Science in medicine

Liver fibrosis
Ramón Bataller1 and David A. Brenner2
1Liver Unit, Institut de Malalties Digestives i Metabòliques, Hospital Clinic, Institut d’Investigació Biomèdiques August Pi i Sunyer (IDIBAPS),
Barcelona, Catalonia, Spain. 2Department of Medicine, Columbia University, New York, New York, USA.

Liver fibrosis is the excessive accumulation of extracellular matrix proteins including collagen
that occurs in most types of chronic liver diseases. Advanced liver fibrosis results in cirrhosis,
liver failure, and portal hypertension and often requires liver transplantation. Our knowledge of
the cellular and molecular mechanisms of liver fibrosis has greatly advanced. Activated hepatic
stellate cells, portal fibroblasts, and myofibroblasts of bone marrow origin have been identi-
fied as major collagen-producing cells in the injured liver. These cells are activated by fibrogenic
cytokines such as TGF-β1, angiotensin II, and leptin. Reversibility of advanced liver fibrosis in
patients has been recently documented, which has stimulated researchers to develop antifibrotic drugs. Emerging
antifibrotic therapies are aimed at inhibiting the accumulation of fibrogenic cells and/or preventing the deposi-
tion of extracellular matrix proteins. Although many therapeutic interventions are effective in experimental mod-
els of liver fibrosis, their efficacy and safety in humans is unknown. This review summarizes recent progress in the
study of the pathogenesis and diagnosis of liver fibrosis and discusses current antifibrotic strategies.

Historical perspective gin have been recently shown to exhibit fibrogenic potential (14, 15).
Liver fibrosis results from chronic damage to the liver in conjunc- At the clinical level, the natural history of liver fibrosis, from early
tion with the accumulation of ECM proteins, which is a charac- changes to liver cirrhosis, was delineated in patients with chronic
teristic of most types of chronic liver diseases (1). The main causes HCV infection (16, 17). Rapid and slower fibrosers were identified,
of liver fibrosis in industrialized countries include chronic HCV and genetic and environmental factors influencing fibrosis progres-
infection, alcohol abuse, and nonalcoholic steatohepatitis (NASH). sion were partially uncovered (18). Since the demonstration, in the
The accumulation of ECM proteins distorts the hepatic architec- 1990s, that even advanced liver fibrosis is reversible, researchers have
ture by forming a fibrous scar, and the subsequent development been stimulated to identify antifibrotic therapies (19). Biotechnol-
of nodules of regenerating hepatocytes defines cirrhosis. Cirrhosis ogy and pharmaceutical companies are increasingly interested in
produces hepatocellular dysfunction and increased intrahepatic developing antifibrotic programs, and clinical trials are currently
resistance to blood flow, which result in hepatic insufficiency and underway. However, the most effective therapy for treating hepatic
portal hypertension, respectively (2). fibrosis to date is still to remove the causative agent (20). A number
Hepatic fibrosis was historically thought to be a passive and of drugs are able to reduce the accumulation of scar tissue in experi-
irreversible process due to the collapse of the hepatic parenchyma mental models of chronic liver injury. Renin-angiotensin system
and its substitution with a collagen-rich tissue (3, 4). Currently, it blockers and antioxidants are the most promising drugs, although
is considered a model of the wound-healing response to chronic their efficacy has not been tested in humans. Lack of clinical trials is
liver injury (5). Early clinical reports in the 1970s suggested that due to the requirement of long follow-up studies and to the fact that
advanced liver fibrosis is potentially reversible (6). However, liver liver biopsy, an invasive procedure, is still the gold-standard method
fibrosis received little attention until the 1980s, when hepatic stel- for detecting changes in liver fibrosis. The current effort to develop
late cells (HSCs), formerly known as lipocytes, Ito cells, or perisinu- noninvasive markers to assess liver fibrosis is expected to facilitate
soidal cells, were identified as the main collagen-producing cells in the design of clinical trials.
the liver (7). This cell type, first described by von Kupffer in 1876, Recently, NASH has been recognized as a major cause of liver
undergoes a dramatic phenotypic activation in chronic liver diseases fibrosis (21). First described by Ludwig et al., it is considered
with the acquisition of fibrogenic properties (8). Methods to obtain part of the spectrum of nonalcoholic fatty liver diseases (22).
HSCs from both rodent and human livers were rapidly standardized These range from steatosis to cirrhosis and can eventually lead to
in the 1980s (9, 10), and prolonged culture of HSCs on plastic was hepatocellular carcinoma. NASH is a component of the metabolic
widely accepted as a model for the study of activated HSCs (11). Key syndrome, which is characterized by obesity, type 2 diabetes mel-
signals that modulate HSCs’ fibrogenic actions were delineated (12). litus, and dyslipidemia, with insulin resistance as a common fea-
Experimental models for studying liver fibrogenesis in rats and in ture. As the prevalence of obesity is rapidly increasing, a rise in the
transgenic mice were developed, which corroborated the cell culture prevalence of NASH is anticipated.
studies and led to the identification of key fibrogenic mediators (13). This review outlines recent progress in the pathogenesis, diagno-
Besides HSCs, portal myofibroblasts and cells of bone marrow ori- sis, and treatment of liver fibrosis, summarizes recent data on the
mechanisms leading to fibrosis resolution, and discusses future
prospects aimed at developing effective antifibrotic therapies.
Nonstandard abbreviations used: CTLA, cytotoxic T lymphocyte antigen; HSC,
hepatic stellate cell; NASH, nonalcoholic steatohepatitis; PBC, primary biliary cirrho-
sis; TIMP-1, tissue inhibitor of metalloproteinase type 1. Natural history and diagnosis
Conflict of interest: The authors have declared that no conflict of interest exists. The onset of liver fibrosis is usually insidious, and most of the
Citation for this article: J. Clin. Invest. 115:209–218 (2005). related morbidity and mortality occur after the development of
doi:10.1172/JCI200524282. cirrhosis (16). In the majority of patients, progression to cirrhosis

The Journal of Clinical Investigation http://www.jci.org Volume 115 Number 2 February 2005 209
science in medicine

Table 1
Genetic and nongenetic factors associated with fibrosis progression in different types of chronic liver diseases

Type of liver disease Candidate genes Candidate genes (full name) Nongenetic factors
Chronic HCV infection HFE Hereditary hemochromatosis gene Alcohol intake
Angiotensinogen Angiotensinogen Coinfection HIV and/or hepatitis B virus
TGF-β1 Transforming growth factor β1 Age at time of acute infection
TNF-α Tumor necrosis factor α Liver transplantation
ApoE Apolipoprotein E Diabetes mellitus
MEH Microsomal epoxide hydroxylase No response to therapy
MCP-1 Monocyte chemotactic protein type 1
MCP-2 Monocyte chemotactic protein type 2
Factor V Factor V (Leiden)
Alcohol-induced IL-10 Interleukin 10 Alcohol intake
IL-1β Interleukin 1β Episodes of alcoholic hepatitis
ADH Alcohol dehydrogenase
ALDH Aldehyde dehydrogenase
CYP2E1 cytochrome P450, family 2, subfamily e, polypeptide 1
TNF-α Tumor necrosis factor α
CTLA-4 Cytotoxic T lymphocyte antigen type 4
TAP2 Transporter-associated antigen-processing type 2
MnSOD Manganese superoxide dismutase
NASH HFE Hereditary hemochromatosis gene Age
Angiotensinogen Angiotensinogen Severity of obesity
TGF-β1 Transforming growth factor β1 Diabetes mellitus
Hypertriglyceridemia
PBC IL-1β Interleukin 1β
TNF-α Tumor necrosis factor α
ApoE Apolipoprotein E
Autoimmune hepatitis HLA-II Human leukocyte antigen type II haplotypes Type II autoimmune hepatitis
No response to therapy

occurs after an interval of 15–20 years. Major clinical complica- an invasive procedure, with pain and major complications occur-
tions of cirrhosis include ascites, renal failure, hepatic encephalop- ring in 40% and 0.5% of patients, respectively (26). Sampling error
athy, and variceal bleeding. Patients with cirrhosis can remain free can occur, especially when small biopsies are analyzed. Histologic
of major complications for several years (compensated cirrhosis). examination is prone to intra- and interobserver variation and
Decompensated cirrhosis is associated with short survival, and does not predict disease progression (27). Therefore, there is a
liver transplantation is often indicated as the only effective therapy need for reliable, simple, and noninvasive methods for assessing
(23). Cirrhosis is also a risk factor for developing hepatocellular liver fibrosis. Scores that include routine laboratory tests, such
carcinoma. Liver fibrosis progresses rapidly to cirrhosis in several as platelet count, aminotransferase serum levels, prothrombin
clinical settings, including repeated episodes of severe acute alco- time, and serum levels of acute phase proteins have been proposed
holic hepatitis, subfulminant hepatitis, and fibrosing cholestasis (28, 29). Serum levels of proteins directly related to the hepatic
in patients with HCV reinfection after liver transplantation (24). fibrogenic process are also used as surrogate markers of liver
The natural history of liver fibrosis is influenced by both genetic fibrosis (30), including N-terminal propeptide of type III colla-
and environmental factors (Table 1). Epidemiological studies have gen, hyaluronic acid, tissue inhibitor of metalloproteinase type 1
identified polymorphisms in a number of candidate genes that (TIMP-1), and YKL-40. Although these scores are useful in detect-
may influence the progression of liver fibrosis in humans (18). ing advanced fibrosis (cirrhosis) in patients, as well as minimal or
These genetic factors may explain the broad spectrum of responses no fibrosis, they are not effective for differentiating intermedi-
to the same etiological agent found in patients with chronic liver ate grades of fibrosis. Also, fibrosis-specific markers may reflect
diseases. However, some studies have yielded contradictory results fibrogenesis in other organs (i.e., pancreatic fibrosis in alcoholic
due to poor study design, and further research is required to clarify patients). Finally, hepatic fibrosis can be estimated by imaging
the actual role of genetic variants in liver fibrosis. techniques. Ultrasonography, computed tomography, and MRI
Liver biopsy is considered the gold-standard method for the can detect changes in the hepatic parenchyma due to moderate
assessment of liver fibrosis (25). Histologic examination is useful to severe fibrosis (31). Due to its low cost, ultrasonography is
in identifying the underlying cause of liver disease and assessing an appealing technique. It is able to detect liver cirrhosis based
the necroinflammatory grade and the stage of fibrosis. Fibrosis on changes in liver echogenicity and nodularity as well as signs
stage is assessed by using scales such as Metavir (stages I–IV) and of portal hypertension. However, ultrasound is highly operator-
Ishak score (stages I–V). Specific staining of ECM proteins (e.g., dependent, and the presence of increased liver echogenicity does
with Sirius red) can be used to quantify the degree of fibrosis, not reliably differentiate hepatic steatosis from fibrosis. Nonin-
using computer-guided morphometric analysis. Liver biopsy is vasive methods currently in development include blood protein

210 The Journal of Clinical Investigation http://www.jci.org Volume 115 Number 2 February 2005
 science in medicine

Figure 1
Changes in the hepatic architecture (A) asso-
ciated with advanced hepatic fibrosis (B). Fol-
lowing chronic liver injury, inflammatory lym-
phocytes infiltrate the hepatic parenchyma.
Some hepatocytes undergo apoptosis, and
Kupffer cells activate, releasing fibrogenic
mediators. HSCs proliferate and undergo a
dramatic phenotypical activation, secreting
large amounts of extracellular matrix proteins.
Sinusoidal endothelial cells lose their fenes-
trations, and the tonic contraction of HSCs
causes increased resistance to blood flow in
the hepatic sinusoid. Figure modified with per-
mission from Science & Medicine (S28).

profiling using proteomic technology and new clinical glycomics accumulate at the sites of tissue repair, secreting large amounts
technology, which is based on DNA sequencer/fragment analyz- of ECM and regulating ECM degradation. PDGF, mainly pro-
ers able to generate profiles of serum protein N-glycans (32). As duced by Kupffer cells, is the predominant mitogen for activated
the technology becomes validated, the noninvasive diagnosis of HSCs. Collagen synthesis in HSCs is regulated at the transcrip-
liver disease may become routine clinical practice. tional and posttranscriptional levels (39). Increased collagen
mRNA stability mediates the increased collagen synthesis in
Pathogenesis of liver fibrosis activated HSCs. In these cells, posttranscriptional regulation
Hepatic fibrosis is the result of the wound-healing response of the of collagen is governed by sequences in the 3′ untranslated
liver to repeated injury (1) (Figure 1). After an acute liver injury region via the RNA-binding protein αCP2 as well as a stem-loop
(e.g., viral hepatitis), parenchymal cells regenerate and replace structure in the 5′ end of collagen mRNA (40). Interestingly,
the necrotic or apoptotic cells. This process is associated with HSCs express a number of neuroendocrine markers (e.g., reelin,
an inflammatory response and a limited deposition of ECM. If nestin, neurotrophins, synaptophysin, and glial-fibrillary acidic
the hepatic injury persists, then eventually the liver regeneration protein) and bear receptors for neurotransmitters (8, 41, 42).
fails, and hepatocytes are substituted with abundant ECM, includ-
ing fibrillar collagen. The distribution of this fibrous material
depends on the origin of the liver injury. In chronic viral hepati-
tis and chronic cholestatic disorders, the fibrotic tissue is initially
located around portal tracts, while in alcohol-induced liver disease,
it locates in pericentral and perisinusoidal areas (33). As fibrotic
liver diseases advance, disease progression from collagen bands to
bridging fibrosis to frank cirrhosis occurs.
Liver fibrosis is associated with major alterations in both the
quantity and composition of ECM (34). In advanced stages, the
liver contains approximately 6 times more ECM than normal,
including collagens (I, III, and IV), fibronectin, undulin, elastin,
laminin, hyaluronan, and proteoglycans. Accumulation of ECM
results from both increased synthesis and decreased degradation Figure 2
(35). Decreased activity of ECM-removing MMPs is mainly due to Expression of collagen α1(I) in a model of cholestasis-induced liver
an overexpression of their specific inhibitors (TIMPs). fibrosis. Transgenic mice with green fluorescence protein reporter
HSCs are the main ECM-producing cells in the injured liver gene under the direction of the collagen α1(I) promoter/enhancers
were subjected to bile duct ligation for 2 weeks. (A) Collagen α1(I)
(36). In the normal liver, HSCs reside in the space of Disse and
was markedly expressed by activated HSCs, but not hepatocytes, in
are the major storage sites of vitamin A. Following chronic inju- the hepatic parenchyma. Magnification, ×200. (B) Collagen α1(I) is
ry, HSCs activate or transdifferentiate into myofibroblast-like markedly expressed by myofibroblasts around proliferating bile ducts.
cells, acquiring contractile, proinflammatory, and fibrogenic HSCs proliferate to initiate collagen deposition in the hepatic paren-
properties (37, 38) (Figure 2A). Activated HSCs migrate and chyma. Magnification, ×40.

The Journal of Clinical Investigation http://www.jci.org Volume 115 Number 2 February 2005 211
science in medicine

212 The Journal of Clinical Investigation http://www.jci.org Volume 115 Number 2 February 2005
 science in medicine

Quiescent HSCs express markers that are characteristic of

adipocytes (PPARγ, SREBP-1c, and leptin), while activated HSCs
express myogenic markers (α smooth muscle actin, c-myb, and
myocyte enhancer factor–2).
Hepatic cell types other than HSCs may also have fibrogenic
potential. Myofibroblasts derived from small portal vessels pro-
liferate around biliary tracts in cholestasis-induced liver fibro-
sis to initiate collagen deposition (43, 44) (Figure 2B). HSCs
and portal myofibroblasts differ in specific cell markers and
response to apoptotic stimuli (45). Culture of CD34 +CD38 –
hematopoietic stem cells with various growth factors has been
shown to generate HSCs and myofibroblasts of bone marrow
origin that infiltrate human livers undergoing tissue remodeling
(15, 46). These data suggest that cells originating in bone mar- Figure 4
row can be a source of fibrogenic cells in the injured liver. Other Reversibility of liver fibrosis in a patient with chronic hepatitis B virus
potential sources of fibrogenic cells (i.e., epithelial-mesenchymal infection after successful treatment with lamivudine. A decrease in
transition and circulating fibrocytes) have not been demonstrat- smooth muscle actin immunostaining, a marker of fibrogenic myofi-
broblasts, can be seen in paired liver biopsies before (A) and after
ed in the liver (47, 48). The relative importance of each cell type
(B) therapy. Dark brown granules represent areas stained for smooth
in liver fibrogenesis may depend on the origin of the liver inju- muscle actin. Magnification, ×40. Reproduced with permission from
ry. While HSCs are the main fibrogenic cell type in pericentral Journal of Hepatology (S2).
areas, portal myofibroblasts may predominate when liver injury
occurs around portal tracts.
A complex interplay among different hepatic cell types takes Genetic studies in rodents and humans
place during hepatic fibrogenesis (Figure 3) (49). Hepatocytes are Extensive studies using models of hepatic fibrosis in transgenic
targets for most hepatotoxic agents, including hepatitis viruses, mice have revealed key genes mediating liver fibrogenesis (1, 18).
alcohol metabolites, and bile acids (50). Damaged hepatocytes Genes regulating hepatocellular apoptosis and/or necrosis (e.g.,
release ROS and fibrogenic mediators and induce the recruit- Bcl-xL, Fas) influence the extent of hepatic damage and the subse-
ment of white blood cells by inflammatory cells. Apoptosis of quent fibrogenic response (60, 61). Genes regulating the inflam-
damaged hepatocytes stimulates the fibrogenic actions of liver matory response to injury (e.g., IL-1β, IL-6, IL-10, and IL-13, IFN-γ,
myofibroblasts (51). Inflammatory cells, either lymphocytes or SOCS-1, and osteopontin) determine the fibrogenic response to
polymorphonuclear cells, activate HSCs to secrete collagen (52). injury (55, 62–65). Genes mediating ROS generation (e.g., NADPH
Activated HSCs secrete inflammatory chemokines, express cell oxidase) regulate both inflammation and ECM deposition (66).
adhesion molecules, and modulate the activation of lymphocytes Fibrogenic growth factors (e.g., TGF-β1, FGF), vasoactive sub-
(53). Therefore, a vicious circle in which inflammatory and fibro- stances (angiotensin II, norepinephrine), and adipokines (leptin
genic cells stimulate each other is likely to occur (54). Fibrosis is and adiponectin) are each required for the development of fibrosis
influenced by different T helper subsets, the Th2 response being (67–70). Finally, removal of excess collagen after cessation of liver
associated with more active fibrogenesis (55). Kupffer cells are injury is regulated by TIMP-1 and TGF-β1 (71, 72).
resident macrophages that play a major role in liver inflammation Association genetic studies have investigated the role of gene
by releasing ROS and cytokines (56, 57). In chronic cholestatic polymorphisms in the progression of liver fibrosis in patients
disorders (i.e., primary biliary cirrhosis [PBC] and primary sclero- with chronic liver diseases (18). In alcoholic liver disease, candidate
sis cholangitis), epithelial cells stimulate the accumulated portal genes include genes encoding for alcohol-metabolizing enzymes
myofibroblasts to initiate collagen deposition around damaged and proteins involved in liver toxicity (73). Polymorphisms in
bile ducts (43). Finally, changes in the composition of the ECM can genes encoding alcohol-dehydrogenase, aldehyde-dehydrogenase,
directly stimulate fibrogenesis. Type IV collagen, fibrinogen, and and cytochrome P450 are involved in individual susceptibility
urokinase type plasminogen activator stimulate resident HSCs by to alcoholism, yet their role in the progression of liver disease
activating latent cytokines such as TGF-β1 (58). Fibrillar collagens remains controversial. Variations in genes encoding inflamma-
can bind and stimulate HSCs via discoidin domain receptor DDR2 tory mediators (e.g., TNF-α, IL-1β, Il-10, and cytotoxic T lympho-
and integrins. Moreover, the altered ECM can serve as a reservoir cyte antigen–4 [CTLA-4]), the lipopolysaccharide receptor CD14,
for growth factors and MMPs (59). and antioxidants (e.g., superoxide dismutase) may influence the

Figure 3
Cellular mechanisms of liver fibrosis. Different types of hepatotoxic agents produce mediators that induce inflammatory actions in hepatic cell
types. Damaged hepatocytes and biliary cells release inflammatory cytokines and soluble factors that activate Kupffer cells and stimulate the
recruitment of activated T cells. This inflammatory milieu stimulates the activation of resident HSCs into fibrogenic myofibroblasts. Activated
HSCs also secrete cytokines that perpetuate their activated state. If the liver injury persists, accumulation of activated HSCs and portal myofibro-
blasts occurs, synthesizing large amounts of ECM proteins and leading to tissue fibrosis. ECM degradation is inhibited by the actions of cytokines
such as TIMPs. Apoptosis of damaged hepatocytes stimulates the fibrogenic actions of HSCs. If the cause of the liver injury is removed, fibrosis
is resolved. This phase includes apoptosis of activated HSCs and regeneration of hepatocytes. Collagen is degraded by increased activity of
MMPs induced by decreased TIMP expression. CCL21, C-C chemokine ligand 21; MCP-1, monocyte chemoattractant protein–1; MIP-2, macro-
phage inflammatory protein–2; NS3, HCV nonstructural protein 3; NS5, HCV nonstructural protein 5; PAF, platelet-activating factor.

The Journal of Clinical Investigation http://www.jci.org Volume 115 Number 2 February 2005 213
science in medicine

progression of alcohol-induced liver disease (74, 75). In chronic angiotensin II) have opposite effects (67, 89). Endothelin-1, a pow-
cholestatic disorders such as PBC, polymorphisms in IL-1β, IL-1 erful vasoconstrictor, stimulates fibrogenesis through its type A
receptor antagonists, and TNF-α genes are associated with faster receptor (90). Among vasoactive cytokines, angiotensin II seems to
disease progression (76). Some alleles of the apolipoprotein E gene play a major role in liver fibrogenesis. Angiotensin II is the effector
influence the response to therapy of PBC with ursodeoxycholic peptide of the renin-angiotensin system, which is a major regulator
acid, which suggests that genetic polymorphisms may predict of arterial pressure homeostasis in humans. Key components of this
therapeutic response (77). In HCV liver disease, genetic varia- system are locally expressed in chronically injured livers, and acti-
tions are involved in susceptibility to persistent HCV infection, vated HSCs de novo generate angiotensin II (91, 92). Importantly,
response to antiviral therapy, and progression of liver disease (78). pharmacological and/or genetic ablation of the renin-angiotensin
Polymorphisms in genes involved in the immune response to HCV system markedly attenuates experimental liver fibrosis (70, 93–98).
infection (e.g., transporter associated with antigen processing 2, Angiotensin II induces hepatic inflammation and stimulates an
mannose-binding lectin, and specific HLA-II alleles) and fibro- array of fibrogenic actions in activated HSCs, including cell prolif-
genic agonists (angiotensinogen and TGF-β1) influence fibrosis eration, cell migration, secretion of proinflammatory cytokines, and
progression (79–81). The fibrogenic effect of heterozygosity in collagen synthesis (66, 99, 100). These actions are largely mediated
the C282Y mutation of the hemochromatosis gene in patients by ROS generated by a nonphagocytic form of NADPH oxidase.
with chronic hepatitis C is controversial (82, 83). Finally, little is Unlike the phagocytic type, NADPH oxidases present in fibrogenic
known about genetic factors and NASH (84), and polymorphisms cell types are constitutively active, producing relatively low levels of
in fibrogenic mediators such as angiotensinogen and TGF-β1 may ROS under basal conditions and generating higher levels of oxidants
be associated with more severe liver disease. in response to cytokines, stimulating redox-sensitive intracellular
pathways. NADPH oxidase also plays a key role in the inflamma-
Key cytokines involved in liver fibrosis tory actions of Kupffer cells (101). Disruption of an active NADPH
Cytokines regulating the inflammatory response to injury mod- oxidase protects mice from developing severe liver injury following
ulate hepatic fibrogenesis in vivo and in vitro (85). Monocyte prolonged alcohol intake and/or bile duct ligation (66, 102).
chemotactic protein type 1 and RANTES stimulate fibrogenesis Adipokines, which are cytokines mainly derived from the adipose
while IL-10 and IFN-γ exert the opposite effect (55, 86). Among tissue, regulate liver fibrogenesis. Leptin is required for HSC activa-
growth factors, TGF-β1 appears to be a key mediator in human tion and fibrosis development (103, 104). In contrast, adiponectin
fibrogenesis (58). In HSCs, TGF-β favors the transition to myofi- markedly inhibits liver fibrogenesis in vitro and in vivo (69). The
broblast-like cells, stimulates the synthesis of ECM proteins, and actions of these cytokines may explain why obesity influences fibro-
inhibits their degradation. Strategies aimed at disrupting TGF-β1 sis development in patients with chronic hepatitis C (105).
synthesis and/or signaling pathways markedly decreased fibrosis
in experimental models (87). PDGF is the most potent mitogen Intracellular signaling pathways
for HSCs and is upregulated in the fibrotic liver (12); its inhibition mediating liver fibrogenesis
attenuates experimental liver fibrogenesis (88). Data on intracellular pathways regulating liver fibrogenesis are
Cytokines with vasoactive properties also regulate liver fibro- mainly derived from studies using cultured HSCs, while under-
genesis. Vasodilator substances (e.g., nitric oxide, relaxin) exert standing of their role in vivo is progressing through experimental
antifibrotic effects while vasoconstrictors (e.g., norepinephrine, fibrogenesis studies using knockout mice (106). Several mitogen-

Table 2
Main antifibrotic drugs in development for the treatment of liver fibrosis

Agent Main mechanism Antifibrotic effects Antifibrotic effects in Antifibrotic effect

in HSCs experimental fibrosis in humans
Angiotensin inhibitors Inhibits HSC activation Consistent positive data Consistent positive data Retrospective study
Colchicine Inhibits inflammatory response Limited data Limited data Discrepant results
Corticosteroids Inhibits inflammatory response Limited data Limited data Effective in
autoimmune hepatitis
Endothelin inhibitors Inhibits HSC function Limited data Limited data Not tested
Interferon-α Inhibits HSC activation Consistent positive data Consistent positive data Effective in chronic hepatitis C
Interleukin 10 Inhibits inflammatory response Limited data Consistent positive data Isolated reports
in chronic hepatitis C
Pentoxifylline Inhibits HSC activation Consistent positive data Consistent positive data Not tested
Phosphatidylcholine Decreases oxidative stress Limited data Consistent positive data Not proven in
alcohol-induced fibrosis
PPAR antagonists Inhibits HSC activation Consistent data Consistent positive data Isolated reports in NASH
S-adenosyl-methionine Antioxidant Limited data Not tested Effective in
alcohol-induced fibrosis
Sho-saiko-to Antioxidant Consistent positive data Consistent positive data Isolated reports in
chronic hepatitis C
TGF-β1 inhibitors Inhibits HSC activation and function Consistent positive data Consistent positive data Not tested
Tocopherol Antioxidant Consistent positive data Limited data Isolated reports in NASH

214 The Journal of Clinical Investigation http://www.jci.org Volume 115 Number 2 February 2005
 science in medicine

activated protein kinases modulate major fibrogenic actions of alcohol-induced liver injury, chronic hepatitis C, B, and D, hemo-
HSCs. Extracellular-regulated kinase, which is stimulated in exper- chromatosis, secondary biliary cirrhosis, NASH, and autoimmune
imentally induced liver injury, mediates proliferation and migra- hepatitis (19, 122, 124, 125, S1, S2) (Figure 4). It may take years for
tion of HSCs (107). In contrast, c-Jun N-terminal kinase regulates significant regression to be achieved; the time varies depending on
apoptosis of hepatocytes as well as the secretion of inflammatory the underlying cause of the liver disease and its severity. Chronic
cytokines by cultured HSCs (66, 108, 109). The focal adhesion HCV infection is the most extensively studied condition, and ther-
kinase PI3K-Akt–signaling pathway mediates agonist-induced apy (IFN-α plus ribavirin) with viral clearance results in fibrosis
fibrogenic actions in HSCs (107). The TGF-β1–activated Smad- improvement. Importantly, nearly half of patients with cirrhosis
signaling pathway stimulates experimental hepatic fibrosis and exhibit reversal to a significant degree (90). Whether this beneficial
is a potential target for therapy (110, 111). The PPAR pathway effect is associated with improvements in long-term clinical out-
regulates HSC activation and experimental liver fibrosis. PPAR-γ come, including decreased portal hypertension, is unknown.
ligands inhibit the fibrogenic actions in HSCs and attenuate liver Increased collagenolytic activity is a major mechanism of fibro-
fibrosis in vivo (112, 113). NF-κB may have an inhibitory action on sis resolution (122). Fibrillar collagens (I and III) are degraded by
liver fibrosis (114, 115). Other transcription factors are involved interstitial MMPs (MMP-1, -8, and -13 in humans and MMP-13 in
in HSC activation and may participate in liver fibrogenesis (116). rodents). During fibrosis resolution, MMP activity increases due to
Recent studies suggest a role for intracellular pathways signaled by a rapid decrease in the expression of TIMP-1. Partial degradation
Toll-like receptors and β-cathepsin (117, 118). of fibrillar collagen occurs, and the altered interaction between
activated HSCs and ECM favors apoptosis (123). Removal of acti-
Pathogenesis of fibrosis in different liver diseases vated HSCs by apoptosis precedes fibrosis resolution. Stimulation
The pathogenesis of liver fibrosis depends on the underlying eti- of death receptors in activated HSCs and a decrease in survival fac-
ology. In alcohol-induced liver disease, alcohol alters the popu- tors, including TIMP-1, can precipitate HSC apoptosis (S3).
lation of gut bacteria and inhibits intestinal motility, resulting Several questions remain unanswered: Can we pharmacologi-
in an overgrowth of Gram-negative flora. Lipopolysaccharide is cally accelerate fibrosis resolution in humans? Can a fibrotic
elevated in portal blood and activates Kupffer cells through the liver completely regress to a normal liver? Does fibrosis reverse
CD14/Toll-like receptor–4 complex to produce ROS via NADPH similarly in all types of liver diseases? Although isolated cases of
oxidase (101). Oxidants activate Kupffer cell NF-κB, causing an complete fibrosis resolution have been reported, it is conceivable
increase in TNF-α production. TNF-α induces neutrophil infil- that some degree of fibrosis cannot be removed (S4). Resolution
tration and stimulates mitochondrial oxidant production in may be limited by ECM cross-linking and a failure of activated
hepatocytes, which are sensitized to undergo apoptosis. Acetal- HSCs to undergo apoptosis.
dehyde, the major alcohol metabolism product, and ROS acti-
vate HSCs and stimulate inflammatory and fibrogenic signals Therapeutic approaches to the treatment of liver fibrosis
(119). The pathogenesis of HCV-induced liver fibrosis is poorly There is no standard treatment for liver fibrosis. Although experi-
understood due to the lack of a rodent model of persistent HCV mental studies have revealed targets to prevent fibrosis progres-
infection (78). HCV escapes surveillance of the HLA-II–directed sion in rodents (20) (Table 2), the efficacy of most treatments has
immune response and infects hepatocytes, causing oxidative not been proven in humans. This is due to the need to perform
stress and inducing the recruitment of inflammatory cells. Both serial liver biopsies to accurately assess changes in liver fibrosis, the
factors lead to HSC activation and collagen deposition. Moreover, necessity of long-term follow-up studies, and the fact that humans
several HCV proteins directly stimulate the inflammatory and are probably less sensitive to hepatic antifibrotic therapies than
fibrogenic actions of HSCs (120). In chronic cholestatic disorders rodents. The development of reliable noninvasive markers of liver
such as PBC, T lymphocytes and cytokines mediate persistent bile fibrosis should have a positive impact on the design of clinical tri-
duct damage (14). Biliary cells secrete fibrogenic mediators acti- als. The ideal antifibrotic therapy would be one that is liver-specific,
vating neighboring portal myofibroblasts to secrete ECM. Even- well tolerated when administered for prolonged periods of time,
tually, perisinusoidal HSCs become activated, and fibrotic bands and effective in attenuating excessive collagen deposition without
develop. The pathogenesis of liver fibrosis due to NASH is poorly affecting normal ECM synthesis.
understood. Obesity, type 2 diabetes mellitus, and dyslipidemia The removal of the causative agent is the most effective interven-
are the most common associated conditions (121). A 2-hit model tion in the treatment of liver fibrosis. This strategy has been shown
has been proposed: hyperglycemia and insulin resistance lead effective in most etiologies of chronic liver diseases (19, 122, 124,
to elevated serum levels of free fatty acids, resulting in hepatic 125, S1, S2). For patients with cirrhosis and clinical complications,
steatosis. In the second hit, oxidative stress and proinflammatory liver transplantation is currently the only curative approach (S5).
cytokines promote hepatocyte apoptosis and the recruitment of Transplantation improves both survival and quality of life. How-
inflammatory cells, leading to progressive fibrosis. ever, in patients with HCV-induced cirrhosis, viral infection recurs
after transplantation (S6), aggressive chronic hepatitis develops,
Is liver fibrosis reversible? and progression to cirrhosis is common.
In contrast with the traditional view that cirrhosis is an irrevers- Because inflammation precedes and promotes the progression
ible disease, recent evidence indicates that even advanced fibrosis of liver fibrosis, the use of antiinflammatory drugs has been
is reversible (122). In experimentally induced fibrosis, cessation proposed. Corticosteroids are only indicated for the treatment
of liver injury results in fibrosis regression (123). In humans, of hepatic fibrosis in patients with autoimmune hepatitis and
spontaneous resolution of liver fibrosis can occur after suc- acute alcoholic hepatitis (S1). Inhibition of the accumulation
cessful treatment of the underlying disease. This observation of activated HSCs by modulating either their activation and/or
has been described in patients with iron and copper overload, proliferation or promoting their apoptosis is another strategy.

The Journal of Clinical Investigation http://www.jci.org Volume 115 Number 2 February 2005 215
science in medicine

Antioxidants such as vitamin E, silymarin, phosphatidylcholine, patients (S21). Sustained virological response is associated with
and S-adenosyl-L-methionine inhibit HSC activation, protect an improvement in liver fibrosis (122). Patients with no sus-
hepatocytes from undergoing apoptosis, and attenuate experi- tained response may also experience improvement of liver fibro-
mental liver fibrosis (S7). Antioxidants exert beneficial effects in sis, which suggests that IFN-α has an intrinsic antifibrotic effect
patients with alcohol-induced liver disease and NASH (S8, S9). (S22). For nonresponder patients, the use of renin-angiotensin
Disrupting TGF-β synthesis and/or signaling pathways pre- system inhibitors is a promising approach. Treatment of the
vents scar formation in experimental liver fibrosis (58). More- metabolic syndrome in patients with chronic hepatitis C may
over, administration of growth factors (e.g., IGF, hepatocyte also decrease fibrosis progression (S23). In patients with alco-
growth factor, and cardiotrophin) or their delivery by gene ther- hol-induced liver disease, the most effective approach is alcohol
apy attenuates experimental liver fibrosis (S10, S11). However, abstinence (124). Antioxidants (e.g., S-adenosyl-L-methionine and
these latter approaches have not been tested in humans and may phosphatidylcholine) and hepatocyte protectors (e.g., silymarin)
favor cancer development. Substances that inhibit key signal slow down the progression of liver fibrosis and can improve sur-
transduction pathways involved in liver fibrogenesis also have the vival (S24). For patients with autoimmune hepatitis, immuno-
potential to treat liver fibrosis (20). They include pentoxifylline suppressant therapy not only decreases inflammation but also
(phosphodiesterase inhibitor), amiloride (Na+/H+ pump inhibi- exerts antifibrotic effects (S25). No antifibrotic therapy is avail-
tor), and S-farnesylthiosalicylic acid (Ras antagonist). Ligands able for patients with chronic cholestatic disorders (i.e., primary
of PPARα and/or PPARγ such as thiazolindiones exert benefi- sclerosing cholangitis and PBC). Ursodeoxycholic acid improves
cial effects in experimental liver fibrosis and in patients with biochemical tests in these patients, but its impact on fibrosis is
NASH (S12, S13). The inhibition of the renin-angiotensin sys- not consistently proven (S26). In patients with NASH, weight loss
tem is probably the most promising strategy in treating liver and specific treatments of the metabolic syndrome can reduce
fibrosis. Renin-angiotensin inhibitors are widely used as antifi- fibrosis development (125). Recent reports have revealed than
brotic agents in patients with chronic renal and cardiac diseases antioxidants and insulin sensitizers (e.g., thiazolindiones) may
and appear to be safe when administered for prolonged periods exert antifibrogenic effects in these patients (S27). Large clinical
of time (S14). Little information is available on the use of this trials are needed to confirm these results.
approach in patients with chronic liver diseases. Preliminary
pilot studies in patients with chronic hepatitis C and NASH Future directions
suggest that renin-angiotensin blocking agents may have benefi- The translation of basic research into improved therapeutics for
cial effects on fibrosis progression (S15). Transplanted patients the management of patients with chronic liver diseases is still poor.
receiving renin-angiotensin system inhibitors as antihyperten- The role of pluripotential stem cells in hepatic wound healing is
sive therapy show less fibrosis progression than patients receiv- one of the most promising fields. Perfusion of these cells may be
ing other types of drugs (S16). However, this approach cannot a potential approach to promoting fibrosis resolution and liver
be recommended in clinical practice until the results of ongoing regeneration. Approaches to removing fibrogenic cells are being
clinical trials become available. The blockade of endothelin-1 evaluated, including development of drug delivery systems that
type A receptors and the administration of vasodilators (prosta- target activated HSCs. Translational research should investigate
glandin E2 and nitric oxide donors) exert antifibrotic activity the molecular mechanisms that cause fibrosis in different types of
in rodents, yet the effects in humans are unknown (90). Dif- human liver diseases in order to identify new targets for therapy.
ferent herbal compounds, many of them traditionally used in In the clinical setting, the identity of the genetic determinants
Asian countries to treat liver diseases, have been demonstrat- that influence fibrosis progression should be uncovered. Well-
ed to have antifibrotic effects (S17). They include Sho-saiko-to, designed large-scale epidemiological genetic studies are clearly
glycyrrhizin, and savia miltiorhiza. An alternative approach is required. Patients at a high risk of progression to cirrhosis should
the inhibition of collagen production and/or the promotion be identified. Developing simple and reliable noninvasive mark-
of its degradation (20). Inhibitors of prolyl-4 hydroxylase and ers of hepatic fibrosis is an important goal in clinical hepatology
halofuginone prevent the development of experimental liver cir- and will facilitate the design of clinical trials. Most importantly,
rhosis by inhibiting collagen synthesis. MMP-8 and urokinase- the efficacy of antifibrotic drugs known to attenuate experimental
type plasminogen activator stimulate collagen degradation in liver fibrosis should be tested in humans.
vivo. The efficacy of these drugs in humans is unknown, and
they may result in undesirable side effects. Finally, infusion of Acknowledgments
mesenchymal stem cells ameliorates experimentally induced The authors’ work is supported by grants from the NIH, the Min-
fibrosis, which suggests a potential for this approach in the isterio de Ciencia y Tecnología de España, and the Instituto de
treatment of chronic liver diseases (S18, S19). Investigación Carlos III (SAF2002-03696 and BFI2002-01202).
A limitation of the current antifibrotic approaches is that anti-
fibrotic drugs are not efficiently taken up by activated HSCs and Due to space constraints, a number of important references could
may produce unwanted side effects. Cell-specific delivery to HSCs not be included in this article. References S1–S27 are available
could provide a solution to these problems. Promising prelimi- online with this article; doi:10.1172/JCI200524282DS1.
nary results have been recently obtained using different carriers
(e.g., cyclic peptides coupled to albumin recognizing collagen Address correspondence to: David A. Brenner, Department of
type VI receptor and/or PDGFR) (S20). Antifibrotic therapy may Medicine, Columbia University Medical Center, College of Physi-
differ depending on the type of liver disease. In patients with cians and Surgeons, 622 West 168th Street, PH 8E-105J, New York,
chronic HCV infection, current antiviral treatments (pegylated New York 10032, USA. Phone: (212) 305-5838; Fax: (212) 305-8466;
IFN plus ribavirin) clear viral infection in more than half of the E-mail: [email protected].

216 The Journal of Clinical Investigation http://www.jci.org Volume 115 Number 2 February 2005
 science in medicine

1. Friedman, S.L. 2003. Liver fibrosis - from bench to C after liver retransplantation for hepatitis C virus- stellate cells. J. Hepatol. 40:261–268.
bedside. J. Hepatol. 38(Suppl. 1):S38–S53. related graft cirrhosis. Liver Transpl. 9:228–235. 47. Phillips, R.J., et al. 2004. Circulating fibrocytes traf-
2. Gines, P., Cardenas, A., Arroyo, V., and Rodes, J. 25. Afdhal, N.H., and Nunes, D. 2004. Evaluation of fic to the lungs in response to CXCL12 and mediate
2004. Management of cirrhosis and ascites. N. Engl. liver fibrosis: a concise review. Am. J. Gastroenterol. fibrosis. J. Clin. Invest. 114:438–446. doi:10.1172/
J. Med. 350:1646–1654. 99:1160–1174. JCI200420997.
3. Popper, H., and Uenfriend, S. 1970. Hepatic fibro- 26. Thampanitchawong, P., and Piratvisuth, T. 1999. 48. Kalluri, R., and Neilson, E.G. 2003. Epithelial-
sis. Correlation of biochemical and morphologic Liver biopsy: complications and risk factors. World mesenchymal transition and its implications for
investigations. Am. J. Med. 49:707–721. J. Gastroenterol. 5:301–304. fibrosis. J. Clin. Invest. 112:1776–1784. doi:10.1172/
4. Schaffner, F., and Klion, F.M. 1968. Chronic hepa- 27. Regev, A., et al. 2002. Sampling error and intra- JCI200320530.
titis. Annu. Rev. Med. 19:25–38. observer variation in liver biopsy in patients 49. Kmiec, Z. 2001. Cooperation of liver cells in
5. Albanis, E., and Friedman, S.L. 2001. Hepatic fibro- with chronic HCV infection. Am. J. Gastroenterol. health and disease. Adv. Anat. Embryol. Cell Biol.
sis. Pathogenesis and principles of therapy. Clin. 97:2614–2618. 161:III–XIII, 1–151.
Liver Dis. 5:315–334, v–vi. 28. Imbert-Bismut, F., et al. 2001. Biochemical 50. Higuchi, H., and Gores, G.J. 2003. Mechanisms of
6. Soyer, M.T., Ceballos, R., and Aldrete, J.S. 1976. markers of liver fibrosis in patients with hepati- liver injury: an overview. Curr. Mol. Med. 3:483–490.
Reversibility of severe hepatic damage caused by tis C virus infection: a prospective study. Lancet. 51. Canbay, A., Friedman, S., and Gores, G.J. 2004.
jejunoileal bypass after re-establishment of normal 357:1069–1075. Apoptosis: the nexus of liver injury and fibrosis.
intestinal continuity. Surgery. 79:601–604. 29. Forns, X., et al. 2002. Identification of chronic Hepatology. 39:273–278.
7. Friedman, S.L., Roll, F.J., Boyles, J., and Bissell, hepatitis C patients without hepatic fibrosis by a 52. Casini, A., et al. 1997. Neutrophil-derived
D.M. 1985. Hepatic lipocytes: the principal colla- simple predictive model. Hepatology. 36:986–992. superoxide anion induces lipid peroxidation and
gen-producing cells of normal rat liver. Proc. Natl. 30. Fontana, R.J., and Lok, A.S. 2002. Noninvasive stimulates collagen synthesis in human hepatic
Acad. Sci. U. S. A. 82:8681–8685. monitoring of patients with chronic hepatitis C. stellate cells: role of nitric oxide. Hepatology.
8. Geerts, A. 2001. History, heterogeneity, develop- Hepatology. 36:S57–S64. 25:361–367.
mental biology, and functions of quiescent hepatic 31. Hirata, M., Akbar, S.M., Horiike, N., and Onji, 53. Vinas, O., et al. 2003. Human hepatic stellate cells
stellate cells. Semin. Liver Dis. 21:311–335. M. 2001. Noninvasive diagnosis of the degree of show features of antigen-presenting cells and
9. Friedman, S.L., et al. 1992. Isolated hepatic lipo- hepatic fibrosis using ultrasonography in patients stimulate lymphocyte proliferation. Hepatology.
cytes and Kupffer cells from normal human liver: with chronic liver disease due to hepatitis C virus. 38:919–929.
morphological and functional characteristics in Eur. J. Clin. Invest. 31:528–535. 54. Maher, J.J. 2001. Interactions between hepatic stel-
primary culture. Hepatology. 15:234–243. 32. Callewaert, N., et al. 2004. Noninvasive diagnosis late cells and the immune system. Semin. Liver Dis.
10. Otto, D.A., and Veech, R.L. 1980. Isolation of a of liver cirrhosis using DNA sequencer-based total 21:417–426.
lipocyte-rich fraction from rat liver nonparenchy- serum protein glycomics. Nat. Med. 10:429–434. 55. Shi, Z., Wakil, A.E., and Rockey, D.C. 1997. Strain-
mal cells. Adv. Exp. Med. Biol. 132:509–517. 33. Pinzani, M. 1999. Liver fibrosis. Springer Semin. specific differences in mouse hepatic wound healing
11. Rockey, D.C., Boyles, J.K., Gabbiani, G., and Fried- Immunopathol. 21:475–490. are mediated by divergent T helper cytokine respons-
man, S.L. 1992. Rat hepatic lipocytes express 34. Benyon, R.C., and Iredale, J.P. 2000. Is liver fibrosis es. Proc. Natl. Acad. Sci. U. S. A. 94:10663–10668.
smooth muscle actin upon activation in vivo and reversible? Gut. 46:443–446. 56. Naito, M., Hasegawa, G., Ebe, Y., and Yamamoto,
in culture. J. Submicrosc. Cytol. Pathol. 24:193–203. 35. Arthur, M.J. 2000. Fibrogenesis II. Metalloproteinases T. 2004. Differentiation and function of Kupffer
12. Pinzani, M., Gesualdo, L., Sabbah, G.M., and and their inhibitors in liver fibrosis. Am. J. Physiol. cells. Med. Electron Microsc. 37:16–28.
Abboud, H.E. 1989. Effects of platelet-derived Gastrointest. Liver Physiol. 279:G245–G249. 57. Thurman, R.G. 1998. Alcoholic liver injury involves
growth factor and other polypeptide mitogens on 36. Gabele, E., Brenner, D.A., and Rippe, R.A. 2003. activation of Kupffer cells by endotoxin. Am. J.
DNA synthesis and growth of cultured rat liver fat- Liver fibrosis: signals leading to the amplification Physiol. 275:G605–G611.
storing cells. J. Clin. Invest. 84:1786–1793. of the fibrogenic hepatic stellate cell. Front. Biosci. 58. Gressner, A.M., Weiskirchen, R., Breitkopf, K.,
13. Wasser, S., and Tan, C.E. 1999. Experimental mod- 8:D69–D77. and Dooley, S. 2002. Roles of TGF-beta in hepatic
els of hepatic fibrosis in the rat. Ann. Acad. Med. 37. Milani, S., et al. 1990. Procollagen expression by fibrosis. Front. Biosci. 7:d793–d807.
Singapore. 28:109–111. nonparenchymal rat liver cells in experimental 59. Olaso, E., et al. 2001. DDR2 receptor promotes
14. Ramadori, G., and Saile, B. 2004. Portal tract fibro- biliary fibrosis. Gastroenterology. 98:175–184. MMP-2–mediated proliferation and invasion by
genesis in the liver. Lab. Invest. 84:153–159. 38. Marra, F. 1999. Hepatic stellate cells and the regula- hepatic stellate cells. J. Clin. Invest. 108:1369–1378.
15. Forbes, S.J., et al. 2004. A significant proportion of tion of liver inflammation. J. Hepatol. 31:1120–1130. doi:10.1172/JCI200112373.
myofibroblasts are of bone marrow origin in human 39. Lindquist, J.N., Marzluff, W.F., and Stefanovic, B. 60. Takehara, T., et al. 2004. Hepatocyte-specific dis-
liver fibrosis. Gastroenterology. 126:955–963. 2000. Fibrogenesis. III. Posttranscriptional regula- ruption of Bcl-xL leads to continuous hepatocyte
16. Poynard, T., et al. 2000. Natural history of HCV tion of type I collagen. Am. J. Physiol. Gastrointest. apoptosis and liver fibrotic responses. Gastro-
infection. Baillieres Best Pract. Res. Clin. Gastroenterol. Liver Physiol. 279:G471–G476. enterology. 127:1189–1197.
14:211–228. 40. Lindquist, J.N., Parsons, C.J., Stefanovic, B., and 61. Canbay, A., et al. 2002. Fas enhances fibrogenesis in
17. Poynard, T., Bedossa, P., and Opolon, P. 1997. Nat- Brenner, D.A. 2004. Regulation of alpha1(I) col- the bile duct ligated mouse: a link between apoptosis
ural history of liver fibrosis progression in patients lagen messenger RNA decay by interactions with and fibrosis. Gastroenterology. 123:1323–1330.
with chronic hepatitis C. The OBSVIRC, META- alphaCP at the 3′-untranslated region. J. Biol. Chem. 62. Safadi, R., et al. 2004. Immune stimulation of
VIR, CLINIVIR, and DOSVIRC groups. Lancet. 279:23822–23829. hepatic fibrogenesis by CD8 cells and attenuation
349:825–832. 41. Sato, M., Suzuki, S., and Senoo, H. 2003. Hepatic by transgenic interleukin-10 from hepatocytes.
18. Bataller, R., North, K.E., and Brenner, D.A. 2003. stellate cells: unique characteristics in cell biology Gastroenterology. 127:870–882.
Genetic polymorphisms and the progression and phenotype. Cell Struct. Funct. 28:105–112. 63. Sahai, A., Malladi, P., Melin-Aldana, H., Green,
of liver fibrosis: a critical appraisal. Hepatology. 42. Oben, J.A., Yang, S., Lin, H., Ono, M., and Diehl, R.M., and Whitington, P.F. 2004. Upregulation of
37:493–503. A.M. 2003. Norepinephrine and neuropeptide Y osteopontin expression is involved in the develop-
19. Hammel, P., et al. 2001. Regression of liver fibro- promote proliferation and collagen gene expres- ment of nonalcoholic steatohepatitis in a dietary
sis after biliary drainage in patients with chronic sion of hepatic myofibroblastic stellate cells. murine model. Am. J. Physiol. Gastrointest. Liver
pancreatitis and stenosis of the common bile duct. Biochem. Biophys. Res. Commun. 302:685–690. Physiol. 287:G264–G273.
N. Engl. J. Med. 344:418–423. 43. Kinnman, N., and Housset, C. 2002. Peribiliary 64. Yoshida, T., et al. 2004. SOCS1 is a suppressor of
20. Bataller, R., and Brenner, D.A. 2001. Hepatic stel- myofibroblasts in biliary type liver fibrosis. Front. liver fibrosis and hepatitis-induced carcinogenesis.
late cells as a target for the treatment of liver fibro- Biosci. 7:d496–d503. J. Exp. Med. 199:1701–1707.
sis. Semin. Liver Dis. 21:437–451. 44. Magness, S.T., Bataller, R., Yang, L., and Brenner, 65. Streetz, K.L., et al. 2003. Interleukin 6/gp130-
21. Brunt, E.M. 2004. Nonalcoholic steatohepatitis. D.A. 2004. A dual reporter gene transgenic mouse dependent pathways are protective during chronic
Semin. Liver Dis. 24:3–20. demonstrates heterognity in hepatic fibrogenic cell liver diseases. Hepatology. 38:218–229.
22. Ludwig, J., Viggiano, T.R., McGill, D.B., and Oh, B.J. populations. Hepatology. 40:1151–1159. 66. Bataller, R., et al. 2003. NADPH oxidase signal
1980. Nonalcoholic steatohepatitis: Mayo Clinic 45. Knittel, T., et al. 1999. Rat liver myofibroblasts and transduces angiotensin II in hepatic stellate cells
experiences with a hitherto unnamed disease. Mayo hepatic stellate cells: different cell populations of and is critical in hepatic fibrosis. J. Clin. Invest.
Clin. Proc. 55:434–438. the fibroblast lineage with fibrogenic potential. 112:1383–1394. doi:10.1172/JCI200318212.
23. Davis, G.L., Albright, J.E., Cook, S.F., and Rosen- Gastroenterology. 117:1205–1221. 67. Oben, J.A., et al. 2004. Hepatic fibrogenesis requires
berg, D.M. 2003. Projecting future complications 46. Suskind, D.L., and Muench, M.O. 2004. Search- sympathetic neurotransmitters. Gut. 53:438–445.
of chronic hepatitis C in the United States. Liver ing for common stem cells of the hepatic and 68. Yu, C., et al. 2003. Role of fibroblast growth fac-
Transpl. 9:331–338. hematopoietic systems in the human fetal liver: tor type 1 and 2 in carbon tetrachloride-induced
24. Berenguer, M., et al. 2003. Severe recurrent hepatitis CD34+ cytokeratin 7/8+ cells express markers for hepatic injury and fibrogenesis. Am. J. Pathol.

The Journal of Clinical Investigation http://www.jci.org Volume 115 Number 2 February 2005 217
science in medicine

163:1653–1662. against PDGF B-chain proves effective in prevent- Fas-mediated apoptosis in hepatocytes. FASEB J.
69. Kamada, Y., et al. 2003. Enhanced carbon tetra- ing experimental liver fibrogenesis. Biochem. Biophys. 18:720–722.
chloride-induced liver fibrosis in mice lacking Res. Commun. 321:413–423. 109. Schwabe, R.F., Schnabl, B., Kweon, Y.O., and
adiponectin. Gastroenterology. 125:1796–1807. 89. Williams, E.J., et al. 2001. Relaxin inhibits effective Brenner, D.A. 2001. CD40 activates NF-kappa
70. Kanno, K., Tazuma, S., and Chayama, K. 2003. collagen deposition by cultured hepatic stellate B and c-Jun N-terminal kinase and enhances
AT1A-deficient mice show less severe progression cells and decreases rat liver fibrosis in vivo. Gut. chemokine secretion on activated human hepatic
of liver fibrosis induced by CCl(4). Biochem. Biophys. 49:577–583. stellate cells. J. Immunol. 166:6812–6819.
Res. Commun. 308:177–183. 90. Cho, J.J., et al. 2000. An oral endothelin-A receptor 110. Schnabl, B., et al. 2001. The role of Smad3 in
71. Ueberham, E., et al. 2003. Conditional tetracycline- antagonist blocks collagen synthesis and deposi- mediating mouse hepatic stellate cell activation.
regulated expression of TGF-beta1 in liver of trans- tion in advanced rat liver fibrosis. Gastroenterology. Hepatology. 34:89–100.
genic mice leads to reversible intermediary fibrosis. 118:1169–1178. 111. Dooley, S., et al. 2003. Smad7 prevents activation
Hepatology. 37:1067–1078. 91. Paizis, G., et al. 2002. Up-regulation of components of hepatic stellate cells and liver fibrosis in rats.
72. Yoshiji, H., et al. 2002. Tissue inhibitor of of the renin-angiotensin system in the bile duct- Gastroenterology. 125:178–191.
metalloproteinases-1 attenuates spontaneous ligated rat liver. Gastroenterology. 123:1667–1676. 112. Marra, F., et al. 2000. Ligands of peroxisome prolif-
liver fibrosis resolution in the transgenic mouse. 92. Bataller, R., et al. 2003. Activated human hepatic erator-activated receptor gamma modulate profi-
Hepatology. 36:850–860. stellate cells express the renin-angiotensin system brogenic and proinflammatory actions in hepatic
73. Agarwal, D.P. 2001. Genetic polymorphisms of alco- and synthesize angiotensin II. Gastroenterology. stellate cells. Gastroenterology. 119:466–478.
hol metabolizing enzymes. Pathol. Biol. 49:703–709. 125:117–125. 113. Galli, A., et al. 2002. Antidiabetic thiazolidinedio-
74. Degoul, F., et al. 2001. Homozygosity for alanine in 93. Jonsson, J.R., et al. 2001. Angiotensin-converting nes inhibit collagen synthesis and hepatic stellate
the mitochondrial targeting sequence of superoxide enzyme inhibition attenuates the progression of cell activation in vivo and in vitro. Gastroenterology.
dismutase and risk for severe alcoholic liver disease. rat hepatic fibrosis. Gastroenterology. 121:148–155. 122:1924–1940.
Gastroenterology. 120:1468–1474. 94. Paizis, G., et al. 2001. Effect of angiotensin II type 114. Boya, P., et al. 2001. Nuclear factor-kappa B in the
75. Jarvelainen, H.A., et al. 2001. Promoter polymor- 1 receptor blockade on experimental hepatic fibro- liver of patients with chronic hepatitis C: decreased
phism of the CD14 endotoxin receptor gene as a genesis. J. Hepatol. 35:376–385. RelA expression is associated with enhanced fibro-
risk factor for alcoholic liver disease. Hepatology. 95. Ramalho, L.N., et al. 2002. Effect of losartan, an sis progression. Hepatology. 34:1041–1048.
33:1148–1153. angiotensin II antagonist, on secondary biliary cir- 115. Rippe, R.A., Schrum, L.W., Stefanovic, B., Solis-
76. Donaldson, P., et al. 2001. HLA and interleukin 1 rhosis. Hepatogastroenterology. 49:1499–1502. Herruzo, J.A., and Brenner, D.A. 1999. NF-kappaB
gene polymorphisms in primary biliary cirrhosis: 96. Wei, H.S., et al. 2000. The regulatory role of AT 1 inhibits expression of the alpha1(I) collagen gene.
associations with disease progression and disease receptor on activated HSCs in hepatic fibrogenesis: DNA Cell Biol. 18:751–761.
susceptibility. Gut. 48:397–402. effects of RAS inhibitors on hepatic fibrosis induced 116. Mann, D.A., and Smart, D.E. 2002. Transcriptional
77. Corpechot, C., et al. 2001. Apolipoprotein E poly- by CCl(4). World J. Gastroenterol. 6:824–828. regulation of hepatic stellate cell activation. Gut.
morphism, a marker of disease severity in primary 97. Wei, H.S., et al. 2000. Effects of AT1 receptor antag- 50:891–896.
biliary cirrhosis? J. Hepatol. 35:324–328. onist, losartan, on rat hepatic fibrosis induced by 117. Paik, Y.H., et al. 2003. Toll-like receptor 4
78. Schuppan, D., Krebs, A., Bauer, M., and Hahn, E.G. CCl(4). World J. Gastroenterol. 6:540–545. mediates inflammatory signaling by bacterial
2003. Hepatitis C and liver fibrosis. Cell Death Differ. 98. Tuncer, I., Ozbek, H., Ugras, S., and Bayram, I. 2003. lipopolysaccharide in human hepatic stellate cells.
10(Suppl. 1):S59–S67. Anti-fibrogenic effects of captopril and candesar- Hepatology. 37:1043–1055.
79. Powell, E.E., et al. 2000. Host genetic factors influ- tan cilexetil on the hepatic fibrosis development 118. Canbay, A., et al. 2003. Cathepsin B inactivation
ence disease progression in chronic hepatitis C. in rat. The effect of AT1-R blocker on the hepatic attenuates hepatic injury and fibrosis during
Hepatology. 31:828–833. fibrosis. Exp. Toxicol. Pathol. 55:159–166. cholestasis. J. Clin. Invest. 112:152–159. doi:10.1172/
80. Sasaki, K., et al. 2000. Mannose-binding lectin 99. Bataller, R., et al. 2000. Angiotensin II induces con- JCI200317740.
polymorphisms in patients with hepatitis C virus traction and proliferation of human hepatic stel- 119.Maher, J.J., Zia, S., and Tzagarakis, C. 1994. Acetal-
infection. Scand. J. Gastroenterol. 35:960–965. late cells. Gastroenterology. 118:1149–1156. dehyde-induced stimulation of collagen synthesis
81. Akuta, N., et al. 2001. Risk factors of hepatitis C 100. Bataller, R., et al. 2003. Prolonged infusion of and gene expression is dependent on conditions
virus-related liver cirrhosis in young adults: posi- angiotensin II into normal rats induces stellate cell of cell culture: studies with rat lipocytes and
tive family history of liver disease and transporter activation and proinflammatory events in liver. Am. fibroblasts. Alcohol Clin. Exp. Res. 18:403–409.
associated with antigen processing 2(TAP2)*0201 J. Physiol. Gastrointest. Liver Physiol. 285:G642–G651. 120. Bataller, R., Paik, Y.H., Lindquist, J.N., Lemasters,
allele. J. Med. Virol. 64:109–116. 101. Wheeler, M.D., et al. 2001. The role of Kupffer cell J.J., and Brenner, D.A. 2004. Hepatitis C virus core
82. Martinelli, A.L., et al. 2000. Are haemochromato- oxidant production in early ethanol-induced liver and nonstructural proteins induce fibrogenic
sis mutations related to the severity of liver dis- disease. Free Radic. Biol. Med. 31:1544–1549. effects in hepatic stellate cells. Gastroenterology.
ease in hepatitis C virus infection? Acta Haematol. 102. Kono, H., et al. 2000. NADPH oxidase–derived free 126:529–540.
102:152–156. radicals are key oxidants in alcohol-induced liver 121. Wanless, I.R., and Shiota, K. 2004. The pathogen-
83. Thorburn, D., et al. 2002. The role of iron and disease. J. Clin. Invest. 106:867–872. esis of nonalcoholic steatohepatitis and other fatty
haemochromatosis gene mutations in the pro- 103. Marra, F. 2002. Leptin and liver fibrosis: a matter of liver diseases: a four-step model including the role
gression of liver disease in chronic hepatitis C. fat. Gastroenterology. 122:1529–1532. of lipid release and hepatic venular obstruction
Gut. 50:248–252. 104.Ikejima, K., et al. 2002. Leptin receptor-medi- in the progression to cirrhosis. Semin. Liver Dis.
84. Day, C.P. 2004. The potential role of genes in ated signaling regulates hepatic fibrogenesis and 24:99–106.
nonalcoholic fatty liver disease. Clin. Liver Dis. remodeling of extracellular matrix in the rat. 122. Arthur, M.J. 2002. Reversibility of liver fibrosis
8:673–691, xi. Gastroenterology. 122:1399–1410. and cirrhosis following treatment for hepatitis C.
85. Marra, F. 2002. Chemokines in liver inflammation 105. Ortiz, V., Berenguer, M., Rayon, J.M., Carrasco, D., Gastroenterology. 122:1525–1528.
and fibrosis. Front. Biosci. 7:d1899–d1914. and Berenguer, J. 2002. Contribution of obesity 123. Issa, R., et al. 2004. Spontaneous recovery from
86. Schwabe, R.F., Bataller, R., and Brenner, D.A. to hepatitis C-related fibrosis progression. Am. J. micronodular cirrhosis: evidence for incomplete
2003. Human hepatic stellate cells express CCR5 Gastroenterol. 97:2408–2414. resolution associated with matrix cross-linking.
and RANTES to induce proliferation and migra- 106. Pinzani, M. 2002. PDGF and signal transduction in Gastroenterology. 126:1795–1808.
tion. Am. J. Physiol. Gastrointest. Liver Physiol. hepatic stellate cells. Front. Biosci. 7:d1720–d1726. 124. Pares, A., Caballeria, J., Bruguera, M., Torres, M.,
285:G949–G958. 107. Marra, F., et al. 1999. Extracellular signal-regulated and Rodes, J. 1986. Histological course of alcoholic
87. Shek, F.W., and Benyon, R.C. 2004. How can trans- kinase activation differentially regulates platelet- hepatitis. Influence of abstinence, sex and extent of
forming growth factor beta be targeted usefully to derived growth factor’s actions in hepatic stellate hepatic damage. J. Hepatol. 2:33–42.
combat liver fibrosis? Eur. J. Gastroenterol. Hepatol. cells, and is induced by in vivo liver injury in the rat. 125. Dixon, J.B., Bhathal, P.S., Hughes, N.R., and
16:123–126. Hepatology. 30:951–958. O’Brien, P.E. 2004. Nonalcoholic fatty liver disease:
88. Borkham-Kamphorst, E., Stoll, D., Gressner, A.M., 108. Schwabe, R.F., et al. 2004. Differential requirement Improvement in liver histological analysis with
and Weiskirchen, R. 2004. Antisense strategy for c-Jun NH2-terminal kinase in TNFalpha- and weight loss. Hepatology. 39:1647–1654.

218 The Journal of Clinical Investigation http://www.jci.org Volume 115 Number 2 February 2005