Pinzani, M., Palu, G.
isassociated with decreased prevalence of 78. Hritz, I. etal. The critical role of toll-like receptor &Martines, D. Exposure to bacterial cell wall
steatohepatitis in patients with non-alcoholic fatty (TLR) 4 in alcoholic liver disease is independent products triggers an inflammatory phenotype
liver disease (NAFLD). J. Hepatol. 57, 384391 of the common TLR adapter MyD88. Hepatology inhepatic stellate cells. Am. J. Physiol. 289,
(2012). 48, 12241231 (2008). G571G578 (2005).
61. Kang, L. et al. Chronic ethanol-induced insulin 79. Zima, T. & Kalousova, M. Oxidative stress and 97. Paik, Y.H. et al. Toll-like receptor 4 mediates
resistance is associated with macrophage signal transduction pathways in alcoholic liver inflammatory signaling by bacterial
infiltration into adipose tissue and altered disease. Alcohol. Clin. Exp. Res. 29 (11 Suppl.) lipopolysaccharide in human hepatic stellate
expression of adipocytokines. Alcohol. Clin. Exp. 110S115S (2005). cells. Hepatology 37, 10431055 (2003).
Res. 31, 15811588 (2007). 80. De Minicis, S. & Brenner, D.A. Oxidative stress 98. Thirunavukkarasu, C., Uemura, T., Wang, L.F.,
62. Xu, A. et al. The fat-derived hormone adiponectin in alcoholic liver disease: role of NADPH oxidase Watkins, S.C. & Gandhi, C.R. Normal rat
alleviates alcoholic and nonalcoholic fatty liver complex. J. Gastroenterol. Hepatol. 23 (Suppl. 1) hepatic stellate cells respond to endotoxin in
diseases in mice. J. Clin. Invest. 112, 91100 S98S103 (2008). LBP-independent manner to produce inhibitor(s)
(2003). 81. Uesugi, T., Froh, M., Arteel, G.E., Bradford, B.U. of DNA synthesis in hepatocytes. J. Cell. Physiol.
63. Yamauchi, T. etal. Adiponectin stimulates & Thurman, R.G. Toll-like receptor 4 is involved 204, 654665 (2005).
glucose utilization and fatty-acid oxidation in the mechanism of early alcohol-induced liver 99. Yan, A.W. etal. Enteric dysbiosis associated
byactivating AMP-activated protein kinase. injury in mice. Hepatology 34, 101108 (2001). with a mouse model of alcoholic liver disease.
Nat.Med. 8, 12881295 (2002). 82. Xiong, S. etal. Hepatic macrophage iron Hepatology 53, 96105 (2011).
64. You, M., Considine, R.V., Leone, T.C., Kelly, D.P. aggravates experimental alcoholic 100. Mutlu, E. et al. Intestinal dysbiosis: a possible
&Crabb, D.W. Role of adiponectin in the steatohepatitis. Am. J. Physiol. 295, G512G521 mechanism of alcohol-induced endotoxemia
protective action of dietary saturated fat (2008). andalcoholic steatohepatitis in rats.
againstalcoholic fatty liver in mice. Hepatology 83. Lin, M., Rippe, R.A., Niemela, O., Brittenham, G. Alcohol.Clin. Exp. Res. 33, 18361846 (2009).
42, 568577 (2005). & Tsukamoto, H. Role of iron in NFkappa B 101. Chen, P., Starkel, P., Turner, J.R., Ho, S.B.
65. Song, Z., Zhou, Z., Deaciuc, I., Chen, T. activation and cytokine gene expression by rat &Schnabl, B. Dysbiosis-induced intestinal
&McClain, C.J. Inhibition of adiponectin hepatic macrophages. Am. J. Physiol. 272, inflammation activates tumor necrosis factor I
production by homocysteine: a potential G1355G1364 (1997). and mediates alcoholic liver disease in mice.
mechanism for alcoholic liver disease. 84. She, H. et al. Iron activates NFkappaB in Kupffer Hepatology http://dx.doi.org/10.1002/
Hepatology 47, 867879 (2008). cells. Am. J. Physiol. 283, G719G726 (2002). hep.27489.
66. Mathurin, P. et al. Exacerbation of alcoholic liver 85. Tsukamoto, H. et al. Iron primes hepatic 102. Schnabl, B. & Brenner, D.A. Interactions between
injury by enteral endotoxin in rats. Hepatology macrophages for NFkappaB activation in alcoholic the intestinal microbiome and liver diseases.
32, 10081017 (2000). liver injury. Am. J. Physiol. 277, G1240G1250 Gastroenterology 146, 15131524 (2014).
67. Parlesak, A., Schafer, C., Schutz, T., Bode, J.C. (1999). 103. Chen, P. etal. Supplementation of saturated
&Bode, C. Increased intestinal permeability to 86. Afford, S.C. etal. Distinct patterns of chemokine long-chain fatty acids maintains intestinal
macromolecules and endotoxemia in patients expression are associated with leukocyte eubiosis and reduces ethanol-induced liver
with chronic alcohol abuse in different stages recruitment in alcoholic hepatitis and alcoholic injury in mice. Gastroenterology 148, 203214
ofalcohol-induced liver disease. J. Hepatol. 32, cirrhosis. J. Pathol. 186, 8289 (1998). (2014).
742747 (2000). 87. Colmenero, J. etal. Hepatic expression of 104. Mutlu, E.A. etal. Colonic microbiome is altered
68. Tamai, H., Horie, Y., Kato, S., Yokoyama, H. candidate genes in patients with alcoholic in alcoholism. Am. J. Physiol. 302, G966G978
&Ishii, H. Long-term ethanol feeding enhances hepatitis: correlation with disease severity. (2012).
susceptibility of the liver to orally administered Gastroenterology 132, 687697 (2007). 105. Qin, N. etal. Alterations of the human gut
lipopolysaccharides in rats. Alcohol. Clin. Exp. Res. 88. Maltby, J., Wright, S., Bird, G. & Sheron, N. microbiome in liver cirrhosis. Nature 513, 5964
26 (8 Suppl.) 75S80S (2002). Chemokine levels in human liver homogenates: (2014).
69. Fujimoto, M. etal. Plasma endotoxin and serum associations between GRO alpha and 106. Forsyth, C.B. et al. Lactobacillus GG treatment
cytokine levels in patients with alcoholic hepatitis: histopathological evidence of alcoholic hepatitis. ameliorates alcohol-induced intestinal oxidative
relation to severity of liver disturbance. Hepatology 24, 11561160 (1996). stress, gut leakiness, and liver injury in a rat
Alcohol.Clin. Exp. Res. 24 (4 Suppl.) 48S54S 89. Taieb, J. etal. Blood neutrophil functions and model of alcoholic steatohepatitis. Alcohol 43,
(2000). cytokine release in severe alcoholic hepatitis: 163172 (2009).
70. Hanck, C., Rossol, S., Bocker, U., Tokus, M. effect of corticosteroids. J. Hepatol. 32, 579586 107. Stadlbauer, V. et al. Effect of probiotic treatment
&Singer, M.V. Presence of plasma endotoxin is (2000). on deranged neutrophil function and cytokine
correlated with tumour necrosis factor receptor 90. Barnes, M.A. etal. Macrophage migration responses in patients with compensated
levels and disease activity in alcoholic cirrhosis. inhibitory factor contributes to ethanol-induced alcoholic cirrhosis. J. Hepatol. 48, 945951
Alcohol Alcohol. 33, 606608 (1998). liver injury by mediating cell injury, (2008).
71. Bhagwandeen, B.S., Apte, M., Manwarring, L. steatohepatitis, and steatosis. Hepatology 57, 108. Kirpich, I.A. etal. Probiotics restore bowel flora
&Dickeson, J. Endotoxin induced hepatic 19801991 (2013). and improve liver enzymes in human alcohol-
necrosis in rats on an alcohol diet. J. Pathol. 91. Mandrekar, P., Ambade, A., Lim, A., Szabo, G. induced liver injury: a pilot study. Alcohol 42,
152, 4753 (1987). &Catalano, D. An essential role for monocyte 675682 (2008).
72. Adachi, Y., Moore, L.E., Bradford, B.U., Gao, W. chemoattractant protein1 in alcoholic liver 109. Delzenne, N.M., Neyrinck, A.M., Backhed, F.
& Thurman, R.G. Antibiotics prevent liver injury injury: regulation of proinflammatory cytokines &Cani, P.D. Targeting gut microbiota in obesity:
in rats following long-term exposure to ethanol. and hepatic steatosis in mice. Hepatology 54, effects of prebiotics and probiotics. Nat. Rev.
Gastroenterology 108, 218224 (1995). 21852197 (2011). Endocrinol. 7, 639646 (2011).
73. Nanji, A.A., Khettry, U. & Sadrzadeh, S.M. 92. Wan, J. etal. M2 Kupffer cells promote M1 110. Malaguarnera, G., Giordano, M., Nunnari, G.,
Lactobacillus feeding reduces endotoxemia and Kupffer cell apoptosis: a protective mechanism Bertino, G. & Malaguarnera, M. Gut microbiota
severity of experimental alcoholic liver (disease). against alcoholic and nonalcoholic fatty liver inalcoholic liver disease: pathogenetic role and
Proc. Soc. Exp. Biol Med. 205, 243247 (1994). disease. Hepatology 59, 130142 (2014). therapeutic perspectives. World J. Gastroenterol.
74. Gustot, T. etal. Differential liver sensitization to 93. Byun, J.S., Suh, Y.G., Yi, H.S., Lee, Y.S. 20, 1663916648 (2014).
toll-like receptor pathways in mice with alcoholic &Jeong, W.I. Activation of Toll-like receptor 3 111. Diehl, A.M. Recent events in alcoholic liver
fatty liver. Hepatology 43, 9891000 (2006). attenuates alcoholic liver injury by stimulating disease, V. effects of ethanol on liver
75. Oak, S., Mandrekar, P., Catalano, D., Kodys, K. Kupffer cells and stellate cells to produce regeneration. Am. J. Physiol. 288, G1G6 (2005).
&Szabo, G. TLR2- and TLR4-mediated signals interleukin10 in mice. J. Hepatol. 58, 342349 112. Michalopoulos, G.K. Liver regeneration:
determine attenuation or augmentation of (2013). alternative epithelial pathways. Int. J. Biochem.
inflammation by acute alcohol in monocytes. 94. Minagawa, M., Deng, Q., Liu, Z.X., Tsukamoto,H. Cell Biol. 43, 173179 (2011).
J.Immunol. 176, 76287635 (2006). & Dennert, G. Activated natural killer Tcells 113. Cosgrove, B.D. et al. An inducible autocrine
76. Guha, M. & Mackman, N. LPS induction of gene induce liver injury by Fas and tumor necrosis cascade regulates rat hepatocyte proliferation
expression in human monocytes. Cell. Signal. factor-alpha during alcohol consumption. and apoptosis responses to tumor necrosis
13, 8594 (2001). Gastroenterology 126, 13871399 (2004). factor-alpha. Hepatology 48, 276288 (2008).

240 | APRIL 2015 | VOLUME 12  www.nature.com/nrgastro

2015 Macmillan Publishers Limited. All rights reserved
 REVIEWS

114. Freimuth, J. etal. Loss of caspase8 in 131. Mathews, S., Xu, M., Wang, H., Bertola, A. 150. Kendrick, S.F., Henderson, E., Palmer, J.,
hepatocytes accelerates the onset of liver &Gao, B. Animals models of gastrointestinal Jones,D.E. & Day, C.P. Theophylline improves
regeneration in mice through premature and liver diseases. Animal models of alcohol- steroid sensitivity in acute alcoholic hepatitis.
nuclearfactor kappa B activation. Hepatology induced liver disease: pathophysiology, Hepatology 52, 126131 (2010).
58, 17791789 (2013). translational relevance, and challenges. Am. J. 151. di Mambro, A.J. et al. Invitro steroid resistance
115. Wang, H. etal. Interplay of hepatic and myeloid Physiol. 306, G819G823 (2014). correlates with outcome in severe alcoholic
signal transducer and activator of transcription 3 132. Affo, S. etal. Transcriptome analysis identifies hepatitis. Hepatology 53, 13161322 (2011).
in facilitating liver regeneration via tempering TNF superfamily receptors as potential 152. Dhanda, A.D. etal. Long-term outcome in
innate immunity. Hepatology 51, 13541362 therapeutic targets in alcoholic hepatitis. Gut patients with severe alcoholic hepatitis can be
(2010). 62, 452460 (2013). reliably determined using an invitro measure of
116. Webber, E.M., Bruix, J., Pierce, R.H. 133. Altamirano, J. etal. A histologic scoring system steroid sensitivity. Hepatology http://dx.doi.org/
&Fausto,N. Tumor necrosis factor primes for prognosis of patients with alcoholic hepatitis. 10.1002/hep.27211.
hepatocytes for DNA replication in the rat. Gastroenterology 146, 12311239 (2014). 153. Kendrick, S.F. etal. Acetate, the key modulator
Hepatology 28, 12261234 (1998). 134. Affo, S. etal. CCL20 mediates of inflammatory responses in acute alcoholic
117. Xu, M.J. etal. Liver is the major source of lipopolysaccharide induced liver injury and is a hepatitis. Hepatology 51, 19881997 (2010).
elevated serum lipocalin2 levels after bacterial potential driver of inflammation and fibrosis in 154. Hamdi, H. etal. Glucocorticoid-induced leucine
infection or partial hepatectomy: a critical role alcoholic hepatitis. Gut 63, 17821792 (2014). zipper: A key protein in the sensitization of
for IL6/STAT3. Hepatology 61, 692702 135. Dominguez, M. etal. Hepatic expression of CXC monocytes to lipopolysaccharide in alcoholic
(2015). chemokines predicts portal hypertension and hepatitis. Hepatology 46, 19861992 (2007).
118. Roskams, T. etal. Oxidative stress and oval survival in patients with alcoholic hepatitis. 155. Mookerjee, R.P. et al. Tumour necrosis factor
cellaccumulation in mice and humans with Gastroenterology 136, 16391650 (2009). alpha is an important mediator of portal and
alcoholic and nonalcoholic fatty liver disease. 136. Morales-Ibanez, O. etal. Human and systemic haemodynamic derangements in
Am. J. Physiol. 163, 13011311 (2003). experimental evidence supporting a role for alcoholic hepatitis. Gut 52, 11821187 (2003).
119. Koteish, A., Yang, S., Lin, H., Huang, X. osteopontin in alcoholic hepatitis. Hepatology 156. Spahr, L. etal. Combination of steroids with
&Diehl,A.M. Chronic ethanol exposure 58, 17421756 (2013). infliximab or placebo in severe alcoholic
potentiates lipopolysaccharide liver injury 137. Yong, K.J. etal. Oncofetal gene SALL4 in hepatitis: a randomized controlled pilot study.
despite inhibiting Jun Nterminal kinase and aggressive hepatocellular carcinoma. N. Engl. J. J.Hepatol. 37, 448455 (2002).
caspase 3 activation. J. Biol. Chem. 277, Med. 368, 22662276 (2013). 157. Tilg, H. etal. Anti-tumor necrosis factor-alpha
1303713044 (2002). 138. OShea, R.S., Dasarathy, S. & McCullough, A.J. monoclonal antibody therapy in severe alcoholic
120. Dippold, R.P., Vadigepalli, R., Gonye, G.E., Alcoholic liver disease. Am. J. Gastroenterol. 105, hepatitis. J. Hepatol. 38, 419425 (2003).
Patra,B. & Hoek, J.B. Chronic ethanol feeding 1432 (2010). 158. Sharma, P., Kumar, A., Sharma, B.C. & Sarin,S.K.
alters miRNA expression dynamics during 139. Mathurin, P. etal. Corticosteroids improve short- Infliximab monotherapy for severe alcoholic
liverregeneration. Alcohol. Clin. Exp. Res. 37 term survival in patients with severe alcoholic hepatitis and predictors of survival: anopen
(Suppl.1) E59E69 (2013). hepatitis (AH): individual data analysis of label trial. J. Hepatol. 50, 584591 (2009).
121. Jung, Y. etal. Accumulation of hedgehog- thelast three randomized placebo controlled 159. Naveau, S. etal. A double-blind randomized
responsive progenitors parallels alcoholic doubleblind trials of corticosteroids in controlled trial of infliximab associated with
liverdisease severity in mice and humans. severeAH. J. Hepatol. 36, 480487 (2002). prednisolone in acute alcoholic hepatitis.
Gastroenterology 134, 15321543 (2008). 140. Mathurin, P. etal. Corticosteroids improve short- Hepatology 39, 13901397 (2004).
122. Sancho-Bru, P. et al. Liver progenitor cell markers term survival in patients with severe alcoholic 160. Boetticher, N.C., etal. A randomized, double-
correlate with liver damage and predict short- hepatitis: meta-analysis of individual patient blinded, placebo-controlled multicenter trial of
term mortality in patients with alcoholic data. Gut 60, 255260 (2011). etanercept in the treatment of alcoholic hepatitis.
hepatitis. Hepatology 55, 19311941 (2012). 141. Forrest, E.H. etal. Analysis of factors predictive Gastroenterology 135, 19531960 (2008).
123. Dubuquoy, L. etal. Progenitor cell expansion of mortality in alcoholic hepatitis and derivation 161. Brenndorfer, E.D. etal. Anti-tumor necrosis
and impaired hepatocyte regeneration in and validation of the Glasgow alcoholic hepatitis factor alpha treatment promotes apoptosis
explanted livers from alcoholic hepatitis. Gut score. Gut 54, 11741179 (2005). andprevents liver regeneration in a transgenic
(inpress). 142. Dominguez, M. etal. A new scoring system mouse model of chronic hepatitis C. Hepatology
124. Tsuchiya, M. etal. Interstrain differences in liver forprognostic stratification of patients with 52, 15531563 (2010).
injury and one-carbon metabolism in alcohol-fed alcoholic hepatitis. Am. J. Gastroenterol. 103, 162. Akriviadis, E. et al. Pentoxifylline improves
mice. Hepatology 56, 130139 (2012). 27472756 (2008). shortterm survival in severe acute alcoholic
125. Lieber, C.S., Jones, D.P. & Decarli, L.M. 143. Louvet, A. etal. The Lille model: a new tool hepatitis: a double-blind, placebo-controlled trial.
Effectsof prolonged ethanol intake: production fortherapeutic strategy in patients with severe Gastroenterology 119, 16371648 (2000).
of fatty liver despite adequate diets. J. Clin. Invest. alcoholic hepatitis treated with steroids. 163. Sidhu, S.S. etal. Corticosteroid plus
44, 10091021 (1965). Hepatology 45, 13481354 (2007). pentoxifylline is not better than corticosteroid
126. Lieber, C.S., DeCarli, L.M. & Sorrell, M.F. 144. Centre Hospitalier Regional, Universitaire alone for improving survival in severe alcoholic
Experimental methods of ethanol deLille. Lille Model. LilleModel.com [online], hepatitis (COPE trial). Dig. Dis. Sci. 57, 16641671
administration. Hepatology 10, 501510 http://www.lillemodel.com/ (2015). (2012).
(1989). 145. Mathurin, P. etal. Early change in bilirubin levels 164. De, B.K. et al. Pentoxifylline versus prednisolone
127. Tsukamoto, H. etal. Severe and progressive is an important prognostic factor in severe for severe alcoholic hepatitis: a randomized
steatosis and focal necrosis in rat liver induced alcoholic hepatitis treated with prednisolone. controlled trial. World J. Gastroenterol. 15,
by continuous intragastric infusion of ethanol Hepatology 38, 13631369 (2003). 16131619 (2009).
and low fat diet. Hepatology 5, 224232 146. Iimuro, Y., Gallucci, R.M., Luster, M.I., Kono, H. 165. Doherty, G.M., Jensen, J.C., Alexander, H.R.,
(1985). & Thurman, R.G. Antibodies to tumor necrosis Buresh, C.M. & Norton, J.A. Pentoxifylline
128. Tsukamoto, H., Reidelberger, R.D., factor alfa attenuate hepatic necrosis and suppression of tumor necrosis factor gene
French,S.W. & Largman, C. Long-term inflammation caused by chronic exposure to transcription. Surgery 110, 192198 (1991).
cannulation model for blood sampling and ethanol in the rat. Hepatology 26, 15301537 166. Han, J., Thompson, P. & Beutler, B.
intragastric infusion in the rat. Am. J. Physiol. (1997). Dexamethasone and pentoxifylline inhibit
247, R595R599 (1984). 147. Yin, M. etal. Essential role of tumor necrosis endotoxin-induced cachectin/tumor necrosis
129. von Montfort, C., Beier, J.I., Guo, L., Kaiser, J.P. factor alpha in alcohol-induced liver injury in factor synthesis at separate points in the
& Arteel, G.E. Contribution of the sympathetic mice. Gastroenterology 117, 942952 (1999). signaling pathway. J. Exp. Med. 172, 391394
hormone epinephrine to the sensitizing effect 148. Naveau, S. etal. Plasma levels of soluble tumor (1990).
ofethanol on LPS-induced liver damage in necrosis factor receptors p55 and p75 in 167. Louvet, A. etal. Early switch to pentoxifylline
mice.Am. J. Physiol. 294, G1227G1234 patients with alcoholic liver disease of inpatients with severe alcoholic hepatitis is
(2008). increasing severity. J. Hepatol. 28, 778784 inefficient in non-responders to corticosteroids.
130. Bertola, A., Park, O. & Gao, B. Chronic plus (1998). J. Hepatol. 48, 465470 (2008).
bingeethanol feeding synergistically induces 149. Spahr, L. et al. Soluble TNFR1, but not tumor 168. Mathurin, P. etal. Prednisolone with vs without
neutrophil infiltration and liver injury in mice: necrosis factor alpha, predicts the 3month pentoxifylline and survival of patients with
acritical role for Eselectin. Hepatology 58, mortality in patients with alcoholic hepatitis. severe alcoholic hepatitis: a randomized clinical
18141823 (2013). J.Hepatol. 41, 229234 (2004). trial. JAMA 310, 10331041 (2013).

NATURE REVIEWS | GASTROENTEROLOGY & HEPATOLOGY VOLUME 12 | APRIL 2015 | 241

2015 Macmillan Publishers Limited. All rights reserved
REVIEWS

169. Park, S.H. etal. Pentoxifylline vs. 179. Gustot, T., Durand, F., Lebrec, D., Vincent, J.L. 190. US National Library of Medicine, ClinicalTrials.gov
corticosteroidto treat severe alcoholic &Moreau, R. Severe sepsis in cirrhosis. [online], https://clinicaltrials.gov/ct2/show/
hepatitis:a randomised, non-inferiority, open Hepatology 50, 20222033 (2009). NCT01809132 (2014).
trial. J.Hepatol. 61, 792798 (2014). 180. Louvet, A. etal. Infection in patients with 191. US National Library of Medicine, ClinicalTrials.gov
170. Thursz, M. etal. Steroids or Pentoxifylline for severealcoholic hepatitis treated with steroids: [online], https://clinicaltrials.gov/ct2/show/
Alcoholic Hepatitis: Results of the STOPAH Trial. early response to therapy is the key factor. NCT02281929 (2014).
Hepatology 60, LB1 (2014). Gastroenterology 137, 541548 (2009). 192. US National Library of Medicine, ClinicalTrials.gov
171. Lebrec, D. etal. Pentoxifylline does not decrease 181. Wasmuth, H.E. etal. Patients with acute on [online], https://clinicaltrials.gov/ct2/show/
short-term mortality but does reduce chronic liver failure display sepsis-like immune NCT02116556 (2014).
complications in patients with advanced cirrhosis. paralysis. J. Hepatol. 42, 195201 (2005). 193. US National Library of Medicine, ClinicalTrials.gov
Gastroenterology 138, 17551762 (2010). 182. Leon, C.G., Tory, R., Jia, J., Sivak, O. [online], https://clinicaltrials.gov/ct2/show/
172. Lieber, C.S. SAdenosyll-methionine and &Wasan,K.M. Discovery and development of NCT02326103 (2014).
alcoholic liver disease in animal models: tolllike receptor 4 (TLR4) antagonists: a new 194. US National Library of Medicine, ClinicalTrials.gov
implications for early intervention in human paradigm for treating sepsis and other diseases. [online], https://clinicaltrials.gov/ct2/show/
beings. Alcohol 27, 173177 (2002). Pharm. Res. 25, 17511761 (2008). NCT01922895 (2014).
173. Colell, A. etal. Transport of reduced glutathione 183. Imperiale, T.F. & McCullough, A.J. 195. US National Library of Medicine, ClinicalTrials.gov
in hepatic mitochondria and mitoplasts from Docorticosteroids reduce mortality from [online], https://clinicaltrials.gov/ct2/show/
ethanol-treated rats: effect of membrane alcoholic hepatitis? A meta-analysis of NCT01820208 (2013).
physical properties and Sadenosyll-methionine. therandomized trials. Ann. Intern. Med. 113, 196. US National Library of Medicine, ClinicalTrials.gov
Hepatology 26, 699708 (1997). 299307 (1990). [online], https://clinicaltrials.gov/ct2/show/
174. Garcia-Ruiz, C. Effect of chronic ethanol feeding 184. Maddrey, W.C. etal. Corticosteroid therapy of NCT02039219 (2014).
on glutathione and functional integrity of alcoholic hepatitis. Gastroenterology 75, 193199 197. US National Library of Medicine, ClinicalTrials.gov
mitochondria in periportal and perivenous rat (1978). [online], https://clinicaltrials.gov/ct2/show/
hepatocytes. J. Clin. Invest. 94, 193201 (1994). 185. Hardy, T. etal. White cell count and platelet count NCT01903798 (2013).
175. Phillips, M. et al. Antioxidants versus associate with histological alcoholic hepatitis in 198. US National Library of Medicine, ClinicalTrials.gov
corticosteroids in the treatment of severe jaundiced harmful drinkers. BMC Gastroenterol. [online], https://clinicaltrials.gov/ct2/show/
alcoholic hepatitis-a randomised clinical trial. 13, 55. (2013). NCT02072746 (2014).
J.Hepatol. 44, 784790 (2006). 186. Mookerjee, R.P. etal. The role of liver biopsy 199. US National Library of Medicine, ClinicalTrials.gov
176. Moreno, C. etal. Enteral nutrition with or without inthe diagnosis and prognosis of patients [online], https://clinicaltrials.gov/ct2/show/
Nacetylcysteine in the treatment of severe acute withacute deterioration of alcoholic cirrhosis. NCT01875081 (2014).
alcoholic hepatitis: a randomized multicenter J.Hepatol. 55, 11031111 (2011). 200. US Department of Health and Human
controlled trial. J. Hepatol. 53, 11171122 187. US National Library of Medicine, ClinicalTrials.gov Services.NIH Research Portfolio Online Report
(2010). [online], https://clinicaltrials.gov/ct2/show/ Tools (RePORT) [online], http://projectreporter.
177. Stewart, S. etal. A randomized trial of antioxidant NCT01912404 (2015). nih.gov/Reporter_Viewsh.cfm?sl=
therapy alone or with corticosteroids in acute 188. US National Library of Medicine, ClinicalTrials.gov 13EBC0034F84C1DE7598B8961CAA4A01A2
alcoholic hepatitis. J. Hepatol. 47, 277283 [online], https://clinicaltrials.gov/ct2/show/ FFCEB861BF (2015).
(2007). NCT02161653 (2014).
178. Nguyen-Khac, E. etal. Glucocorticoids plus 189. US National Library of Medicine, ClinicalTrials.gov Author contributions
Nacetylcysteine in severe alcoholic hepatitis. [online], https://clinicaltrials.gov/ct2/show/ Both authors contributed equally to all aspects
N.Engl. J. Med. 365, 17811789 (2011). NCT02019056 (2013). ofthemanuscript.

242 | APRIL 2015 | VOLUME 12  www.nature.com/nrgastro

2015 Macmillan Publishers Limited. All rights reserved