Oncogene (2010) 29, 4989–5005 & 2010 Macmillan Publishers Limited All rights reserved 0950-9232/10 www.nature.com/onc REVIEW The role of signaling pathways in the development and treatment of hepatocellular carcinoma

S Whittaker1,2, R Marais3 and AX Zhu4

1Dana-Farber Cancer Institute, Boston, MA, USA; 2The Broad Institute, Cambridge, MA, USA; 3Institute of Cancer Research, London, UK and 4Massachusetts General Hospital Cancer Center, Harvard Medical School, Boston, MA, USA

Hepatocellular carcinoma (HCC) is a highly prevalent, malignancy in adults (Pons-Renedo and Llovet, 2003). treatment-resistant malignancy with a multifaceted mole- For the vast majority of patients, HCC is a late cular pathogenesis. Current evidence indicates that during complication of chronic liver disease, and as such, is hepatocarcinogenesis, two main pathogenic mechanisms often associated with cirrhosis. The main risk factors for prevail: (1) cirrhosis associated with hepatic regeneration the development of HCC include infection with hepatitis after tissue damage caused by hepatitis infection, toxins B virus (HBV) or hepatitis C virus (HCV). Hepatitis (for example, alcohol or aflatoxin) or metabolic influ- infection is believed to be the main etiologic factor in ences, and (2) mutations occurring in single or multiple 480% of cases (Anzola, 2004). Other risk factors oncogenes or tumor suppressor genes. Both mechanisms include excessive alcohol consumption, nonalcoholic have been linked with alterations in several important steatohepatitis, autoimmune hepatitis, primary biliary cellular signaling pathways. These pathways are of cirrhosis, exposure to environmental carcinogens (parti- interest from a therapeutic perspective, because targeting cularly aflatoxin B) and the presence of various genetic them may help to reverse, delay or prevent tumorigenesis. metabolic diseases (for example, hereditary hemo- In this review, we explore some of the major pathways chromatosis, tyrosinemia and a1-antitrypsin deficiency; implicated in HCC. These include the RAF/MEK/ERK Llovet et al., 1999; Roberts and Gores, 2005; Thomas pathway, phosphatidylinositol-3 kinase (PI3K)/AKT/ and Abbruzzese, 2005). mammalian target of rapamycin (mTOR) pathway, One of the main reasons for the high mortality rate in WNT/b-catenin pathway, insulin-like growth factor patients with HCC is the lack of effective treatment pathway, hepatocyte growth factor/c-MET pathway and options, especially for those with advanced disease. growth factor-regulated angiogenic signaling. We focus on Although surgery and percutaneous ablation can the role of these pathways in hepatocarcinogenesis, how achieve long-term control in some patients with early they are altered, and the consequences of these abnor- HCC, recurrence rates are high (approximately 50% malities. In addition, we also review the latest preclinical at 3 years; Mulcahy, 2005). Moreover, because of the and clinical data on the rationally designed targeted asymptomatic nature of early HCC, lack of awareness agents that are now being directed against these pathways, and poorly defined screening strategies, most patients with early evidence of success. (approximately 80%) present with advanced or unre- Oncogene (2010) 29, 4989–5005; doi:10.1038/onc.2010.236; sectable disease (Thomas and Abbruzzese, 2005). These published online 19 July 2010 patients generally have a very poor prognosis and treatment options are mainly palliative. Even with Keywords: epidermal growth factor receptor; hepato- treatment (such as transarterial chemoembolization, cellular carcinoma; multikinase inhibitor; signaling; intra-arterial or systemic chemotherapy, radiotherapy, sorafenib; vascular endothelial growth factor immunotherapy or hormonal therapy), the 5-year relative survival rate for patients with HCC is only 7%, and very few patients with symptomatic disease Introduction survive for 41 year (Bosch et al., 2004). The paucity of effective and well-tolerated treatments for advanced Liver cancer is the sixth most common malignancy and HCC (Simonetti et al., 1997; Zhu, 2006) highlights the the third most common cause of cancer-related morta- need for new therapeutic approaches. lity worldwide (Parkin et al., 2005). Hepatocellular In recent years, improved knowledge of oncogenic carcinoma (HCC) is the most common primary liver processes and the signaling pathways that regulate tumor cell proliferation, differentiation, angiogenesis, invasion and metastasis has led to the identification of Correspondence: Dr AX Zhu, Massachusetts General Hospital Cancer several possible therapeutic targets that have driven the Center, Harvard Medical School, Lawrence House/POB 232, 55 Fruit development of molecularly targeted therapies. These Street, Boston, MA 02114, USA. E-mail: [email protected] drugs, which act directly on components of the signaling Received 4 March 2010; accepted 13 May 2010; published online 19 July pathways regulating tumorigenesis, have showed clinical 2010 benefit in patients with various tumor types, including HCC signaling review S Whittaker et al 4990 colorectal, renal, breast and lung cancers, and more larly, tumor growth, vascular invasion (the hallmark of recently, HCC. Because many of the pathways are invasive disease) and metastasis are also critically implicated in the pathogenesis of various tumor types, dependent on efficient angiogenesis (Semela and Du- targeted agents seem to be effective in more than one four, 2004). In HCC, angiogenesis relies on autocrine malignancy. Here we review several major molecular and paracrine interactions between tumor cells, vascular signaling pathways implicated in the pathogenesis of endothelial cells and pericytes (Figure 3) (Folkman, HCC. In addition, we explore the rationale for targeting 2003; Roberts and Gores, 2006). During the angiogenic molecular components of these signaling pathways, process, the existing microvasculature is destabilized, describe the preclinical activity of targeted therapies leading to vascular hyperpermeability, remodeling of the currently in development and approved for HCC, and cellular matrix and activation of endothelial cells briefly review the evidence for their clinical activity in (Papetti and Herman, 2002). Once activated, endothelial patients with HCC. cells proliferate, migrate and undergo cord formation to form new microvessels (Papetti and Herman, 2002). Finally, pericytes are activated and recruited to stabilize Search strategy and selection criteria the new blood vessels. Normal angiogenesis is maintained by the balance Data on the molecular drivers of HCC were identified between proangiogenic and antiangiogenic factors (Seme- for this review by searches of MEDLINE, EMBASE, la and Dufour, 2004). The angiogenic balance is disturbed BIOSIS, ISI and PubMed. Only papers published in in HCC as tumor cells, endothelial cells and pericytes English from 1987 to 2010 were assessed. Recent (2006– secrete a net excess of angiogenic factors, which support 2010) abstracts and reports from meetings were also the recruitment and activation of endothelial cells included if relevant. In addition, the clinicaltrials.gov and pericytes. A number of angiogenic growth factors, website was used to identify studies of agents of including VEGF-A, angiopoietin-2 and PDGF, have been potential interest in HCC. shown to be upregulated in HCC tumors at the level of gene expression and at the plasma protein level in patients with HCC compared with cirrhotic patients (Mas et al., Signaling transduction pathways implicated in HCC 2007). The principal angiogenic factors involved are VEGFs, PDGFs, placental growth factors, transforming HCC has a very complex molecular pathogenesis in growth factor (TGF)-a and -b, basic fibroblast growth which two mechanisms predominate: (1) cirrhosis factor, EGF, HGF, angiopoietins and interleukin-4 and -8 associated with hepatic regeneration after tissue damage (Folkman, 2003; Semela and Dufour, 2004). These growth caused by hepatitis infection, toxin/environmental fac- factors and cytokines induce angiogenic signaling through tors (for example, alcohol or aflatoxin B) or metabolic a variety of mechanisms, including activation of the influences (for example, insulin resistance, obesity, type RAF/MEK/ERK, PI3K/AKT/mTOR and Janus kinase II diabetes or dyslipidemia in nonalcoholic steatohepa- (JAK)/signal transducer and activator of transcription titis-induced HCC; Bugianesi, 2005); and (2) mutations pathways (Roberts and Gores, 2005). occurring in one or more oncogenes or tumor suppres- sor genes (Figure 1) (Thorgeirsson and Grisham, 2002; Wang et al., 2002; Feitelson et al., 2004; Marotta et al., VEGF receptor signaling 2004; Villanueva et al., 2007, 2008). Both of these mechanisms have been associated with abnormalities in One of the most intensely studied growth factors several critical cell signaling pathways that perpetuate involved in angiogenesis is VEGF. Studies of human the carcinogenic process. These signaling cascades are of tumor xenografts in immunodeficient mice showed that interest from a therapeutic perspective, because target- neutralization of VEGF inhibited tumor growth and ing them may help to reverse, delay or prevent hepato- decreased blood vessel density in a variety of tumor carcinogenesis. In this respect, growth factor-mediated types (Kim et al., 1993). Overexpression of VEGF angiogenic signaling (vascular endothelial growth factor may be induced by the hypoxic tumor environment (VEGF), platelet-derived growth factor (PDGF), (mediated by hypoxia-inducible factor 2-a), activation epidermal growth factor (EGF), insulin-like growth of EGF receptor (EGFR) and cyclo-oxygenase-2 factor (IGF), hepatocyte growth factor (HGF)/c-MET), signaling (Avila et al., 2006; Bangoura et al., 2007). and the mitogen-activated protein kinase (MAPK), Increased expression of VEGF and VEGF receptors phosphatidylinositol-3 kinase (PI3K)/AKT/mammalian (VEGFRs; which include VEGFR-1, -2 and -3) has been target of rapamycin (mTOR) and WNT/b-catenin observed in HCC cell lines and tissues, as well as in the pathways are among the most important (Figure 2). serum of patients with HCC (Shimamura et al., 2000; Ng et al., 2001; Dhar et al., 2002; Poon et al., 2004a). The hepatitis B x antigen has also been associated Receptor tyrosine kinases and growth factor-mediated with the upregulation of VEGFR-3 (Lian et al., 2007). angiogenic signaling Furthermore, increased VEGF expression has been reported in cirrhotic and dysplastic liver tissue, suggest- The liver is a highly vascular organ that depends on ing a possible role for VEGF-mediated angiogenesis effective angiogenesis for cellular regeneration. Simi- in hepatocarcinogenesis (El-Assal et al., 1998). VEGF

Oncogene HCC signaling review S Whittaker et al 4991

HBV/HCV, Alcohol, AFB1, NASH

MYC, TGF- amplification, IGF-2 overexpression, RB1 loss, hTERT expression

Loss of p53, p16, IGFR-2, PTEN

-catenin mutation, AXIN1/2 loss, TGF-β activation, HGF/MET overexpression, PIK3CA mutation, in VEGF/microvessel density

Figure 1 Key pathways of HCC development. AFB1, aflatoxin B1; AXIN1/2, axin proteins; AXIN1 is a gene that encodes a cytoplasmic protein that contains a regulation of G-protein signaling (RGS) domain and a Dishevelled and axin domain; mutations in AXIN1 have been associated with hepatocellular carcinoma; AXIN2 (axin-related protein) regulates the stability of b-catenin in the WNT/b-catenin signaling pathway; HBV, hepatitis B virus; HCC, hepatocellular carcinoma; HCV, hepatitis C virus; hTERT, human telomerase reverse transcriptase; overexpression of hTERT leads to dysregulation of the cell cycle; IGF-2, insulin-like growth factor 2; ligand for insulin-like growth factor 1 (IGF-1); IGFR-2, IGF-2 receptor; MET, mesenchymal–epithelial transition factor; also known as hepatocyte growth factor receptor (HGFR); MYC, oncogene that transcriptionally regulates other genes and is associated with carcinogenesis; NASH, nonalcoholic steatohepatitis; p16, cyclin-dependent kinase inhibitor 2A (CDKN2A); a tumor suppressor protein; p53, a tumor suppressor protein; PTEN, phosphatase and tensin homolog; regulates cell- survival pathway; PIK3CA, phosphatidylinositol-3-kinase 110 kDa catalytic subunit; an oncogenic protein; RB1, retinoblastoma protein 1; a tumor suppressor protein that is downregulated in many types of cancer; TGF-a, transforming growth factor-a; TGF-b, transforming growth factor-b; ligand for TGF-b receptor; ligand for epidermal growth factor receptor (EGFR); b-catenin, a subunit of the nuclear protein complex; an integral component of the WNT/b-catenin signaling pathway. clearly has an important regulatory role in HCC; high high frequency (60–80%) of EGFR expression in HCC, levels of VEGF expression have been linked with HCC EGFR mutation in exons 18–21 is not detected (Su et al., tumor grade (Yamaguchi et al., 1998), poor outcome 2006) or rare (p1%) (Wellcome Trust, 2008). However, after resection (Poon et al., 2004b), disease recurrence, the elevated expression of TGF-a reported in preneo- poor disease-free and overall survival (OS) (Chao et al., plastic HCC (Feitelson et al., 2004) and as a ligand of 2003), vascular invasion (Li et al., 1998) and portal vein the EGFR may indicate a role for EGFR signaling emboli (Zhou et al., 2000). during early HCC. The IGF signaling pathway is activated by binding of the ligands (IGF-1 and -2) to the membrane-bound EGFR, IGF and HGF/c-MET signaling receptor (IGF-1R) (Alexia et al., 2004). IGF-1R signaling regulates several cellular processes, including Although the role of VEGF/VEGFR in HCC has proliferation, motility and inhibition of apoptosis. gained more supportive evidence, the relevance of Moreover, expression of IGF-1R is a key regulator for EGFR signaling in HCC is still debatable. Despite the anchorage-independent growth. Ligand binding to

Oncogene HCC signaling review S Whittaker et al 4992 VEGFR WNT PDGFR Cell Membrane Receptor IGFR EGFR SOS Cell Proliferation GRB2 RAS SHC2 DSH PI3K RAF Cell Differentiation PLCx Cell Survival GBP PTEN MEK1/2

AKT PKC GSK3-β BAD ERK1/2 mTOR DSH

FOXO BCL-XL C-MYC C-JUN FOXO β-catenin

p53 Nucleus Cell Transcription/Translation

Figure 2 Cellular signaling pathways implicated in the pathogenesis of hepatocellular carcinoma. AKT, a protein kinase family of genes involved in regulating cell survival; members of this family are also called protein kinases B (PKB); BAD, BCL-2-associated death promoter; BCL-XL, a member of the BCL-2 family of proteins; involved in regulating apoptosis; C-JUN, a protein that interacts with c-FOS, thereby forming the activator protein-1 (AP-1) early-response transcription factor; C-MYC, gene that encodes for a protein that binds to the DNA of other genes; overexpression of C-MYC is associated with carcinogenesis; DSH, downstream effector Dishevelled; EGFR, epidermal growth factor receptor; ERK 1/2, extracellular signal-regulated kinases; FOXO, a protein belonging to the O subclass of the forkhead family of transcription factors; GBP, guanylate-binding protein; GRB2, growth factor receptor-bound protein 2; an adaptor protein involved in signal transduction/cell communication; GSK-3b, glycogen synthase kinase-3b; IGFR, insulin-like growth factor receptor; MEK 1/2, kinases that phosphorylate mitogen-activated protein (MAP) kinase (MAPK); mTOR, mammalian target of rapamycin; p53, a tumor suppressor protein; PDGFR, platelet-derived growth factor receptor; PI3K, phosphatidylinositol-3-kinase; PKC, protein kinase C; PLCx, phospholipase Cx; 13 kinds of mammalian PLCs are classified into six models (b, g, d, e, z and Z); PTEN, phosphatase and tensin homolog; regulates cell-survival pathway; RAF, a MAP kinase kinase kinase (MAP3K) that functions in the MAPK/ERK signal transduction pathway; a serine/threonine-specific kinase; RAS, prototypical member of the RAS superfamily of proteins; activation of RAS signaling causes cell growth, differentiation and survival; dysregulated RAS signaling can lead to oncogenesis and cancer; SHC2, SRC homology 2 domain-containing (SHC)-transforming protein 2; SOS, son of sevenless; a guanine nucleotide exchange factor that acts on RAS-GTPases; so named because the SOS protein that it encoded was found to operate downstream of the sevenless gene in Drosophila melanogaster in a RAS/MAP kinase pathway; VEGFR, vascular endothelial growth factor receptor; WNT, a signaling pathway made up of a complex network of proteins involved in embryogenesis, normal physiological processes and carcinogenesis; b-catenin, a subunit of the nuclear protein complex; an integral component of the WNT/b-catenin signaling pathway.

IGF-1R triggers rapid receptor autophosphorylation, overexpression correlated with increased cell proli- followed by phosphorylation of intracellular targets feration (Schirmacher et al., 1992), whereas inhibition (mainly insulin receptor substrates 1–4), which in turn of IGF-2 production in HCC cells reduced cell proli- initiates downstream cellular effectors, ultimately leading feration and increased apoptosis (Lund et al., 2004). to activation of PI3K, protein kinase B and the HGF is a multifunctional cytokine that has been RAF/MEK/ERK pathway (Pollak et al., 2004). implicated in tumor invasion in several malignancies In HCC, dysregulation of IGF signaling occurs (Matsumoto and Nakamura, 1996). The HGF ligand predominantly at the level of IGF-2 bioavailability. exerts its effect by binding the high-affinity tyrosine kinase IGF-2 is upregulated by epigenetic mechanisms after an receptor c-MET, which is predominantly expressed on the inflammatory response to liver damage or viral trans- surface of epithelial and endothelial cells. In addition to activation, altered methylation of the IGF-2 gene, tissue regeneration, c-MET regulates cell proliferation, inactivation of the P1 promoter or activation of the P3 migration, survival, branching morphogenesis and angio- promoter (Feitelson et al., 2004). IGF-2 is overexpressed genesis. HGF-induced activation of c-MET leads to in 16–40% of human HCCs (Cariani et al., 1988), and phosphorylation of adaptor proteins (including GRB2 IGF-2R (an alternative receptor for IGF-2) is under- (growth factor receptor-bound protein 2) and GAB1 expressed in approximately 80% of HCCs (De Souza (GRB2-associated-binding protein 1)), which then activate et al., 1995). This excess ligand availability leads to downstream effector molecules (including phospholipase elevated receptor binding and subsequent down- C, PI3K and ERK). For example, activation of c-MET stream signaling through the MAPK and PI3K/AKT/ leads to recruitment of the GRB2/SOS (son of sevenless) mTOR pathways. In animal models of HCC, IGF-2 complex to the plasma membrane (Ponzetto et al., 1994).

Oncogene HCC signaling review S Whittaker et al 4993 Tumor Cell

IGF FGF VEGF PDGF

ANG-2 IL-8 HIF

RAS PI3K STATs

Autocrine Loop

Paracrine Loop

RAS PI3K STATs

VEGF Proliferation and Migration

Endothelial Cell Pericyte Figure 3 Angiogenic pathways. ANG-2, angiopoietin-2; FGF, fibroblast growth factor; HIF, hypoxia-inducible factor; IGF, insulin- like growth factor; IL-8, interleukin-8; PDGF, platelet-derived growth factor; PI3K, phosphatidylinositol-3-kinase; RAS, prototypical member of the RAS superfamily of proteins; activation of RAS signaling causes cell growth, differentiation and survival; dysregulated RAS signaling can lead to oncogenesis and cancer; STAT, signal transducer and activator of transcription; VEGF, vascular endothelial growth factor. Adapted from Roberts and Gores (2006).

This translocation can result in GTP-loading on RAS, pathway that regulates crucial cellular processes, which initiates a protein cascade that subsequently leads to including proliferation, differentiation, angiogenesis downstream activation of ERK through RAF and MEK and survival (Gollob et al., 2006). Importantly, the (Schlessinger, 2000). Presumably, other pathways down- overexpression or activation of components of this stream of RAS are also activated. pathway is believed to contribute to tumorigenesis, c-MET overexpression, relative to levels in peri- tumor progression and disease metastasis in a variety tumorous liver tissue, is observed in 20–48% of HCC of solid tumors (Leicht et al., 2007). The ERK/MAPK samples (Boix et al., 1994; Suzuki et al., 1994; Kiss et al., pathway lies downstream of the various growth factors 1997; Ueki et al., 1997; Tavian et al., 2000). Dysregula- described above, and is thus a valid therapeutic target tion of c-MET is associated with various molecular– in HCC. Growth-factor binding results in receptor genetic factors, and overexpression has been linked phosphorylation, which activates an adapter molecule with decreased 5-year survival in patients with HCC complex known as GRB2/SHC/SOS. This in turn activates (Ueki et al., 1997). Moreover, a c-MET-regulated the RAF/MEK/ERK pathway, which triggers a cascade expression signature defines a subset of HCC in humans; of specific phosphorylation events (Avila et al., 2006). these patients have a poor prognosis and an aggressive Within the RAF/MEK/ERK pathway, the small phenotype (Kaposi-Novak et al., 2006). GTPase RAS and the serine/threonine kinase RAF (of which there are three main isoforms—ARAF, BRAF and CRAF) are the key molecular signal regulators The ERK/MAPK pathway (Kolch, 2000; Avila et al., 2006). Intermediate signaling is regulated by the mitogen/extracellular protein kinase The ERK/MAPK pathway (also known as the RAF/ kinases MEK1 and MEK2, which are responsible for MEK/ERK pathway) is a ubiquitous signal transduction phosphorylating and activating the final downstream

Oncogene HCC signaling review S Whittaker et al 4994 signaling molecules extracellular-regulated protein ki- second main mechanism for activating the RAF/MEK/ nases ERK1 and ERK2 (Roberts and Gores, 2005; ERK pathway is constitutive CRAF activation resulting Avila et al., 2006). ERK1/2 regulate cellular activity by from dysregulated overexpression of growth factors and acting on more than 100 substrates in the cytoplasm and their receptors (Gollob et al., 2006). Finally, the RAF/ nucleus, including indirect inducers of gene expression, MEK/ERK pathway has been reported to be activated transcription factors and cell cycle-related kinases through HBV infection in HCC. During chronic (Kolch, 2000; Schulze et al., 2001; Xaus et al., 2001; infection, HBV integrates into host DNA and expresses Sananbenesi et al., 2002; Halaschek-Wiener et al., 2004). two transcriptional activators, both of which trigger Importantly, RAS also has a regulatory role in other activation of the RAF/MEK/ERK pathway (Stockl signaling pathways, most notably the PI3K/AKT/ et al., 2003). The first activator, HBX protein, interacts mTOR pathway, the phospholipase C/protein kinase C with PIN1 (also overexpressed in HCC) (Pang et al., pathway and the RALGDS (ral guanine nucleotide 2007), which may enhance activation of the MAPK dissociation stimulator) pathway (To et al., 2005; pathway through dephosphorylation of CRAF, promot- Harden and Sondek, 2006). ing its activation by RAS (Dougherty et al., 2005), In HCC, the RAF/MEK/ERK pathway is constitu- whereas the second activator, PreS2-activator large tively activated, suggesting a possible role for this surface protein, activates the pathway through a protein pathway in tumorigenesis. Hwang et al. (2004) reported kinase C-dependent mechanism (Stockl et al., 2003). overexpression of CRAF in 100% of 30 HCC tissue Note that we have previously shown that when CRAF is specimens tested, and concluded that CRAF activation activated downstream of protein kinase C, it occurs in a may have an important role in HCC. Furthermore, after RAS-dependent manner (Marais et al., 1998) and hence immunohistochemical evaluation of tissue samples, it seems likely that CRAF activation downstream of there was an approximately sevenfold increase in protein kinase C in HCC cells is also RAS dependent. MEK1/2 phosphorylation in HCC tissues, compared HCV may also activate RAF kinase through its core with that in adjacent benign liver tissues (Huynh et al., protein (Aoki et al., 2000). 2003). The MAPK negative regulatory proteins RKIP (phosphatidylethanolamine binding protein 1) and the Sprouty (Spry)/Spred proteins are downregulated in PI3K/AKT/mTOR signaling pathway HCC tumors, and this decrease is postulated to have a role in the excessive activation of the MAPK pathway Constitutive activation of the PI3K/AKT/mTOR observed in HCC (Schuierer et al., 2006; Yoshida et al., signaling pathway has been firmly established as a 2006). Other studies using in vivo HCC models and major determinant of tumor cell growth and survival in human HCC tissue specimens have shown an increase in a multitude of solid tumors (Chen et al., 2005). In the expression and activity of signaling intermediates, the PI3K/AKT/mTOR signaling pathway, binding of such as phosphorylated ERK, compared with surround- growth factors (most notably IGF and EGF) to their ing liver tissue (McKillop et al., 1997; Ito et al., 1998; receptors activates PI3K (Avila et al., 2006). PI3K Feng et al., 2001). In terms of the clinical relevance of subsequently produces the lipid second messenger RAF/MEK/ERK activation, MAPK/ERK activity was PIP3b (phosphoinositoltriphosphate), which in turn shown to correlate positively with tumor size, but not activates the serine/threonine kinase AKT. In addition with HCC stage or the degree of differentiation in a to regulating various transcription factors (for example, histopathological investigation of 26 human HCC tissue FOXO (mammalian forkhead members of the class O)) samples (Ito et al., 1998). In a more recent immunohis- (Greer and Brunet, 2005), activated AKT also phos- tochemical study of tumor samples taken from 208 phorylates several cytoplasmic proteins, most notably patients with HCC who underwent resection or liver mTOR and BCL-2-associated death promoter (BAD; transplantation, ERK1/2 activation was shown to be Avila et al., 2006). The activation of mTOR increases associated with aggressive tumor behavior (Schmitz cellular proliferation, and inactivation of BAD et al., 2008). More importantly, ERK1/2 activation was decreases apoptosis and increases cell survival (Roberts also shown to be an independent prognostic marker for and Gores, 2005). In normal tissue, this pathway is decreased OS. negatively regulated by the phosphatase and tumor Activation of the RAF/MEK/ERK pathway in solid suppressor phosphatase on chromosome 10 (phos- tumors usually occurs by one of two main mechanisms. phatase and tensin homolog (PTEN)), which targets The first is through oncogenic mutations within the the lipid products of PI3K for dephosphorylation RAS gene, which lead to constitutive pathway activa- (Roberts and Gores, 2005). tion through CRAF (Gollob et al., 2006). Mutation of The PI3K/AKT/mTOR signaling pathway can be the NRAS gene has been reported in up to 30% of HCC overactivated by enhanced stimulation of receptor tumors (Scharovsky et al., 2000; Downward, 2003); tyrosine kinases, particularly the IGF receptor and however, the COSMIC (Catalog of Somatic Mutations EGFR. Expression of both IGF and IGF receptor is in Cancer) database suggests that this figure is closer to upregulated in HCC and human cirrhotic liver (Alexia 4%, with KRAS being mutated in another 6% of cases et al., 2004), resulting in stimulation of the PI3K/AKT/ (Wellcome Trust, 2008). Mutations within the BRAF mTOR signaling pathway, in addition to activation gene are extremely rare in HCC; only two are described of the RAF/MEK/ERK and WNT/b-catenin pathways on the COSMIC website (Tannapfel et al., 2003). The (Desbois-Mouthon et al., 2001; Alexia et al., 2004).

Oncogene HCC signaling review S Whittaker et al 4995 Similarly, EGF and related growth factors are com- form of mTOR have been shown to be elevated in monly overexpressed in HCC (Yeh et al., 1987; Carlin 15% of cases of HCC, and levels of total p70 S6 kinase et al., 1988; Kira et al., 1997; Ito et al., 2001). (the immediate substrate for phosphorylated mTOR) Evidence also suggests that anomalies in PTEN have been shown to be increased in 45% of cases function may lead to overactivation of the PI3K/AKT/ (Sahin et al., 2004). mTOR pathway in HCC. The PTEN gene is mutated in 5% of HCCs (Wellcome Trust, 2008), is frequently deleted, and its expression is reduced in nearly half of The WNT/b-catenin pathway all HCC tumors, resulting in constitutive activation of the PI3K/AKT/mTOR pathway (Hu et al., 2003). A major and early carcinogenic event in the develop- In addition, PTEN expression can be downregulated ment of HCC seems to be the abnormal regulation of directly by the HBX protein in HBV-infected patients the transcription factor b-catenin, a key component of (Feitelson et al., 2004). Importantly, downregulation the WNT signaling pathway (De La et al., 1998). In the of PTEN expression has been shown to correlate normal state, the binding of members of a family of with increased tumor grade, advanced disease stage soluble cysteine-rich glycoprotein ligands, the WNTs, to and reduced OS in patients with HCC (Hu et al., 2003). members of the Frizzled family of cell-surface receptors In a transgenic hepatocyte-specific, PTEN-deficient results in the activation of the WNT signaling pathway mouse model, the livers of 40-week-old PTEN-deficient (Avila et al., 2006). Receptor binding activates DSH mice revealed the macrovesicular steatosis, ballooning (downstream effector Dishevelled), which consequently hepatocytes, lobular inflammatory cell infiltration and prevents phosphorylation of b-catenin by glycogen perisinusoidal fibrosis that are characteristic of human synthase kinase-3b and its subsequent ubiquitination nonalcoholic steatohepatitis; by 80 weeks of age, 100% and proteasomal degradation. An ensuing increase in of PTEN-deficient mouse livers showed adenomas and the cytoplasmic concentrations of b-catenin results in its 66% of the mice had HCC (Watanabe et al., 2005). translocation from the cytoplasm to the nucleus (Avila AKT phosphorylation has been implicated in early et al., 2006). Once in the nucleus, b-catenin acts as a co- HCC recurrence and poor prognosis (Nakanishi et al., activator to stimulate the transcription of genes and 2005), and a recent microarray study found that 23% of expression of gene products involved in cell proliferation HCC patients had elevated levels of AKT phosphoryla- (for example, MYC, MYB, CJUN and CYCD1), angio- tion on Ser473 (Boyault et al., 2007). There is also some genesis, antiapoptosis and the formation of extracellular evidence suggesting that PI3K/AKT/mTOR signaling matrix (Avila et al., 2006). may be activated by somatic mutations in the PI3K According to the COSMIC database, b-catenin is catalytic a gene PIK3CA, which encodes the p110a catalytic mutated in 17% of HCCs, with sites of mutation subunit of PI3K. In one study, somatic PIK3CA mutations including those phosphorylated by glycogen synthase were detected in 26 of 73 (35.6%) HCC specimens tested kinase-3b (which regulates degradation of b-catenin)— (Lee et al., 2005), whereas much lower frequencies have an early event in HCC. It has also been reported that in been reported in studies from other geographic regions 12–26% of HCCs, b-catenin accumulation and stabili- (Tanaka et al., 2006; Boyault et al., 2007). zation results from mutations in the b-catenin gene, and More recently, Villanueva et al. (2008) performed an in 8–13% of cases it results from mutations in the genes analysis of integrative genomic data derived from a large that activate b-catenin signaling (Giles et al., 2003). cohort of human tissue samples (314 from HCC tumoral In either case, these mutations seem to be particularly tissue and 37 from nontumoral tissue). Assessments common in HCCs associated with chronic HCV included direct-sequencing mutation analysis, a single- infection (Giles et al., 2003). In HCC, approximately nucleotide polymorphism array to detect DNA copy 50–70% of tumors have increased levels of b-catenin number changes, real-time PCR and gene expression in the cytoplasm and nucleus (Wong et al., 2001). This microarray to detect mRNA levels and immunostaining accumulation of b-catenin provides a growth advantage to detect protein activation. Dual blockade of mTOR to tumor cells by promoting proliferation and suppres- signaling was achieved using a rapamycin analog sing differentiation. b-catenin accumulation alone, how- (everolimus) and an EGFR/VEGFR inhibitor, and the ever, does not seem to cause progression to HCC from effects of this blockade were evaluated in HCC cell lines a nonmalignant state. Studies with b-catenin transgenic and in a tumor xenograft model. Aberrant mTOR mouse models indicate that abnormal WNT signaling signaling was observed with activation of either the IGF can cause severe hepatomegaly, but is not sufficient for and/or EGF cascade. mTOR blockade with everolimus carcinogenic transformation (Giles et al., 2003). Inter- slowed tumor growth and increased survival in the estingly, targeted disruption of the Iqgap2 gene in mice HCC xenograft model, an effect that was enhanced leads to the development of HCC, which is associated in vivo with EGFR/VEGFR blockade. These results with Iqgap1 overexpression and the activation of add considerably to the body of research suggesting b-catenin and cyclin D1. When Iqgap2-null mice were that mTOR pathway activation has a crucial role in the bred with Iqgap1-null mice, the HCC phenotype was pathogenesis of HCC. reduced. This suggests that Iqgap2 acts as a tumor Taken together, these data suggest that the PI3K/ suppressor, and its loss can lead to b-catenin activation AKT/mTOR pathway has a critical role in the patho- and the development of HCC, and it further implicates genesis of HCC. Indeed, levels of the phosphorylated b-catenin as a key driver of HCC (Schmidt et al., 2008).

Oncogene HCC signaling review S Whittaker et al 4996 Direct mutation of b-catenin is not the only route Bevacizumab (Avastin; Genentech, Inc., South San through which the WNT pathway can be aberrantly Francisco, CA, USA) is a recombinant, humanized, activated in HCC. Hoshida et al. (2009) performed an anti-VEGF monoclonal antibody (Presta et al., 1997) analysis of gene expression profiles from 603 HCC that has shown encouraging early evidence of efficacy in patients in an effort to define molecular drivers of the patients with advanced HCC (Siegel et al., 2008; Zhu, disease. Three subclasses of HCC were characterized, two 2008). In two small monotherapy studies, treatment of which showed either increased WNT pathway activity with bevacizumab 5–10 mg/kg resulted in partial res- or increased MYC/AKT pathway activity. The mechan- ponse (PR) in 8–13% of patients and stable disease (SD) ism through which WNT pathway activation may occur in 54–72% of patients (Malka et al., 2007; Siegel et al., was determined to be mediated by TGF-b. A strong 2008). The combination of bevacizumab with either association between WNT activation and TGF-b target cytotoxic regimens or erlotinib ((OSI-774) Tarceva; genes was found, and treatment of an HCC cell line with OSI Pharmaceuticals, Inc., Melville, NY, USA) has TGF-b led to activation of a b-catenin reporter construct. also shown encouraging results in patients with Given the dearth of precise information about the key advanced HCC in four phase II trials (Zhu et al., molecular instigators of HCC, such classifications may 2006; Sun et al., 2007; Thomas et al., 2009; Hsu et al., aid the stratification of patients for future clinical trials of 2010). In one phase II trial reported by Zhu et al. (2006), agents targeting these pathways. the combination of bevacizumab plus gemcitabine and oxaliplatin showed moderate activity, with a 20% overall response rate in evaluable patients. SD was Therapeutic targets in HCC achieved in an additional 27% of patients (median duration 9 months; range 4.5–13.7 months). Median OS Anti-VEGF/VEGFR therapies and median progression-free survival (PFS) were 9.6 As previously mentioned, several studies have estab- and 5.3 months, respectively. Sun et al. (2007) reported lished that tumor growth and invasion in HCC are the results of a phase II trial performed to evaluate dependent on dysregulated angiogenesis. There is, there- bevacizumab in combination with capecitabine and fore, a strong rationale for targeting growth factors that oxaliplatin. Of 30 evaluable patients, 4 (13.3%) had a drive the angiogenic process as a potential therapeutic PR and 23 (76.6%) had SD. Median OS was 10.3 strategy for the treatment of HCC. VEGF is a key months, and the mean time to progression (TTP) was angiogenic factor, and several agents that target VEGF 4.5 months. In a third phase II trial performed by or VEGFR are currently in development for the Hsu et al. (2010), the combination of bevacizumab plus treatment of HCC (Table 1) (Koch et al., 2005; Alberts capecitabine was evaluated. On the basis of data from et al., 2007; Malka et al., 2007; Siegel et al., 2008; Toh 44 evaluable patients, the overall response rate was et al., 2009). 9% and the disease control rate (defined in this study as

Table 1 Molecularly targeted therapies currently available or in development for the treatment of hepatocellular carcinoma Therapy Mode(s) of action Stage of References developmenta

Anti-VEGF/VEGFR therapies Bevacizumab Recombinant, humanized, IgG1, anti-VEGF monoclonal antibody Phase II Siegel et al. (2008); Malka et al. (2007) Vatalanib Oral tyrosine kinase inhibitor of VEGFR-1, VEGFR-2, VEGFR-3, Phase I Koch et al. (2005) PDGFR, c-KIT Cediranib Oral tyrosine kinase inhibitor of VEGFR-1, VEGFR-2, VEGFR-3 Phase II Alberts et al. (2007) Linifanib VEGFR, PDGFR Phase II/III Toh et al. (2009)

Anti-EGF/EGFR therapies Erlotinib Oral tyrosine kinase inhibitor of EGFR Phase II Philip et al. (2005); Thomas et al. (2007) Lapatinib Oral, dual tyrosine kinase inhibitor of EGFR and HER-2 Phase II Ramanathan et al. (2006) Gefitinib Oral tyrosine kinase inhibitor of EGFR Phase II O’Dwyer et al. (2006) Cetuximab Recombinant, chimeric human/mouse, IgG1, anti-EGFR monoclonal Phase II Zhu et al. (2007); antibody Gruenwald et al. (2007)

Multikinase inhibitors Sorafenib Oral multikinase inhibitor of VEGFR-1, VEGFR-2, VEGFR-3, PDGFR-a, Approved for treat- Llovet et al. (2008) PDGFR-b, p38MAPK, FLT-3, c-KIT, RET ment of HCC Sunitinib Oral tyrosine kinase inhibitor of VEGFR-1, VEGFR-2, VEGFR-3, Phase II/III Zhu et al. (2009); PDGFR-b, FLT-3, c-KIT, RET Faivre et al. (2009) Brivanib VEGFR-2 and FGFR-1 Phase II/III Raoul et al. (2009)

Abbreviations: EGF, epidermal growth factor; EGFR, EGF receptor; FGFR, fibroblast growth factor receptor; FLT-3, FMS-like tyrosine kinase 3; HCC, hepatocellular carcinoma; HER-2, human epidermal growth factor receptor-2; IgG1, immunoglobulin 1; MAPK, mitogen-activated protein kinase; PDGFR, platelet-derived growth factor receptor; VEGF, vascular endothelial growth factor; VEGFR, VEGF receptor. aOngoing clinical trials are listed on www.clinicaltrials.gov.

Oncogene HCC signaling review S Whittaker et al 4997 complete response [CR] þ PR þ SD) was 52%. In patients decreased interleukin-6 and sc-KIT after 14 days of treat- with the CLIP (Cancer of the Liver Italian Program) ment showed improved PFS and OS. In a second phase II score of p3, median OS and median PFS were 8.2 study (sunitinib 50 mg/day for 4 weeks out of every 6), and 3.6 months, respectively. Finally, in a single-center one patient (out of 37) had a confirmed PR and 35% of phase II trial reported by Thomas et al. (2009), the patients had SD as their best response (Faivre et al., 2009). combination of bevacizumab plus erlotinib resulted in Median OS and PFS were 8.0 and 3.7 months, respectively. PR in 10 of 40 evaluable patients (25%), with a median However, significant toxicities, including four deaths, PFS of 9 months and a median OS of 15.7 months. were observed. A low response rate and failure to meet a Sunitinib (Sutent; Pfizer Labs, New York, NY, USA) primary end point based on RECIST (Response Evaluation is an orally administered small molecule that inhibits Criteria in Solid Tumors) led to the termination of the members of the split-kinase domain family of receptor trial before entering the second phase. A phase III study tyrosine kinases, including VEGFR-1 and -2, PDGF comparing sunitinib with sorafenib was discontinued receptor (PDGFR)-a and -b, the stem cell factor following a review by the Independent Data Monitoring receptor c-KIT, FLT3 (FMS-like tyrosine kinase 3) Committee both because of a greater incidence of adverse kinase, RET kinase and many others (Mendel et al., events in the sunitinib group and because sunitinib failed 2003; Chow and Eckhardt, 2007; Zhu, 2008). Preclinical to demonstrate that it was either superior or not inferior to studies in a variety of tumor cell lines have shown that sorafenib in extending OS. the antiangiogenic effects of sunitinib are mediated Sorafenib (Nexavar; Bayer HealthCare Pharmaceuticals through VEGFR and PDGFR-b (Murray et al., 2003; Inc., Wayne, NJ, USA) is an orally available multikinase Abrams et al., 2003a, b), although the primary target of inhibitor originally designed to target CRAF. Although sunitinib is likely to be VEGFR-2. Two phase II studies sorafenib inhibits oncogenic BRAF in vitro andincell of sunitinib in patients with advanced HCC have been lines, it does not seem to be sufficiently potent to target undertaken. In the first study, sunitinib was adminis- BRAF in vivo (Eisen et al., 2006). Indeed, we recently tered at a dose of 37.5 mg/day (4 weeks on, 2 weeks off). demonstrated that the anti-tumor activity of sorafenib is One patient (out of 34) had a PR and 17 patients (50%) independent of any activity that it may possess against achieved SD for at least 12 weeks (Zhu et al., 2009). BRAF (Whittaker et al., 2010), supporting the view that Median PFS in this cohort was 3.9 months, and TTP sorafenib is not a BRAF inhibitor in vivo. However, was 4.1 months. Sunitinib treatment resulted in increased sorafenib is a potent inhibitor of VEGFR and PDGFR, VEGF, PDGF and stromal-derived factor-1a, whereas and these receptor tyrosine kinases are likely to be among decreases were observed in soluble VEGFR-2, soluble its clinical targets. Consequently, there is a good rationale VEGFR-3 and circulating progenitor cells. Patients with for testing sorafenib in HCC (Figure 4) (Wilhelm et al.,

G VEGF PDGFR EGF VEGFR IGFR EGFR A Lapatinib E Cell Membrane F B E WNT F B Cetuximab A H I HER2/neu C Frizzled C Erlotinib D D Gefitinib A E Sorafenib F Sunitinib G Bevacizumab PI3K RAS H Cediranib DSH I Vatalanib PTEN RAF J Everolimus AKT GSK3-β K Sirolimus L AZD6244 MEK L β-catenin

BAD mTOR J ERK K Figure 4 Targeted therapies currently available or in development for the treatment of hepatocellular carcinoma, and the molecular targets on which they are believed to act. AKT, a protein kinase family of genes involved in regulating cell survival; members of this family are also called protein kinases B (PKB); BAD, BCL-2-associated death promoter; DSH, downstream effector Dishevelled; EGF, epidermal growth factor; EGFR, EGF receptor; ERK, extracellular signal-regulated kinase; Frizzled, a family of G-protein- coupled receptor proteins that serve as receptors in the WNT/b-catenin signaling pathway and other signaling pathways; when activated, Frizzled leads to activation of Dishevelled in the cytoplasm; GSK-3b, glycogen synthase kinase 3b; HER2/neu, human epidermal growth factor receptor 2, also known as ERBB-2, CD 340 and p185; a cell membrane surface-bound receptor tyrosine kinase involved in the signal transduction pathways leading to cell growth and differentiation; MEK, kinases that phosphorylate mitogen- activated protein (MAP) kinase (MAPK); mTOR, mammalian target of rapamycin; PDGFR, platelet-derived growth factor receptor; PI3K, phosphatidylinositol-3-kinase; PTEN, phosphatase and tensin homolog; regulates cell-survival pathway; RAF, a MAP kinase kinase kinase (MAP3K) that functions in the MAPK/ERK signal transduction pathway; a serine/threonine-specific kinase; RAS, prototypical member of the RAS superfamily of proteins; activation of RAS signaling causes cell growth, differentiation and survival; dysregulated RAS signaling can lead to oncogenesis and cancer; VEGF, vascular endothelial growth factor; VEGFR, VEGF receptor; IGFR, insulin-like growth factor receptor; WNT, a signaling pathway made up of a complex network of proteins involved in embryogenesis, normal physiological processes and carcinogenesis; b-catenin, a subunit of the nuclear protein complex; an integral component of the WNT/b-catenin signaling pathway.

Oncogene HCC signaling review S Whittaker et al 4998 2004; Carlomagno et al., 2006; Lierman et al., 2006; Wilhelm activity is mediated through its ability to target the et al., 2006). Consistent with this, preclinical evaluations receptor tyrosine kinases that drive tumor neoangio- suggest that the antitumor activity of sorafenib in HCC genesis and/or through ‘off-target’ effects. The avail- may be mediated by inhibition of angiogenesis (through ability of potent new anti-RAF drugs, such as PLX4032 inhibition of receptor tyrosine kinases), although direct (Flaherty et al., 2009) and XL281 (Schwartz et al., effects on cell proliferation and survival cannot be 2009), will allow the importance of RAF as a target in excluded (Wilhelm et al., 2004; Liu et al., 2005). HCC to be tested in preclinical models. However, care The promising preclinical activity of sorafenib has must be exercised. PLX4032 is a highly potent BRAF since been supported by clinical studies (Zhu, 2008). In a drug and is unlikely to be of value in HCC because phase II study of 137 patients with advanced inoperable BRAF is not implicated in HCC. More importantly, HCC, single-agent sorafenib (400 mg twice daily) BRAF-selective inhibitors cause paradoxical pathway induced a PR in 2% of patients, a minor response in activation in tumor cells in which RAS is mutated or 6% and SD lasting X4 months in 34% of patients activated by upstream components (Hatzivassilou et al., (Abou-Alfa et al., 2006). In this study, median TTP was 2010; Heidorn et al., 2010; Poulikakos et al.,2010).BRAF 4.2 months and median OS was 9.2 months. In the inhibitors may therefore accelerate rather than suppress pivotal, randomized phase III SHARP (Sorafenib HCC the growth of some HCC tumors. Nevertheless, results Assessment Randomized Protocol) trial, sorafenib 400 mg with the MEK inhibitors suggest the potential value of twice daily significantly prolonged OS compared with developing pan-RAF or CRAF-specific inhibitors in HCC; placebo in patients with advanced HCC (10.7 months in possibly as prophylaxis in HCV-infected patients. the sorafenib group vs 7.9 months in the placebo group; Vatalanib (PTK787) is an oral, small-molecular-weight Llovet et al., 2008). Median time to radiological progres- tyrosine kinase inhibitor that potently inhibits VEGFR-1 sion was significantly longer in the sorafenib group (5.5 vs and -2 (Wood et al., 2000). In a preclinical study, it 2.8 months). Significantly, this is the first time a systemic significantly reduced tumor volume and microvessel therapy has showed a survival advantage in a phase III density in an HCC xenograft model (Liu et al., 2005). study in patients with HCC. In another randomized It also inhibited tumor cell proliferation and induced phase III study performed in the Asia-Pacific region, apoptosis in a dose-dependent manner (Liu et al., 2005). sorafenib also showed improved OS in patients with In addition, it induced growth arrest in HCC cell lines, advanced HCC, most of whom had HBV infection as an which was associated with cell cycle arrest at G1 and etiologic factor (Cheng et al., 2009). OS was 6.5 months partial blockade of the G2-M checkpoint. These preclinical in the sorafenib group compared with 4.2 months in findings have prompted a phase I study in which vatalanib the placebo group (hazard ratio in the sorafenib group, was given at either 750 or 1250 mg/day. Patients were 0.68; P ¼ 0.014). On the basis of these data, sorafenib stratified into three groups: those with mild, moderate has become the only targeted therapy approved for or severe hepatic dysfunction based on total bilirubin clinical use in several countries, including the United and aspartate transaminase/alanine transaminase levels. States (for the treatment of patients with unresectable The maximum tolerated dose of PTK787 was defined as HCC) and in Europe (for the treatment of HCC). 750 mg daily. Of patients in all groups, 18 were evaluable Recently, encouraging data emerged for the combina- for efficacy. No CR or PR was observed. Nine patients had tion of sorafenib and doxorubicin. An interim analysis a best response of SD (Koch et al., 2005). of a randomized phase II study found a prolonged Cediranib (AZD2171) is another selective inhibitor of TTP (median 8.6 months) and OS (median 13.7 months) VEGFR-1, -2 and -3. Preclinical studies showed that with this combination in patients with advanced cediranib inhibits the growth of a variety of human HCC (Abou-Alfa et al., 2008). tumor xenografts, and reduces microvessel density in Although RAS mutations are relatively rare in human lung xenografts in a dose-dependent manner HCC, and RAF mutations are almost nonexistent, the (Wedge et al., 2005). Given at 45 mg daily, it is currently importance of the MAPK pathway in HCC is high- being evaluated in a phase II trial in patients with lighted by preclinical studies showing that the MEK1/2 unresectable HCC (Alberts et al., 2007). inhibitor AZD6244 blocks growth and promotes apop- Brivanib (BMS-582664) is a dual inhibitor of VEGFR tosis in primary HCC cells in vitro, and suppresses and fibroblast growth factor receptor signaling path- tumor growth in HCC xenograft models. The antitumor ways that can induce tumor growth inhibition in mouse activity of AZD6244 correlates with levels of phos- HCC xenograft models. A phase II study was conducted phorylated MEK (Huynh et al., 2007b) and one study to assess the efficacy and safety of brivanib in patients suggests that AZD6244 may be effectively combined with unresectable locally advanced or metastatic HCC with doxorubicin in patients with HCC (Huynh et al., who had received either no previous systemic therapy 2007a). These data suggest that CRAF drugs could also (cohort A) or one previous regimen of an angiogenesis produce antitumor activity in HCC, and a recent report inhibitor (cohort B) (Raoul et al., 2009). The treatment has highlighted the importance of CRAF in HCV schedule consisted of continuous daily dosing of viral replication, suggesting that it may also have a role brivanib 800 mg. Of the 96 patients enrolled, 55 were in HCC tumor initiation (Himmelsbach et al., 2009). in cohort A and 41 (including 38 in whom sorafenib Although sorafenib was designed as a CRAF inhibitor, had failed) in cohort B. In cohort A, median OS was there is no evidence to suggest that it targets CRAF in 10 months and median TTP was 2.8 months (95% HCC tumors—rather, the data suggest that its clinical confidence interval (CI) 1.4–3.9). PR was observed in

Oncogene HCC signaling review S Whittaker et al 4999 5% of patients and the disease control rate was 47%. 150 mg induced a PR in 9% of patients (n ¼ 3) and a Brivanib is undergoing additional evaluation in phase disease control rate (tumor response or SD X8 weeks) III studies in both the first-line setting in comparison of 59%. In all, 12 patients (32%) were progression free with sorafenib and in the sorafenib-refractory setting in at 24 weeks, and median OS was 13 months. In a report comparison with best supportive care in advanced HCC. by Thomas et al. (2009), 17 out of 40 patients (43%) Linifanib (ABT-869) is a novel receptor tyrosine with unresectable HCC achieved PFS after 16 weeks kinase inhibitor with potent activity against members of of treatment with erlotinib 150 mg daily; at 24 weeks, the VEGFR and PDGFR families (Albert et al., 2006). PFS was 28%. No complete or partial responses Although not a general antiproliferative agent, it can were observed. The median time to failure (defined as reduce cell growth and increase apoptosis in tumors with either disease progression or death) was 13.3 weeks. The FLT3-dependent proliferation. Preliminary results from median OS was 25 weeks (95% CI 17.9–42.3) after an open-label, multicenter, phase II study of linifanib in initiation of erlotinib therapy. The phase III placebo- advanced HCC have been reported (Toh et al., 2009). controlled, double-blind SEARCH (Sorafenib and Linifanib 0.25 mg/kg was administered daily to patients Erlotinib, a Randomized Trial Protocol for the Treat- with Child–Pugh class A cirrhosis and once every other ment of Patients with HCC) trial, which is being day to those with Child–Pugh class B cirrhosis until conducted in patients with advanced HCC and under- progressive disease or intolerable toxicity. Of the 44 lying Child–Pugh class A cirrhosis, seeks to determine patients enrolled, 34 (28 with Child–Pugh class A and 6 whether the OS observed with sorafenib in advanced with Child–Pugh class B cirrhosis) were available for HCC can be improved by combining sorafenib with analysis. The estimated response rate was 8.7% (95% CI erlotinib to achieve combined inhibition of the EGF, 1.1–28) for the 23 patients with Child–Pugh class A VEGF and RAS/RAF/MEK signaling pathways. cirrhosis. For all 34 patients, median TTP was 112 days Lapatinib (GW572016) is an oral dual tyrosine kinase (95% CI 110–not estimable), median PFS was 112 days inhibitor that targets both EGFR and the human (95% CI 61–168) and median OS was 295 days (95% CI EGFR-2 (HER-2) (Johnston and Leary, 2006). Phase 182–333). II results indicate that lapatinib is well tolerated and shows preliminary evidence of antitumor activity in HCC (Ramanathan et al., 2006). Among 40 patients with Anti-EGF/EGFR therapies advanced HCC, the response rate was 5%, median PFS Ligands that bind to the EGFR, such as EGF, have a was 2.3 months (95% CI 1.7–5.6) and median OS was vital role in both tumor angiogenesis and proliferation, 6.2 months (95% CI 5.1–N)(Ramanathanet al.,2009). thought to be primarily through activation of the RAF/ Gefitinib ((ZD1839) Iressa; Astrazeneca Pharmaceuticals MEK/ERK and PI3K/AKT/mTOR pathways. Because LP, Wilmington DE, USA) is an oral small-molecule of their efficacy in other solid tumors and the integral EGFR tyrosine kinase drug that showed activity role of growth factors in HCC development and in HCC cell lines and animal models of HCC. In progression, it was hypothesized that agents specifically these preclinical studies, gefitinib inhibited tumor targeting EGF/EGFR signaling may also be beneficial cell growth (Matsuo et al., 2003; Okano et al., 2006), in HCC. These agents include the tyrosine kinase angiogenesis (Ueda et al., 2006) and intrahepatic inhibitors erlotinib, lapatinib and gefitinib, and the metastasis (Matsuo et al., 2003); induced apoptosis monoclonal antibody cetuximab. Because of their (Hopfner et al., 2004) and cell cycle arrest (Hopfner promising activity in other tumor types and the et al., 2004); and prevented the development of HCC rationale for targeting EGFR in HCC, these agents in cirrhotic livers (Schiffer et al., 2005). Gefitinib treat- have also been tested in HCC. ment resulted in reduced RAF/MEK/ERK, AKT and Erlotinib is an orally active, low-molecular-weight TNF-a signaling; cell cycle arrest; suppression of drug that acts as a potent and reversible inhibitor of antiapoptotic protein expression; and stimulation of EGFR tyrosine kinase activity (Grunwald and Hidalgo, proapoptotic protein expression (Hopfner et al., 2004; 2003). In an in vitro study, it was shown to potently Okano et al., 2006; Ueda et al., 2006). However, these suppress the growth of human EGFR-expressing HCC encouraging preclinical results have not been matched cell lines (Huh-7 and HepG2) in a dose-dependent in clinical studies. In a phase II study of patients manner by inducing both apoptosis and cell cycle arrest with advanced unresectable HCC, single-agent gefitinib at the G1/S-transition (Huether et al., 2005). In a 250 mg/day showed low activity (O’Dwyer et al., 2006). subsequent study, it was shown to inhibit the RAF/ Of 31 patients, 1 had a PR and 7 had SD. Median PFS MEK/ERK signaling pathway and block signal trans- was 2.8 months and median OS was 6.5 months. Given ducer and activator of transcription-mediated signaling these results, it is unlikely that gefitinib will be investi- in Huh-7 and HepG2 cells (Huether et al., 2006). These gated further in advanced HCC, although studies are actions resulted in overexpression of proapoptotic ongoing in the adjuvant setting (www.clinicaltrials.gov). factors and downregulation of antiapoptotic factors The recombinant chimeric monoclonal immuno- (for example, BCL-2, BCL-XL and JUN D) and cell globulin 1 antibody cetuximab ((IMC-C225) Erbitux; cycle-regulating genes that induce G1/G0 arrest. Prelimi- ImClone LLC, New York, NY, and Bristol-Myers nary antitumor activity has been reported in a phase II Squibb, Princeton, NJ, USA), which targets the extra- study of 38 patients with unresectable or metastatic HCC cellular domain of the EGFR, is also under investigation (Philip et al., 2005). Daily administration of erlotinib in HCC. In an initial in vitro study, cetuximab inhibited

Oncogene HCC signaling review S Whittaker et al 5000 cell growth, induced cell cycle arrest and moderately mTOR inhibition with sirolimus significantly reduced increased apoptosis in p53 wild-type HepG2 cells HCC growth and improved survival compared with no (Huether et al., 2005). However, it only had limited treatment. The antitumor activity of sirolimus was effects in Huh-7 cells that carried a mutated p53 tumor attributed primarily to its antiangiogenic effects. A suppressor gene (a common mutation in HCC). As with study by Kneteman et al. (2004) indicated that a gefitinib, clinical results have so far been disappointing. sirolimus-based immunosuppression protocol seems to In a phase II study of 30 patients with unresectable have beneficial effects on tumor recurrence and survival or metastatic HCC, cetuximab (400 mg/m2 initial dose and an acceptable rate of rejection and toxicity in liver followed by 250 mg/m2 given weekly) produced no CRs transplant patients, with extended criteria for HCC. or PRs, and only five patients achieved SD (Zhu et al., Further evaluation is warranted, and both everolimus 2007). Moreover, median PFS was just 1.4 months. and temsirolimus are undergoing early-phase clinical A further phase II study of single-agent cetuximab in development in patients with HCC. 32 patients with advanced HCC (Gruenwald et al., 2007) indicated only limited activity for the drug (44% SD rate and a median TTP of 2 months). Conclusions and future directions Together, these clinical trials have suggested that other than erlotinib, which may have modest activity against Advanced HCC is a notoriously difficult disease to treat. HCC, EGFR inhibitors do not have promising activity in Currently, sorafenib is the only effective systemic HCC. Whether EGFR inhibitors may augment the treatment option available. In recent years, research antitumor activity of other targeted or chemotherapeutic has focused on uncovering the cellular signaling agents remains to be examined in future clinical trials. mechanisms that underlie HCC. Genetic alterations clearly have a major role in hepatocarcinogenesis, and abnormalities in several critical molecular signaling Miscellaneous therapies pathways have been identified as contributing to tumor Activation of the IGF signaling pathway induces development and progression. These pathways include potent proliferative and antiapoptotic effects; molecular the MAPK, PI3K/AKT/mTOR, WNT/b-catenin and components of this pathway may represent valid IGF pathways, and growth factor-regulated angiogenic targets for therapeutic intervention. Indeed, blockade signaling. Evidently, a high dependency on angiogenesis of IGF-1R activity through anti-IGF-1R monoclonal is an exploitable feature for the treatment of HCC. antibodies can induce growth inhibition, apoptosis and Importantly, HCC clearly has a complex pathogen- cell cycle arrest in HCC cells (Hopfner et al., 2006; esis, with the potential for considerable crosstalk and Zhang et al., 2006). IMC-A12, a monoclonal antibody redundancy in signaling pathways, and hence targeting directed at IGF-1R, has been shown to downregulate single molecules or pathways may have a limited benefit IGF-1R levels and inhibit downstream signaling in in treatment. Use of combinations of therapies may be solid tumors (Burtrum et al., 2003; Wu et al., 2005). On needed to overcome the complex network of signaling the basis of these results, clinical trials are currently pathways, and ultimately inhibit the signaling that under way to assess the efficacy of IMC-A12 in liver controls tumor growth and survival. However, use of a cancer (www.clinicaltrials.gov). combination regimen can lead to tolerability and drug– The high proportion of HCC tumors expressing drug interaction problems, and hence an alternative c-MET has also led to the study of agents that may approach is to use molecularly targeted agents that target this molecular pathway. Two small-molecule have multiple modes of action, targeting several signal- inhibitors of c-MET, PHA665752 and SU11274, have ing pathways concurrently. As described above, both shown inhibition of cancer cell growth in vitro (Ma sorafenib and sunitinib are multitargeted agents that et al., 2005a, b), and in preclinical studies, a monoclonal inhibit several kinases. It is unclear whether their clinical antibody to the b-chain of HGF seems to completely efficacy is due to their broad specificity or their targeting inhibit HGF-mediated activities (Burgess et al., 2006). of a single, as-yet-unidentified, protein. However, as Notably, c-MET can be activated downstream of the kinase inhibitors become more and more selective, it will receptor tyrosine kinases in other cancers (Huang et al., be possible to test which targets are important in specific 2007), suggesting that this kinase could be an important diseases. Unfortunately, at present there are no target in many cancers. Further studies of c-MET- approved drugs to effectively target the WNT/b-catenin targeted therapies will be needed to determine whether pathway, which clearly has an important role in the the targeting of MET signaling will have clinical activity pathogenesis of HCC. These drugs are eagerly awaited. in patients with HCC. These issues point to the important role that increased In an in vitro study (Sieghart et al., 2007), the mTOR knowledge of the molecular pathways involved in the inhibitor everolimus (RAD001) showed marked anti- pathogenesis of HCC will have in guiding the develop- proliferative activity when applied as a single agent to ment of effective, rationally designed therapeutic ap- Hep3B and SNU398 HCC cells. The drug also had a proaches. Currently, a plethora of kinase inhibitors and chemosensitizing effect when applied in combination antibodies directed against specific molecular targets are with doxorubicin. In a separate study, the activity of being investigated in HCC. Agents that target angio- another mTOR inhibitor, sirolimus, was investigated in genesis pathways seem to offer the most promise, but a tumor xenograft model of HCC (Semela et al., 2007). further research dissecting the molecular mechanisms

Oncogene HCC signaling review S Whittaker et al 5001 mediating both the clinical benefits and escape or is on the editorial board of Oncogene. Andrew X Zhu resistance is needed. Clearly, the efficacy of these agents has participated in advisory activities and has received must be balanced with toxicity in this highly compro- research support from Bayer HealthCare Pharmaceu- mised patient population. Nevertheless, the approval ticals and Genentech, Inc. and wide use of sorafenib for the treatment of HCC has shown the potential of targeting the molecular pathways involved in HCC, and offers hope that other effective therapies will ultimately be developed. Acknowledgements We would like to thank John D Zoidis, MD (Bayer Conflict of interest HealthCare Pharmaceuticals, Montville, NJ, USA), and Anna Hunt and Catherine Crookes (GeoMed, Macclesfield, Steven Whittaker and Richard Marais have no competing Cheshire, UK) for writing and editorial support, with funding financial interests in relation to this work. Richard Marais from Bayer HealthCare Pharmaceuticals.

References

Abou-Alfa GK, Johnson P, Knox J, Davidenko I, Lacava J, Leung T Bugianesi E. (2005). Review article: steatosis, the metabolic syndrome et al. (2008). Final results from a phase II (PHII), randomized, double- and cancer. Aliment Pharmacol Ther 22(Suppl 2): 40–43. blind study of sorafenib plus doxorubicin (S+D) versus placebo plus Burgess T, Coxon A, Meyer S, Sun J, Rex K, Tsuruda T et al. (2006). doxorubicin (P+D) in patients (pts) with advanced hepatocellular Fully human monoclonal antibodies to hepatocyte growth factor carcinoma (AHCC). ASCO Gastrointestinal Cancers Symposium. with therapeutic potential against hepatocyte growth factor/c-Met- Orlando, FL. dependent human tumors. Cancer Res 66: 1721–1729. Abou-Alfa GK, Schwartz L, Ricci S, Amadori D, Santoro A, Figer A Burtrum D, Zhu Z, Lu D, Anderson DM, Prewett M, Pereira DS et al. et al. (2006). Phase II study of sorafenib in patients with advanced (2003). A fully human monoclonal antibody to the insulin-like hepatocellular carcinoma. J Clin Oncol 24: 4293–4300. growth factor I receptor blocks ligand-dependent signaling Abrams TJ, Lee LB, Murray LJ, Pryer NK, Cherrington JM. (2003a). and inhibits human tumor growth in vivo. Cancer Res 63: SU11248 inhibits KIT and platelet-derived growth factor receptor 8912–8921. beta in preclinical models of human small cell lung cancer. Mol Cariani E, Lasserre C, Seurin D, Hamelin B, Kemeny F, Franco D Cancer Ther 2: 471–478. et al. (1988). Differential expression of insulin-like growth factor II Abrams TJ, Murray LJ, Pesenti E, Holway VW, Colombo T, Lee LB mRNA in human primary liver cancers, benign liver tumors, and et al. (2003b). Preclinical evaluation of the tyrosine kinase inhibitor liver cirrhosis. Cancer Res 48: 6844–6849. SU11248 as a single agent and in combination with ‘standard Carlin CR, Simon D, Mattison J, Knowles BB. (1988). Expression and of care’ therapeutic agents for the treatment of breast cancer. biosynthetic variation of the epidermal growth factor receptor in Mol Cancer Ther 2: 101–1021. human hepatocellular carcinoma-derived cell lines. Mol Cell Biol 8: Albert DH, Tapang P, Magoc TJ, Pease LJ, Reuter DR, Wei RQ et al. 25–34. (2006). Preclinical activity of ABT-869, a multitargeted receptor Carlomagno F, Anaganti S, Guida T, Salvatore G, Troncone G, tyrosine kinase inhibitor. Mol Cancer Ther 5: 995–1006. Wilhelm SM et al. (2006). BAY 43-9006 inhibition of oncogenic Alberts SR, Morlan BW, Kim GP, Piyoy HC, Quevedo FJ, Dakhil SR RET mutants. J Natl Cancer Inst 98: 326–334. et al. (2007). NCCTG phase II trial (N044J) of AZD2171 for Chao Y, Li CP, Chau GY, Chen CP, King KL, Lui WY et al. (2003). patients with hepatocellular carcinoma (HCC)-interim review of Prognostic significance of vascular endothelial growth factor, basic toxicity. ASCO Gastrointestinal Cancers Symposium. Orlando, FL. fibroblast growth factor, and angiogenin in patients with resectable Alexia C, Fallot G, Lasfer M, Schweizer-Groyer G, Groyer A. (2004). hepatocellular carcinoma after surgery. Ann Surg Oncol 10: 355–362. An evaluation of the role of insulin-like growth factors (IGF) and of Chen YL, Law PY, Loh HH. (2005). Inhibition of PI3K/Akt signaling: type-I IGF receptor signalling in hepatocarcinogenesis and in the an emerging paradigm for targeted cancer therapy. Curr Med Chem resistance of hepatocarcinoma cells against drug-induced apoptosis. Anticancer Agents 5: 575–589. Biochem Pharmacol 68: 1003–1015. Cheng AL, Kang YK, Chen Z, Tsao CJ, Qin S, Kim JS et al. (2009). Anzola M. (2004). Hepatocellular carcinoma: role of hepatitis B and Efficacy and safety of sorafenib in patients in the Asia-Pacific region hepatitis C viruses proteins in hepatocarcinogenesis. J Viral Hepat with advanced hepatocellular carcinoma: a phase III randomised, 11: 383–393. double-blind, placebo-controlled trial. Lancet Oncol 10: 25–34. Aoki H, Hayashi J, Moriyama M, Arakawa Y, Hino O. (2000). Chow LQ, Eckhardt SG. (2007). Sunitinib: from rational design to Hepatitis C virus core protein interacts with 14-3-3 protein and clinical efficacy. J Clin Oncol 25: 884–896. activates the kinase Raf-1. J Virol 74: 1736–1741. De La CA, Romagnolo B, Billuart P, Renard CA, Buendia MA, Avila MA, Berasain C, Sangro B, Prieto J. (2006). New therapies for Soubrane O et al. (1998). Somatic mutations of the beta-catenin hepatocellular carcinoma. Oncogene 25: 3866–3884. gene are frequent in mouse and human hepatocellular carcinomas. Bangoura G, Liu ZS, Qian Q, Jiang CQ, Yang GF, Jing S. (2007). Proc Natl Acad Sci USA 95: 8847–8851. Prognostic significance of HIF-2alpha/EPAS1 expression in hepato- De Souza AT, Hankins GR, Washington MK, Fine RL, Orton TC, cellular carcinoma. World J Gastroenterol 13: 3176–3182. Jirtle RL. (1995). Frequent loss of heterozygosity on 6q at the Boix L, Rosa JL, Ventura F, Castells A, Bruix J, Rode´sJet al. (1994). mannose 6-phosphate/insulin-like growth factor II receptor locus in c-MET mRNA overexpression in human hepatocellular carcinoma. human hepatocellular tumors. Oncogene 10: 1725–1729. Hepatology 19: 88–91. Desbois-Mouthon C, Cadoret A, Blivet-Van Eggelpoel MJ, Bertrand Bosch FX, Ribes J, Diaz M, Cleries R. (2004). Primary liver F, Cherqui G, Perret C et al. (2001). Insulin and IGF-1 stimulate the cancer: worldwide incidence and trends. Gastroenterology 127: beta-catenin pathway through two signalling cascades involving S5–S16. GSK-3beta inhibition and Ras activation. Oncogene 20: 252–259. Boyault S, Rickman DS, de RA, Balabaud C, Rebouissou S, Jeannot E Dhar DK, Naora H, Yamanoi A, Ono T, Kohno H, Otani H. (2002). et al. (2007). Transcriptome classification of HCC is related to Requisite role of VEGF receptors in angiogenesis of hepatocellular gene alterations and to new therapeutic targets. Hepatology 45: carcinoma: a comparison with angiopoietin/Tie pathway. Anti- 42–52. cancer Res 22: 379–386.

Oncogene HCC signaling review S Whittaker et al 5002 Dougherty MK, Muller J, Ritt DA, Zhou M, Zhou XZ, Copeland TD Hoshida Y, Nijman SM, Kobayashi M, Chan JA, Brunet JP, Chiang et al. (2005). Regulation of Raf-1 by direct feedback phosphory- DY et al. (2009). Integrative transcriptome analysis reveals common lation. Mol Cell 17: 215–224. molecular subclasses of human hepatocellular carcinoma. Cancer Downward J. (2003). Targeting RAS signalling pathways in cancer Res 69: 7385–7392. therapy. Nat Rev Cancer 3: 11–22. Hsu CH, Yang TS, Hsu C, Toh HC, Epstein RJ, Hsiao LT et al. Eisen T, Ahmad T, Flaherty KT, Gore M, Kaye S, Marais R et al. (2010). Efficacy and tolerability of bevacizumab plus capecitabine as (2006). Sorafenib in advanced melanoma: a phase II randomised first-line therapy in patients with advanced hepatocellular carcinoma. discontinuation trial analysis. Br J Cancer 95: 581–586. Br J Cancer 102: 981–986. El-Assal ON, Yamanoi A, Soda Y, Yamaguchi M, Igarashi M, Hu TH, Huang CC, Lin PR, Chang HW, Ger LP, Lin YW et al. Yamamoto A et al. (1998). Clinical significance of microvessel (2003). Expression and prognostic role of tumor suppressor gene density and vascular endothelial growth factor expression in PTEN/MMAC1/TEP1 in hepatocellular carcinoma. Cancer 97: hepatocellular carcinoma and surrounding liver: possible involve- 1929–1940. ment of vascular endothelial growth factor in the angiogenesis of Huang PH, Mukasa A, Bonavia R, Flynn RA, Brewer ZE, Cavenee cirrhotic liver. Hepatology 27: 1554–1562. WK et al. (2007). Quantitative analysis of EGFRvIII cellular Faivre S, Raymond E, Boucher E, Douillard J, Lim HY, Kim JS et al. signaling networks reveals a combinatorial therapeutic strategy for (2009). Safety and efficacy of sunitinib in patients with advanced glioblastoma. Proc Natl Acad Sci USA 104: 12867–12872. hepatocellular carcinoma: an open-label, multicentre, phase II Huether A, Hopfner M, Sutter AP, Baradari V, Schuppan D, Scherubl study. Lancet Oncol 10: 794–800. H. (2006). Signaling pathways involved in the inhibition of Feitelson MA, Pan J, Lian Z. (2004). Early molecular and genetic epidermal growth factor receptor by erlotinib in hepatocellular determinants of primary liver malignancy. Surg Clin North Am 84: cancer. World J Gastroenterol 12: 5160–5167. 339–354. Huether A, Hopfner M, Sutter AP, Schuppan D, Scherubl H. (2005). Feng DY, Zheng H, Tan Y, Cheng RX. (2001). Effect of Erlotinib induces cell cycle arrest and apoptosis in hepatocellular phosphorylation of MAPK and Stat3 and expression of c-fos and cancer cells and enhances chemosensitivity towards cytostatics. c-jun proteins on hepatocarcinogenesis and their clinical signifi- J Hepatol 43: 661–669. cance. World J Gastroenterol 7: 33–36. Huynh H, Chow PK, Soo KC. (2007a). AZD6244 and doxorubicin Flaherty KT, Puzanov I, Sosman J, Kim K, Ribas A, McArthur G induce growth suppression and apoptosis in mouse models of et al. (2009). Phase I study of PLX4032: proof of concept for V600E hepatocellular carcinoma. Mol Cancer Ther 6: 2468–2476. BRAF mutation as a therapeutic target in human cancer. J Clin Huynh H, Nguyen TT, Chow KH, Tan PH, Soo KC, Tran E. (2003). Oncol 27(Suppl): 461s. Over-expression of the mitogen-activated protein kinase (MAPK) Folkman J. (2003). Fundamental concepts of the angiogenic process. kinase (MEK)-MAPK in hepatocellular carcinoma: its role in tumor Curr Mol Med 3: 643–651. progression and apoptosis. BMC Gastroenterol 3: 19. Giles RH, van Es JH, Clevers H. (2003). Caught up in a Wnt storm: Huynh H, Soo KC, Chow PK, Tran E. (2007b). Targeted inhibition of Wnt signaling in cancer. Biochim Biophys Acta 1653: 1–24. the extracellular signal-regulated kinase kinase pathway with Gollob JA, Wilhelm S, Carter C, Kelley SL. (2006). Role of Raf kinase AZD6244 (ARRY-142886) in the treatment of hepatocellular in cancer: therapeutic potential of targeting the Raf/MEK/ERK carcinoma. Mol Cancer Ther 6: 138–146. signal transduction pathway. Semin Oncol 33: 392–406. Hwang YH, Choi JY, Kim S, Chung ES, Kim T, Koh SS et al. (2004). Greer EL, Brunet A. (2005). FOXO transcription factors at the Over-expression of c-raf-1 proto-oncogene in liver cirrhosis and interface between longevity and tumor suppression. Oncogene 24: hepatocellular carcinoma. Hepatol Res 29: 113–121. 7410–7425. Ito Y, Sasaki Y, Horimoto M, Wada S, Tanaka Y, Kasahara A Gruenwald V, Wilkens V, Gebel M, Greten TF, Kubicka S, Ganser A et al. (1998). Activation of mitogen-activated protein kinases/ et al. (2007). A phase II open-label study of cetuximab in unresectable extracellular signal-regulated kinases in human hepatocellular hepatocellular carcinoma: final results. JClinOncol25(Suppl): 222s. carcinoma. Hepatology 27: 951–958. Grunwald V, Hidalgo M. (2003). Developing inhibitors of the Ito Y, Takeda T, Higashiyama S, Sakon M, Wakasa KI, Tsujimoto M epidermal growth factor receptor for cancer treatment. J Natl et al. (2001). Expression of heparin binding epidermal growth Cancer Inst 95: 851–867. factor-like growth factor in hepatocellular carcinoma: an immuno- Halaschek-Wiener J, Wacheck V, Kloog Y, Jansen B. (2004). Ras histochemical study. Oncol Rep 8: 903–907. inhibition leads to transcriptional activation of p53 and down- Johnston SR, Leary A. (2006). Lapatinib: a novel EGFR/HER2 regulation of Mdm2: two mechanisms that cooperatively increase tyrosine kinase inhibitor for cancer. Drugs Today (Barc) 42: p53 function in colon cancer cells. Cell Signal 16: 1319–1327. 441–453. Harden TK, Sondek J. (2006). Regulation of phospholipase C Kaposi-Novak P, Lee JS, Gomez-Quiroz L, Coulouarn C, Factor VM, isozymes by ras superfamily GTPases. Annu Rev Pharmacol Toxicol Thorgeirsson SS. (2006). Met-regulated expression signature defines 46: 355–379. a subset of human hepatocellular carcinomas with poor prognosis Hatzivassilou G, Song K, Yen I, Brandhuber BJ, Anderson DJ, and aggressive phenotype. J Clin Invest 116: 1582–1595. Alvarado R et al. (2010). RAF inhibitors prime wild-type RAF Kim KJ, Li B, Winer J, Armanini M, Gillett N, Phillips HS et al. to activate the MAPK pathway and enhance growth. Nature 464: (1993). Inhibition of vascular endothelial growth factor-induced 431–435. angiogenesis suppresses tumour growth in vivo. Nature 362: Heidorn SJ, Milagre C, Whittaker S, Nourry A, Niculescu-Duvas I, 841–844. Dhomen N et al. (2010). Kinase-dead BRAF and oncogenic RAS Kira S, Nakanishi T, Suemori S, Kitamoto M, Watanabe Y, cooperate to drive tumor progression through CRAF. Cell 140: Kajiyama G (1997). Expression of transforming growth factor 209–221. alpha and epidermal growth factor receptor in human hepato- Himmelsbach K, Sauter D, Baumert TF, Ludwig L, Blum HE, cellular carcinoma. Liver 17: 177–182. Hildt E. (2009). New aspects of an anti-tumour drug: sorafenib Kiss A, Wang NJ, Xie JP, Thorgeirsson SS. (1997). Analysis of efficiently inhibits HCV replication. Gut 58: 1644–1653. transforming growth factor (TGF)-alpha/epidermal growth factor Hopfner M, Huether A, Sutter AP, Baradari V, Schuppan D, Scherubl receptor, hepatocyte growth factor/c-met, TGF-beta receptor H. (2006). Blockade of IGF-1 receptor tyrosine kinase has type II, and p53 expression in human hepatocellular carcinomas. antineoplastic effects in hepatocellular carcinoma cells. Biochem Clin Cancer Res 3: 1059–1066. Pharmacol 71: 1435–1448. Kneteman NM, Oberholzer J, Al SM, Meeberg GA, Blitz M, Ma MM Hopfner M, Sutter AP, Huether A, Schuppan D, Zeitz M, Scherubl H. et al. (2004). Sirolimus-based immunosuppression for liver trans- (2004). Targeting the epidermal growth factor receptor by gefitinib plantation in the presence of extended criteria for hepatocellular for treatment of hepatocellular carcinoma. J Hepatol 41: 1008–1016. carcinoma. Liver Transp 10: 1301–1311.

Oncogene HCC signaling review S Whittaker et al 5003 Koch I, Baron A, Roberts S, Junker U, Palacay-Romano M, Mason E Mendel DB, Laird AD, Xin X, Louie SG, Christensen JG, Li G et al. et al. (2005). Influence of hepatic dysfunction on safety, tolerability, (2003). In vivo antitumor activity of SU11248, a novel tyrosine and pharmacokinetics (PK) of PTK787/ZK 222584 in patients (Pts) kinase inhibitor targeting vascular endothelial growth factor and with unresectable hepatocellular carcinoma (HCC). J Clin Oncol platelet-derived growth factor receptors: determination of a 23(Suppl): 341s. pharmacokinetic/pharmacodynamic relationship. Clin Cancer Res Kolch W. (2000). Meaningful relationships: the regulation of the 9: 327–337. Ras/Raf/MEK/ERK pathway by protein interactions. Biochem J Mulcahy MF. (2005). Management of hepatocellular cancer. Curr 351(Part 2): 289–305. Treat Options Oncol 6: 423–435. Lee JW, Soung YH, Kim SY, Lee HW, Park WS, Nam SW et al. Murray LJ, Abrams TJ, Long KR, Ngai TJ, Olson LM, Hong W et al. (2005). PIK3CA gene is frequently mutated in breast carcinomas (2003). SU11248 inhibits tumor growth and CSF-1R-dependent and hepatocellular carcinomas. Oncogene 24: 1477–1480. osteolysis in an experimental breast cancer bone metastasis model. Leicht DT, Balan V, Kaplun A, Singh-Gupta V, Kaplun L, Dobson M Clin Exp Metastasis 20: 757–766. et al. (2007). Raf kinases: function, regulation and role in human Nakanishi K, Sakamoto M, Yamasaki S, Todo S, Hirohashi S. (2005). cancer. Biochim Biophys Acta 1773: 1196–1212. Akt phosphorylation is a risk factor for early disease recurrence Li XM, Tang ZY, Zhou G, Lui YK, Ye SL. (1998). Significance of and poor prognosis in hepatocellular carcinoma. Cancer 103: vascular endothelial growth factor mRNA expression in invasion 307–312. and metastasis of hepatocellular carcinoma. J Exp Clin Cancer Res Ng IO, Poon RT, Lee JM, Fan ST, Ng M, Tso WK. (2001). 17: 13–17. Microvessel density, vascular endothelial growth factor and its Lian Z, Liu J, Wu M, Wang HY, Arbuthnot P, Kew M et al. (2007). receptors Flt-1 and Flk-1/KDR in hepatocellular carcinoma. Am J Hepatitis B x antigen up-regulates vascular endothelial growth factor Clin Pathol 116: 838–845. receptor 3 in hepatocarcinogenesis. Hepatology 45: 1390–1399. O’Dwyer PJ, Giantonio BJ, Levy DE, Kauh JS, Fitzgerald DB, Lierman E, Folens C, Stover EH, Mentens N, Van MH, Scheers W Benson III AB. (2006). Gefitinib in advanced unresectable hepato- et al. (2006). Sorafenib is a potent inhibitor of FIP1L1-PDGFRal- cellular carcinoma: Results from the Eastern Cooperative Oncology pha and the imatinib-resistant FIP1L1-PDGFRalpha T674I Group’s Study E1203. J Clin Oncol 24(Suppl): 213s. mutant. Blood 108: 1374–1376. Okano J, Matsumoto K, Nagahara T, Murawaki Y. (2006). Gefitinib Liu Y, Poon RT, Li Q, Kok TW, Lau C, Fan ST. (2005). Both and the modulation of the signaling pathways downstream of antiangiogenesis- and angiogenesis-independent effects are respon- epidermal growth factor receptor in human liver cancer cells. sible for hepatocellular carcinoma growth arrest by tyrosine kinase J Gastroenterol 41: 166–176. inhibitor PTK787/ZK222584. Cancer Res 65: 3691–3699. Pang R, Lee TK, Poon RT, Fan ST, Wong KB, Kwong YL et al. Llovet JM, Fuster J, Bruix J. (1999). Intention-to-treat analysis of (2007). Pin1 interacts with a specific serine-proline motif of surgical treatment for early hepatocellular carcinoma: resection hepatitis B virus X-protein to enhance hepatocarcinogenesis. versus transplantation. Hepatology 30: 1434–1440. Gastroenterology 132: 1088–1103. Llovet JM, Ricci S, Mazzaferro V, Hilgard P, Gane E, Blanc JF et al. Papetti M, Herman IM. (2002). Mechanisms of normal and tumor- (2008). Sorafenib in advanced hepatocellular carcinoma. N Engl J derived angiogenesis. Am J Physiol Cell Physiol 282: C947–C970. Med 359: 378–390. Parkin DM, Bray F, Ferlay J, Pisani P. (2005). Global cancer statistics, Lund P, Schubert D, Niketeghad F, Schirmacher P. (2004). Autocrine 2002. CA Cancer J Clin 55: 74–108. inhibition of chemotherapy response in human liver tumor cells by Philip PA, Mahoney MR, Allmer C, Thomas J, Pitot HC, Kim G et al. insulin-like growth factor-II. Cancer Lett 206: 85–96. (2005). Phase II study of erlotinib (OSI-774) in patients with Ma PC, Jagadeeswaran R, Jagadeesh S, Tretiakova MS, Nallasura V, advanced hepatocellular cancer. J Clin Oncol 23: 6657–6663. Fox EA et al. (2005a). Functional expression and mutations of Pollak MN, Schernhammer ES, Hankinson SE. (2004). Insulin-like c-Met and its therapeutic inhibition with SU11274 and small growth factors and neoplasia. Nat Rev Cancer 4: 505–518. interfering RNA in non-small cell lung cancer. Cancer Res 65: Pons-Renedo F, Llovet JM. (2003). Hepatocellular carcinoma: a 1479–1488. clinical update. Med Gen Med 5: 11. Ma PC, Schaefer E, Christensen JG, Salgia R. (2005b). A selective Ponzetto C, Bardelli A, Zhen Z, Maina F, dalla ZP, Giordano S et al. small molecule c-MET inhibitor, PHA665752, cooperates with (1994). A multifunctional docking site mediates signaling and rapamycin. Clin Cancer Res 11: 2312–2319. transformation by the hepatocyte growth factor/scatter factor Malka D, Dromain C, Farace F, Horn S, Pignon J, Ducreux M et al. receptor family. Cell 77: 261–271. (2007). Bevacizumab in patients (pts) with advanced hepatocellular Poon RT, Ho JW, Tong CS, Lau C, Ng IO, Fan ST. (2004a). carcinoma (HCC): preliminary results of a phase II study with Prognostic significance of serum vascular endothelial growth factor circulating endothelial cell (CEC) monitoring. J Clin Oncol and endostatin in patients with hepatocellular carcinoma. Br J Surg 18(Suppl): 215s. 91: 1354–1360. Marais R, Light Y, Mason C, Paterson H, Olson MF, Marshall CJ. Poon RT, Lau C, Yu WC, Fan ST, Wong J. (2004b). High serum levels (1998). Requirement of Ras-GTP-Raf complexes for activation of of vascular endothelial growth factor predict poor response to Raf-1 by protein kinase C. Science 280: 109–112. transarterial chemoembolization in hepatocellular carcinoma: Marotta F, Vangieri B, Cecere A, Gattoni A. (2004). The pathogenesis a prospective study. Oncol Rep 11: 1077–1084. of hepatocellular carcinoma is multifactorial event. Novel immuno- Poulikakos PI, Zhang C, Bollag G, Shokat KM, Rosen N. (2010). logical treatment in prospect. Clin Ter 155: 187–199. RAF inhibitors transactivate RAF dimers and ERK signalling in Mas VR, Maluf DG, Archer KJ, Yanek KC, Fisher RA. (2007). cells with wild-type BRAF. Nature 464: 427–430. Angiogenesis soluble factors as hepatocellular carcinoma noninva- Presta LG, Chen H, O’Connor SJ, Chisholm V, Meng YG, Krummen sive markers for monitoring hepatitis C virus cirrhotic patients L et al. (1997). Humanization of an anti-vascular endothelial growth awaiting liver transplantation. Transplantation 84: 1262–1271. factor monoclonal antibody for the therapy of solid tumors and Matsumoto K, Nakamura T. (1996). Emerging multipotent aspects other disorders. Cancer Res 57: 4593–4599. of hepatocyte growth factor. J Biochem 119: 591–600. Ramanathan RK, Belani CP, Singh DA, Tanaka M, Lenz HJ, Yen Y Matsuo M, Sakurai H, Saiki I. (2003). ZD1839, a selective epidermal et al. (2006). Phase II study of lapatinib, a dual inhibitor of growth factor receptor tyrosine kinase inhibitor, shows antimeta- epidermal growth factor receptor (EGFR) tyrosine kinase 1 and 2 static activity using a hepatocellular carcinoma model. Mol Cancer (Her2/Neu) in patients (pts) with advanced biliary tree cancer (BTC) Ther 2: 557–561. or hepatocellular cancer (HCC). A California Consortium (CCC-P) McKillop IH, Schmidt CM, Cahill PA, Sitzmann JV. (1997). Altered Trial. J Clin Oncol 24(Suppl): 181s. expression of mitogen-activated protein kinases in a rat model of Ramanathan RK, Belani CP, Singh DA, Tanaka M, Lenz HJ, Yen Y experimental hepatocellular carcinoma. Hepatology 26: 1484–1491. et al. (2009). A phase II study of lapatinib in patients with advanced

Oncogene HCC signaling review S Whittaker et al 5004 biliary tree and hepatocellular cancer. Cancer Chemother Pharmacol is essential for hepatitis B virus gene expression. Oncogene 22: 64: 777–783. 2604–2610. Raoul JL, Finn RS, Kang YK, Park JW, Harris R, Coric V et al. Su MC, Hsu C, Kao HL, Jeng YM. (2006). CD24 expression is a (2009). An open-label phase II study of first- and second-line prognostic factor in intrahepatic cholangiocarcinoma. Cancer Lett treatment with brivanib in patients with hepatocellular carcinoma 235: 34–39. (HCC). J Clin Oncol 27(Suppl): 15s. Sun W, Haller DG, Mykulowycz K, Rosen M, Soulen M, Capparo M Roberts LR, Gores GJ. (2005). Hepatocellular carcinoma: molecular et al. (2007). Combination of capecitabine, oxaliplatin with pathways and new therapeutic targets. Semin Liver Dis 25: 212–225. bevacizumab in treatment of advanced hepatocellular carcinoma Roberts LR, Gores GJ. (2006). Emerging drugs for hepatocellular (HCC): a phase II study. J Clin Oncol 25(Suppl): 18s. carcinoma. Expert Opin Emerg Drugs 11: 469–487. Suzuki K, Hayashi N, Yamada Y, Yoshihara H, Miyamoto Y, Ito Y Sahin F, Kannangai R, Adegbola O, Wang J, Su G, Torbenson M. et al. (1994). Expression of the c-met protooncogene in human (2004). mTOR and P70 S6 kinase expression in primary liver hepatocellular carcinoma. Hepatology 20: 1231–1236. neoplasms. Clin Cancer Res 10: 8421–8425. Tanaka Y, Kanai F, Tada M, Asaoka Y, Guleng B, Jazag A et al. Sananbenesi F, Fischer A, Schrick C, Spiess J, Radulovic J. (2002). (2006). Absence of PIK3CA hotspot mutations in hepatocellular Phosphorylation of hippocampal Erk-1/2, Elk-1, and p90-Rsk-1 carcinoma in Japanese patients. Oncogene 25: 2950–2952. during contextual fear conditioning: interactions between Erk-1/2 Tannapfel A, Sommerer F, Benicke M, Katalinic A, Uhlmann D, and Elk-1. Mol Cell Neurosci 21: 463–476. Witzigmann H et al. (2003). Mutations of the BRAF gene in Scharovsky OG, Rozados VR, Gervasoni SI, Matar P. (2000). cholangiocarcinoma but not in hepatocellular carcinoma. Gut 52: Inhibition of ras oncogene: a novel approach to antineoplastic 706–712. therapy. J Biomed Sci 7: 292–298. Tavian D, De PG, Benetti A, Portolani N, Giulini SM, Barlati S. Schiffer E, Housset C, Cacheux W, Wendum D, Sbois-Mouthon C, (2000). u-PA and c-MET mRNA expression is co-ordinately Rey C et al. (2005). Gefitinib, an EGFR inhibitor, prevents enhanced while hepatocyte growth factor mRNA is down-regulated hepatocellular carcinoma development in the rat liver with cirrhosis. in human hepatocellular carcinoma. Int J Cancer 87: 644–649. Hepatology 41: 307–314. Thomas MB, Abbruzzese JL. (2005). Opportunities for targeted Schirmacher P, Held WA, Yang D, Chisari FV, Rustum Y, Rogler CE. therapies in hepatocellular carcinoma. J Clin Oncol 23: 8093–8108. (1992). Reactivation of insulin-like growth factor II during Thomas MB, Chadha R, Glover K, Wang X, Morris J, Brown T et al. hepatocarcinogenesis in transgenic mice suggests a role in malignant (2007). Phase 2 study of erlotinib in patients with unresectable growth. Cancer Res 52: 2549–2556. hepatocellular carcinoma. Cancer 110: 1059–1067. Schlessinger J. (2000). Cell signaling by receptor tyrosine kinases. Thomas MB, Morris JS, Chadha R, Iwasaki M, Kaur H, Lin E et al. Cell 103: 211–225. (2009). Phase II trial of the combination of bevacizumab and Schmidt VA, Chiariello CS, Capilla E, Miller F, Bahou WF. (2008). erlotinib in patients who have advanced hepatocellular carcinoma. Development of hepatocellular carcinoma in Iqgap2-deficient mice J Clin Oncol 27: 843–850. is IQGAP1 dependent. Mol Cell Biol 28: 1489–1502. Thorgeirsson SS, Grisham JW. (2002). Molecular pathogenesis of Schmitz KJ, Wohlschlaeger J, Lang H, Sotiropoulos GC, Malago M, human hepatocellular carcinoma. Nat Genet 31: 339–346. Steveling K et al. (2008). Activation of the ERK and AKT signalling To MD, Perez-Losada J, Mao JH, Balmain A. (2005). Crosstalk pathway predicts poor prognosis in hepatocellular carcinoma and between PTEN and Ras signaling pathways in tumor development. ERK activation in cancer tissue is associated with hepatitis C virus Cell Cycle 4: 1185–1188. infection. J Hepatol 48: 83–90. Toh H, Chen P, Carr BI, Knox JJ, Gill S, Steinberg J et al. (2009). Schuierer MM, Bataille F, Weiss TS, Hellerbrand C, Bosserhoff AK. A phase II study of ABT-869 in hepatocellular carcinoma (HCC): (2006). Raf kinase inhibitor protein is downregulated in hepato- interim analysis. J Clin Oncol 27(Suppl): 15s. cellular carcinoma. Oncol Rep 16: 451–456. Ueda S, Basaki Y, Yoshie M, Ogawa K, Sakisaka S, Kuwano M et al. Schulze A, Lehmann K, Jefferies HB, McMahon M, Downward J. (2006). PTEN/Akt signaling through epidermal growth factor (2001). Analysis of the transcriptional program induced by Raf in receptor is prerequisite for angiogenesis by hepatocellular carcino- epithelial cells. Genes Dev 15: 981–994. ma cells that is susceptible to inhibition by gefitinib. Cancer Res Schwartz GK, Robertson S, Shen A, Wang E, Pace L, Dials H et al. 66: 5346–5353. (2009). A phase I study of XL281, a selective oral RAF kinase inhibitor, Ueki T, Fujimoto J, Suzuki T, Yamamoto H, Okamoto E. (1997). in patients (Pts) with advanced solid tumors. J Clin Oncol 27(Suppl): 15s. Expression of hepatocyte growth factor and its receptor, the c-met Semela D, Dufour JF. (2004). Angiogenesis and hepatocellular proto-oncogene, in hepatocellular carcinoma. Hepatology 25: 619–623. carcinoma. J Hepatol 41: 864–880. Villanueva A, Chiang DY, Newell P, Peix J, Thung S, Alsinet C et al. Semela D, Piguet AC, Kolev M, Schmitter K, Hlushchuk R, Djonov V (2008). Pivotal role of mTOR signaling in hepatocellular carcinoma. et al. (2007). Vascular remodeling and antitumoral effects of mTOR Gastroenterology 135: 1972–1983. inhibition in a rat model of hepatocellular carcinoma. J Hepatol 46: Villanueva A, Newell P, Chiang DY, Friedman SL, Llovet JM. (2007). 840–848. Genomics and signaling pathways in hepatocellular carcinoma. Shimamura T, Saito S, Morita K, Kitamura T, Morimoto M, Kiba T Semin Liver Dis 27: 55–76. et al. (2000). Detection of vascular endothelial growth factor and its Wang XW, Hussain SP, Huo TI, Wu CG, Forgues M, Hofseth LJ receptor expression in human hepatocellular carcinoma biopsy et al. (2002). Molecular pathogenesis of human hepatocellular specimens. J Gastroenterol Hepatol 15: 640–646. carcinoma. Toxicology 181–182: 43–47. Siegel AB, Cohen EI, Ocean A, Lehrer D, Goldenberg A, Knox JJ Watanabe S, Horie Y, Suzuki A. (2005). Hepatocyte-specific PTEN- et al. (2008). Phase II trial evaluating the clinical and biologic effects deficient mice as a novel model for nonalcoholic steatohepatitis of bevacizumab in unresectable hepatocellular carcinoma. J Clin and hepatocellular carcinoma. Hepatol Res 33: 161–166. Oncol 26: 2992–2998. Wedge SR, Kendrew J, Hennequin LF, Valentine PJ, Barry ST, Sieghart W, Fuereder T, Schmid K, Cejka D, Werzowa J, Wrba F Brave SR et al. (2005). AZD2171: a highly potent, orally bioavailable, et al. (2007). Mammalian target of rapamycin pathway activity in vascular endothelial growth factor receptor-2 tyrosine kinase hepatocellular carcinomas of patients undergoing liver transplan- inhibitor for the treatment of cancer. Cancer Res 65: 4389–4400. tation. Transplantation 83: 425–432. Wellcome Trust Genome Campus, Sanger Institute (2008). Catalogue Simonetti RG, Liberati A, Angiolini C, Pagliaro L. (1997). Treatment Of Somatic Mutations In Cancer (COSMIC) database http://www. of hepatocellular carcinoma: a systematic review of randomized sanger.ac.uk/genetics/CGP/cosmic/. Accessed 4 February 2010. controlled trials. Ann Oncol 8: 117–136. Whittaker S, Kirk R, Hayward R, Zambon A, Viros A, Cantarino N Stockl L, Berting A, Malkowski B, Foerste R, Hofschneider PH, Hildt et al. (2010). Gatekeeper mutations mediate resistance to BRAF- E. (2003). Integrity of c-Raf-1/MEK signal transduction cascade targeted therapies. Sci Transl Med 2: 35ra41.

Oncogene HCC signaling review S Whittaker et al 5005 Wilhelm S, Carter C, Lynch M, Lowinger T, Dumas J, Smith RA et al. relationship to the epidermal growth factor and alpha-fetoprotein (2006). Discovery and development of sorafenib: a multikinase levels in patients with hepatocellular carcinoma. Cancer Res 47: inhibitor for treating cancer. Nat Rev Drug Discov 5: 835–844. 896–901. Wilhelm SM, Carter C, Tang L, Wilkie D, McNabola A, Rong H et al. Yoshida T, Hisamoto T, Akiba J, Koga H, Nakamura K, Tokunaga Y (2004). BAY 43-9006 exhibits broad spectrum oral antitumor et al. (2006). Spreds, inhibitors of the Ras/ERK signal transduction, activity and targets the RAF/MEK/ERK pathway and receptor are dysregulated in human hepatocellular carcinoma and linked to tyrosine kinases involved in tumor progression and angiogenesis. the malignant phenotype of tumors. Oncogene 25: 6056–6066. Cancer Res 64: 7099–7109. Zhang YC, Wang XP, Zhang LY, Song AL, Kou ZM, Li XS. (2006). Wong CM, Fan ST, Ng IO. (2001). Beta-catenin mutation and Effect of blocking IGF-I receptor on growth of human hepato- overexpression in hepatocellular carcinoma: clinicopathologic and cellular carcinoma cells. World J Gastroenterol 12: 3977–3982. prognostic significance. Cancer 92: 136–145. Zhou J, Tang ZY, Fan J, Wu ZQ, Li XM, Liu YK et al. (2000). Wood JM, Bold G, Buchdunger E, Cozens R, Ferrari S, Frei J et al. Expression of platelet-derived endothelial cell growth factor (2000). PTK787/ZK 222584, a novel and potent inhibitor of and vascular endothelial growth factor in hepatocellular carcinoma vascular endothelial growth factor receptor tyrosine kinases, and portal vein tumor thrombus. J Cancer Res Clin Oncol 126: impairs vascular endothelial growth factor-induced responses and 57–61. tumor growth after oral administration. Cancer Res 60: 2178–2189. Zhu AX. (2006). Systemic therapy of advanced hepatocellular Wu JD, Odman A, Higgins LM, Haugk K, Vessella R, Ludwig DL carcinoma: how hopeful should we be? Oncologist 11: 790–800. et al. (2005). In vivo effects of the human type I insulin-like Zhu AX. (2008). Development of sorafenib and other molecularly growth factor receptor antibody A12 on androgen-dependent targeted agents in hepatocellular carcinoma. Cancer 112: 250–259. and androgen-independent xenograft human prostate tumors. Zhu AX, Blaszkowsky LS, Ryan DP, Clark JW, Muzikansky A, Clin Cancer Res 11: 3065–3074. Horgan K et al. (2006). Phase II study of gemcitabine and Xaus J, Comalada M, Valledor AF, Cardo M, Herrero C, Soler C oxaliplatin in combination with bevacizumab in patients with et al. (2001). Molecular mechanisms involved in macrophage advanced hepatocellular carcinoma. J Clin Oncol 24: 1898–1903. survival, proliferation, activation or apoptosis. Immunobiology Zhu AX, Sahani DV, Duda DG, di TE, Ancukiewicz M, Catalano OA 204: 543–550. et al. (2009). Efficacy, safety, and potential biomarkers of sunitinib Yamaguchi R, Yano H, Iemura A, Ogasawara S, Haramaki M, monotherapy in advanced hepatocellular carcinoma: a phase II Kojiro M. (1998). Expression of vascular endothelial growth factor study. J Clin Oncol 27: 3027–3035. in human hepatocellular carcinoma. Hepatology 28: 68–77. Zhu AX, Stuart K, Blaszkowsky LS, Muzikansky A, Reitberg DP, Yeh YC, Tsai JF, Chuang LY, Yeh HW, Tsai JH, Florine DL et al. Clark JW et al. (2007). Phase 2 study of cetuximab in patients with (1987). Elevation of transforming growth factor alpha and its advanced hepatocellular carcinoma. Cancer 110: 581–589.

Oncogene