genetic landscape and biomarkers of hepatocellular …genetic landscape and biomarkers of...

Gastroenterology 2015;149:1226–1239

GENETICSAND

GENETICTESTING

Genetic Landscape and Biomarkers of Hepatocellular Carcinoma

Jessica Zucman-Rossi1,2,3,4 Augusto Villanueva5,6 Jean-Charles Nault1,2,7 Josep M. Llovet5,8,9

1Inserm, UMR-1162, Génomique Fonctionnelle des Tumeurs Solides, Equipe Labellisée Ligue Contre le Cancer, InstitutUniversitaire d’Hematologie, Paris, France; 2Université Paris Descartes, Labex Immuno-Oncology, Sorbonne Paris Cité,Faculté de Médecine, Paris, France; 3Université Paris 13, Sorbonne Paris Cité, Unité de Formation et de Recherche Santé,Médecine, Biologie Humaine, Bobigny, France; 4Université Paris Diderot, Paris; 5Liver Cancer Program, Division of LiverDiseases, Department of Medicine, Tisch Cancer Institute, Icahn School of Medicine at Mount Sinai, New York, New York;6Division of Hematology and Medical Oncology, Department of Medicine, Icahn School of Medicine at Mount Sinai, NewYork, New York; 7Service d’hépatologie, Hôpital Jean Verdier, Hôpitaux Universitaires Paris-Seine-Saint-Denis,Assistance-Publique Hôpitaux de Paris, Bondy, France; 8Liver Cancer Translational Research Laboratory, Barcelona-ClínicLiver Cancer Group, Institut d’Investigacions Biomèdiques August Pi i Sunyer (IDIBAPS), Liver Unit, CIBEREHD,Hospital Clínic, Barcelona, Catalonia, Spain; 9Institució Catalana de Recerca i Estudis Avançats (ICREA), Barcelona,Catalonia, Spain

Abbreviations used in this paper: AFB1, aflatoxin B1; EGF, epidermalgrowth factor; HBV, hepatitis B virus; HCC, hepatocellular carcinoma;HCV, hepatitis C virus; HGDN, high-grade dysplastic nodules; LGDN, low-

Hepatocellular carcinoma (HCC) has emerged as amajor cause of cancer-related death. Its mortality hasincreased in Western populations, with a minority ofpatients diagnosed at early stages, when curativetreatments are feasible. Only the multikinase inhibitorsorafenib is available for the management of advancedcases. During the last 10 years, there has been a cleardelineation of the landscape of genetic alterations inHCC, including high-level DNA amplifications in chro-mosome 6p21 (VEGFA) and 11q13 (FGF19/CNND1), aswell as homozygous deletions in chromosome 9(CDKN2A). The most frequent mutations affect TERTpromoter (60%), associated with an increased telo-merase expression. TERT promoter can also be affectedby copy number variations and hepatitis B DNA in-sertions, and it can be found mutated in preneoplasticlesions. TP53 and CTNNB1 are the next most prevalentmutations, affecting 25%L30% of HCC patients, that,in addition to low-frequency mutated genes (eg, AXIN1,ARID2, ARID1A, TSC1/TSC2, RPS6KA3, KEAP1, MLL2),help define some of the core deregulated pathways inHCC. Conceptually, some of these changes behave asprototypic oncogenic addiction loops, being ideal bio-markers for specific therapeutic approaches. Data fromgenomic profiling enabled a proposal of HCC in 2 ma-jor molecular clusters (proliferation and nonprolifera-tion), with differential enrichment in prognosticsignatures, pathway activation and tumor phenotype.Translation of these discoveries into specific thera-peutic decisions is an unmeet medical need in thisfield.

grade dysplastic nodules; SNP, single nucleotide polymorphism.

Most current article

© 2015 by the AGA Institute0016-5085/$36.00

http://dx.doi.org/10.1053/j.gastro.2015.05.061

Keywords: Liver Cancer; Genomics; Signaling Pathways;Molecular Therapies.

ur understanding of the molecular pathogenesis of
Ohepatocellular carcinoma (HCC) has improvedsignificantly in the last decade. Certainly, cumulative datafrom high-throughput analyses of large number of sampleshave provided an accurate landscape of HCC genetic alter-ations. This has enabled us to delineate some of the keyevents that might dominate tumor development and pro-gression. Hopefully, translation of this knowledge into newtargets and biomarkers might impact HCC decision making,and ultimately improve patient’s outcomes. There are,however, some challenges ahead. First, the most prevalentoncogenic mutations are currently undruggable. Second, inthose cases where a molecular subclass or a dominantpathway can be identified, there have been minimal effortsto translating this knowledge into clinical trials. Third, thedebate on the role of tumor heterogeneity and its assess-ment in advanced stages has fueled some skepticism onthe future of single-biopsy stratified treatments. Finally, ourability to manipulate the immune system in the therapeuticarsenal against HCC is promising but remains to beevaluated. This review will provide an overview of the ge-netic changes involved in HCC development and progres-sion, as well as discuss the role of molecular markers tostratify patients based on their prognosis and response totherapies.
http://crossmark.crossref.org/dialog/?doi=10.1053/j.gastro.2015.05.061&domain=pdf


October 2015 Genetic Landscape in HCC 1227

GENE

TICS

AND

GENE

TICTESTING

Genetic Predisposition, EnvironmentalFactors, and Mechanism of MalignantTransformation in HepatocellularCarcinomaGene�Environment Interaction in HepatocellularCarcinoma Predisposition

HCC occurrence results from a complex interplay amonggenetic and nongenetic host factors, exposure to environ-mental carcinogens and virus, and development of an un-derlying chronic liver disease, which, at its ultimate stage(ie, cirrhosis), becomes a certain procarcinogenic field.Although cirrhosis is the “soil” where most of HCC grow, itsdevelopment on noncirrhotic liver helps us to reappraisethe risk factors and mechanisms that lead to HCC develop-ment without background liver damage (Figure 1).

Mendelian genetic predisposition to hepatocellu-lar carcinoma. The liver is a very rare target of classicalcancer predisposition syndromes. The most frequentmonogenic syndromes that predispose to breast, ovarian, orcolorectal cancers are usually not associated with develop-ment of liver tumors.1 Exceptionally, HCC can develop inpatients with APC germ-line mutations.2 However, HCCpredispositions are observed in several genetic metabolicdiseases, mainly via development of cirrhosis. These raremonogenic diseases include iron (hemochromatosis, HFE1gene) or copper (Wilson disease, ATP7B gene) overload,tyrosinemia type 1 (FAH gene), porphyria acute intermittent(HMBS gene) or cutanea tarda (UROD gene), and the a1antitrypsin deficiency (SERPINA1 gene).1 In addition, ge-netic alterations of glucose metabolism leading to glycogenstorage diseases (particularly type Ia or von Gierke disease,G6PC gene) or to specific maturity onset diabetes of theyoung type 3 (MODY3, HNF1A gene) can promote geneticliver adenomatosis occurrence and, in a second step, theirrare malignant transformation to HCC without cirrhosis.3–5

Multifactorial genetic predisposition to hepato-cellular carcinoma. Several single nucleotide poly-morphisms (SNPs) were identified to be associated withHCC risk.6 Among them, many polymorphisms alter biologicpathways of carcinogenesis, including inflammation viaTNFA, IL1B, or TGFB7–9; oxidative stress through SOD2 orMPO10,11; iron metabolism with a cooperation betweenalcohol intake and the HFE1 C282Y mutant variant12,13;DNA repair (MTHFR or XRCC3)14,15; cell cycle with a majorrole of MDM216 and TP5317; or growth factor with EGF.18

Polymorphisms modulate HCC risk at different steps ofthe disease, including predisposition to risk factors (eg, viralhepatitis, alcohol intake, or obesity), to the severity of thechronic liver disease and its evolution to cirrhosis, or to themalignant transformation and tumor progression.6 Theycould be used to stratify patients in personalized sur-veillance policies, and act as candidate targets for chemo-prevention strategies.19 As an example, rs4444903polymorphism in the 50 untranslated region that activatesEGF was associated with HCC on hepatitis C virus (HCV)-related cirrhosis and validated in independent studies.18

Epidermal growth factor (EGF) signaling is also involved

in HCC development in mice and rat models that can beprevented by tyrosine kinase inhibitors targeting EGF re-ceptor, such as gefitinib.20 In addition, EGF is among the top-ranked genes of a signature highly correlated with HCCdevelopment in HCV cirrhosis.21 An early clinical chemo-prevention trial is currently evaluating whether erlotinib isable to revert this high-risk HCC signature into a low risk(NCT02273362). To date, no drug, including sorafenib, hasbeen able to decrease HCC incidence in high-risk patients, orprevent tumor recurrence after curative treatments.22

Most of the polymorphisms associated with HCC devel-opment on chronic liver disease are related to specific riskfactors. This observation highlights the close relationshipsamong the nature of the exposure, the genetic background,and the mechanism of hepatocyte transformation/prolifera-tion.23 The association between aflatoxin B1 (AFB1), hepatitisB virus (HBV), and SNP of GTSM1 and GSTT1 is the prototypeof a strong interplay between specific exposure to genotoxiccontaminant together with viral infection and common poly-morphisms to dramatically increase HCC risk.24,25 A PNPLA3polymorphism (coding for a lipase that mediates tri-acylglycerol hydrolysis) is strongly associated with fatty andalcoholic chronic liver diseases; it is also associated with HCCoccurrence.25–27 In contrast, contribution of PNPLA3 poly-morphism seems only minor in the risk of HCV-HCC.27–29

Recently, genome-wide association studies allowed theanalysis of thousands of SNPs in HCC patients comparedwith controls. A genome-wide association study performedin HCV patients in Japan identified a polymorphism in MICA(gene involved in immune regulation),30 and another studyin the same population identified a different polymorphismin DEPDC5 (gene of unknown function),31 associated withHCC development. In Chinese patients infected by HBV,SNPs were identified in STAT4 (a key protein of the in-flammatory pathway),32 TPTE2 (a homolog of PTEN),33

DCL1 (a tumor suppressor gene implicated in HCC patho-genesis),34 and in a region containing the UBEB4, KIF1B,PGD genes35 in 4 different studies as being associated withHCC occurrence. All these polymorphisms require validationin patients with various etiologies and at different stages ofthe liver diseases.6

Early Genetic Alterations inPrecancerous Lesions Involvedin Malignant TransformationCirrhosis as a Cancer Field

There is a key event during malignant transformation incirrhosis that involves damaged cells, possibly hepatocytes,surrounded by fibrosis and vascularized mainly by theportal system switching to highly proliferative cells, vascu-larized by arterial neovessels with an incremental invasiveand metastatic potential.36 This is a multistep processdefined by a precise sequence of lesions: cirrhosis / low-grade dysplastic nodules (LGDN) / high-grade dysplasticnodules (HGDN) / early HCC / progressed HCC andadvanced HCC.37 Several lines of evidence underlined the

Figure 1. Interaction be-tween genetic predisposi-tion, environmental factors,and HCC occurrence.Red, Mendelian inheri-tance pattern, blue, poly-morphisms involved inHCC occurrence, green,HCC genetic imprint ofgenotoxic exposures, andpurple, the potential thera-pies used to reduce HCCincidence.

1228 Zucman-Rossi et al Gastroenterology Vol. 149, No. 5

GENETICSAND

GENETICTESTING

key role of telomere and telomerase in cirrhosis pathogen-esis and tumor initiation.38 Telomeres are repeat sequencesof DNA essential to protect chromosome ends where theyare located.39 They shorten during cell division leading tocell senescence and apoptosis.36 Telomerase is required fortelomere synthesis and is composed of TERC, the RNAtemplate, and of TERT, the enzyme at rate-limiting compo-nent of the complex.40 In human livers, telomerase is notexpressed in mature hepatocytes; cirrhotic tissue exhibitedtelomere shortening with replicative senescence andgermline TERT mutations, predicted to decrease telomeraseactivity, are associated with a higher risk of cirrhosis.39,41

Conversely, telomerase-deficient mouse with experimentalliver injury develop cirrhosis, showing that telomereshortening fosters hepatocytes senescence.42 In contrast,this model enlightened that telomerase reactivation isrequired in a second step to ensure HCC development.43

In humans, telomerase is reactivated in >90% of HCC45

due to somatic TERT promoter mutations (54%�60%),44,45

TERT amplification (5%�6%),45 or HBV insertion in theTERT promoter (10%�15%)46,47 (Figure 2). Interestingly,premalignant lesions exhibit TERT promoter mutations in6% of LGDN and 19% of HGDN, and frequency of TERTpromoter mutations dramatically increases in early HCC

Figure 2.Mechanisms ofmalignant transformationin HCC.


GENE

TICS

AND

GENE

TICTESTING

(61%) and remains stable in progressed and advancedHCC.48 In addition, whole-exome sequencing of early HCCand of premalignant lesions revealed no additional recur-rent mutations in classical HCC driver genes.49 These resultsshow that TERT promoter mutation is the earliest recurrentsomatic genetic alteration, behaving as a “gatekeeper” dur-ing the transformation sequence. It also suggests that earlyHCC are monoclonal, genetically closer to LGDN and HGDNharboring TERT promoter mutations compared with pro-gressed HCC that harbored mutations in many more cancergenes. Concordantly, there is a phenotypic proximity be-tween these lesions, what frequently renders pathologicdiscrimination between HGDN and early HCC a difficult task,even among expert liver pathologists.50 This paradigm

suggests that early HCC, LGDN, and HGDN harboring TERTpromoter mutations are at high risk for full malignanttransformation/progression to advance HCC, which alsorequires additional hits in other cancer genes.

Malignant Transformation of HepatocellularAdenoma

In normal liver, HCC can sometimes arise from the ma-lignant transformation of hepatocellular adenoma, a raremonoclonal benign proliferation of hepatocytes usuallyobserved in young women taking oral contraception.51 Ade-nomas are classified in 4 molecular subgroups according tomutations in either HNF1A, CTNNB1 coding for b-catenin, or


GENETICSAND

GENETICTESTING

in the genes activating the inflammatory pathway as IL6ST,FRK, STAT3, JAK1, and GNAS.51–53 Hepatocellular adenomawith exon 3 b-catenin�activating mutations have a higherrisk of transformation, and TERT promoter mutation areinvolved at the last step of malignant transformationconcomitantly with increased global hypomethylation, andchromosomal aberrations.54,55 However, the sequence ofevents differs in hepatocellular adenoma compared withcirrhosis: in normal hepatocytes, CTNNB1 activatingmutation is first and associated with monoclonal benignproliferation at risk of transformation, and in cirrhotichepatocytes, TERT promoter mutation occurs earlier.48,55

Likely, cirrhotic hepatocytes might require early telomerasereactivation to proliferate and escape from senescence.

Genotoxic Effect as a Cause of HepatocellularCarcinoma Development

HCC related to HBV can occur as a consequence of viralproteins, such as HBxwith oncogenic properties or due to theviral genome itself with HBV insertional mutagenesis incancer genes like TERT, CCNE1, andMLL4.23,46,56,57 Recently,the adeno-associated virus type 2 (AAV2), a DNA defectivevirus, was associated with HCC development on normal liverdue to insertional mutagenesis in TERT, CCNA2, CCNE1,TNFSF10 and MLL4.58 Another layer of complexity emergedfrom the analysis of the cancer genome that carries theimprint of exposure to environmental carcinogens during apatient’s lifetime.59 Exposure to AFB1, a mycotoxin contam-inating food in certain areas of Asia and Africa, is a well-known HCC carcinogen, a phenomena fostered by concomi-tant chronic HBV infection. AFB1-related HCC have a specificmutational signature due to AFB1-DNA adducts at guanineleading to a high rate of C>Amutations and a specific hot spotof mutations R249S in TP53.23 In addition, null genotypes ofGTSM1 and GSTT1, 2 genes important for carcinogen detox-ification that belong to glutathione S transferase family, areassociated with an increased risk of HCC development inAsian patients in cooperation with AFB1 exposure andchronic HBV infection.24 Analysis of the spectrum of nucleo-tide mutations in HCC enabled to identify signatures over thetumor genome specific of the AFB1 exposure.49

Exposure to aristolochic acid, contained in a plant usedin Chinese traditional medicine, is associated with urothelialcancer development that bears a highly specific high rate ofmutations with T>A transversion.60 Interestingly, the samemutational signature was identified in a subset of HCC inChina, suggesting that aristolochic acid could be primarilyinvolved in HCC pathogenesis in a subset of patients.60 Inaddition, next-generation sequencing identified a group ofHCCs developed on noncirrhotic liver in France with anoverrepresentation of T>C at ApTpX with transcriptionstrand bias, a pattern known to be strongly associated withgenotoxic injury.49 Interestingly, these tumors developed innoncirrhotic patients with high alcohol and tobaccoconsumption.49

All of these results showed that the analysis of thecancer genome is an identity card bearing the archeologicalfeatures of exposure to carcinogens or perturbance of DNA

maintenance. This could be useful to better understand thebiologic events that shape HCC development and helpidentify new risk factors using a molecular-epidemiologicapproach.61

Genomic Landscape and Driver Pathwaysin Liver Carcinogenesis

Each HCC genome is the result of a unique combinationof somatic genetic alterations with a mean numbers ofmutations in coding regions varying from 35 to 80 pertumor.45,49,62–66 If most of the genetic alterations occurredin passenger genes without any predicted functional carci-nogenic consequences, then a small numbers of mutationsoccurred in cancer driver genes that belong to key signalingpathways involved in liver carcinogenesis.67 An integratedview of recent whole-exome sequencing studies allowsidentifying the major pathways recurrently mutated in HCC(Figure 3, and Table 1).

Telomere maintenance. As telomerase reactivation iskey during malignant transformation, 90% of human HCCsharbor an increased telomerase expression. The mecha-nisms of telomerase reactivation are mutually exclusive andinclude TERT promoter mutations (54%�60%),44 TERTamplification (5%�6%),45 and HBV insertion in TERT pro-moter (10%�15%).46,47 TERT promoter mutations arefrequently associated with CTNNB1 mutations, suggestingcooperation between telomerase maintenance and b-cateninpathway in liver tumorigenesis.44,45,49 Other mechanisms oftelomerase re-expression remain to be discovered and therole of alternative pathways leading to telomere synthesis,such as alternative lengthening of telomeres, need to bespecifically evaluated.

WNT/b-catenin pathway. The WNT/b-catenin path-way is pivotal in physiologic embryogenesis, zonation, andmetabolic control in the liver. It is the oncogenic pathwaymost frequently activated in HCC by activating mutations ofCTNNB1 (11%�37%),68 and inactivating mutations ofAXIN1 (5%�15%),69 or APC (1%�2%). CTNNB1 mutations,coding for b-catenin, are substitutions or in-frame deletionsin a hotspot located in the domain targeted by the APC/AXIN1/GSK3B inhibitory complex.68 Tumors mutated forCTNNB1 have also a specific transcriptomic profile withoverexpression of classical target genes like GLUL andLGR5,70 and a specific histologic pattern with intratumorcholestasis.71,72

P53 cell cycle pathway. The P53 cell cycle pathway isaltered in at least half of HCC patients with frequent TP53mutations (12%�48%),23,49,62,73,74 the tumor suppressorgene more frequently mutated in cancer. Except for theR249S mutation related to AFB1 exposure, no other recur-rent TP53 mutation hotspot has been identified.62,75 Theretinoblastoma pathway that control progression from G1 toS phase of the cell cycle is inactivated in HCC mainly byhomozygous deletion of CDKN2A (2%�12%)45,62 or RB1mutations (3%�8%).65 Interestingly, genetic alterations ofCDKN2A and RB1 were enriched in tumors with poorprognosis suggesting a role of P21 pathway inactivation intumor aggressiveness.49,65 Finally, recurrent HBV insertionsin CCNE1 (cyclin E1, 5%)46 and amplification of the CCND1/

Figure 3. The geneticlandscape of HCC. Red,activating mutations ofoncogenes, blue, inacti-vating mutations of tumorsuppressors. *Gene wasrecurrently targeted byHBV integration.


GENE

TICS

AND

GENE

TICTESTING

FGF19 locus (5%�14%),76,77 two key proteins involved incell cycle progression, have been reported in HCC.

Epigenetic modifiers. Epigenetic modifiers arerecurrently altered in HCC,49 with inactivating mutations ofARID1A (4%�17%)62 and ARID2 (3%�18%)78 underliningthe key role of SWI/SNF chromatin remodeling complexes(BAF and PBAF) as tumor suppressors. The physiologic roleof these complexes is to modify chromatin structure andnucleosome position. Indirectly, they modify transcriptionfate of the cell. Recurrent somatic alterations in the histone

methylation writer family mainly in MLL (3%�4%), MLL2(2%�3%), MLL3 (3%�6%), and MLL4 (2%�3%) genes bymutation or HBV insertions in MLL4 (10%) are alsofrequent in HCC.46,49,64 In physiologic state, these genescode for proteins that modify histone methylation by addingand removing H3K4 methyl. Altogether, the functionalconsequences of ARID1A, ARID2, and MLL genes mutationsin hepatocarcinogenesis remain to be further explored.

Oxidative stress pathway. The oxidative stresspathway is altered by activating mutations of NRF2 (coded

Table 1.Candidate Drivers as Targets for Therapies in Hepatocellular Carcinoma

Candidateoncogene

addiction loop

Somaticaberrationin human

samples, % Altered pathway

Experimental evidence of “driver”properties

Drug (clinical trialsin any tumor type)

GEM impactliver tumor

development /progression

In vivo tumorresponse after

selective induction/blockade

DNA amplificationsFGF19 5�14a AKT/MTOR; RAS/MAPK Yes Yes BLU9931CCND1 5�14a Cell cycle Palbociclibb

VEGFA 7�11a AKT/MTOR; RAS/MAPK;angiogenesis

NA Yes Bevacizumab, aflibercept

TERT 5�6a Telomere maintenance Yes Yes GX301ImtelstatGV1001

MYC 4 Cell cycle Yes Yes DCR-MYC (RNAi)FAK 4 RAS/MAPK Yes Yes GSK2256098

Gene mutationsTERT promoter 54�60 Telomere maintenance Yes Yes GX301

ImtelstatGV1001

CTNNB1 11�37 WNT pathway Yes Yes PRI-724TP53 12�48 Cell cycle Yes Yes NAARID1Ac 4�17 Epigenetic modifiers NA No GSK126ARID2 3�18 Epigenetic modifiers NA No NAAXIN1 5�15 WNT pathway NA No NATSC1/TSC2 3�8 AKT/MTOR Yes Yes Everolimus,c Temsirolimus,c

Sirolimusc

NFE2L2 3�6 Oxidative stress Yes Yes HSP90 inhibitors (17-AAG,17-DMAG)

RPS6KA3 2�9 RAS/MAPK No No RefametinibATM 2�6 Cell cycle Yes No NAKEAP1 2�8 Oxidative stress Yes No HSP90 inhibitors (17-AAG,

17-DMAG)RB1 3�8 Cell cycle Yes Yes NANRAS/HRAS/BRAF 2 RAS/MAPK Yes Yes Refametinib

OverexpressionMET 30�50d AKT/MTOR; RAS/MAPK Yes Yes Cabozantinib, tivantinibe

IGF2 10d AKT/MTOR; RAS/MAPK Yes Yes IMC-A12,f BI 836845PD1/PD-L1 NA Immune checkpoint NA Yes Nivolumab

NA, not available.aHigh-level DNA focal gains.bCDK4/6 inhibitor.cTarget is MTOR (not TSC1/TSC2).dFrequency estimation not based on high-throughput RNA sequencing studies.eData suggest that tivantinib might not be a specific MET inhibitor.146fTarget is IGF-IR.


GENETICSAND

GENETICTESTING

by NFE2L2) or inactivating KEAP1 in 5%�15% of the cases,preventing proteasome degradation of NRF2 physiologicallyinduced by KEAP1/CUL3 complex ubiquitinylation.62,79

NRF2 pathway activation was previously shown to protectmice against chronic oxidative stress and tumor initiation.79

In contrast, recurrent mutations identified in HCC revealedNRF2 activation as a driver event in tumor progression.62,64

In vitro study has demonstrated that NRF2 activation pre-serves tumor cells from toxic exposure to reactive oxygenspecies and subsequent death.80

PI3K/AKT/MTOR and RAS/RAF/mitogen-activatedprotein kinase pathways. The PI3K/AKT/MTOR and

RAS/RAF/mitogen-activated protein kinase pathways areactivated in around 5%�10% of HCC by amplification of theFGF19/CCND1 locus.76,81 Also, activating mutations ofPIK3CA (0%�2%) and inactivating mutations of TSC1 orTSC2 (3%�8%) lead to activation of the AKT/MTORsignaling in a subset of HCC.45,49,65 In addition, homozygousdeletion of PTEN, an inhibitor of the PI3K kinase, has beenidentified in 1%�3% of the HCC. However, some HCC withactivation of the PI3K/AKT/MTOR cascade have no geneticalterations in this pathway.82 Indirect upstream activationthrough insulin growth factor pathway has been proposed asan alternative mechanism of AKT/MTOR activation.83


D ING

Activating mutations of genes belonging to the RAS family arerarely observed in HCC (<2%), and inactivating mutations ofRP6SKA3, coding for the RAS inhibitor RSK2, were identifiedin 2%�9% of tumors.62 RSK2, is located downstreammitogen-activated protein kinase kinase and extracellularsignal-regulated kinase, and is a known negative control loopof RAS signaling.84 Inactivation of RSK2 released this negativefeedback and induced a constitutive activation of thepathway.49 Experimental data also suggest that persistentRAS activation may be a mechanism of HCC resistance tosorafenib.85

GENE

TICS

ANGE

NETICTEST

Molecular ClassesWe have described the landscape of mutations and

critical pathways involved in the development and pro-gression of HCC. But 2 important questions emerge: Are weable to classify tumors according to the molecular eventsdescribed here? And can we treat HCC patients based onthose biomarkers? Molecular classifications are aimed atproviding a molecular understanding of the different bio-logic events that drive tumor subclasses and also at definingspecific biomarkers/targets for therapies. Development ofwhole-genome expression profiling in the late 1990s had atremendous impact in biomedical research.86 The premisewas that, despite being at the same clinical stage, tumors aresignificantly different at a molecular level. Implications arehuge, because decision making in oncology relies mainly onclinical staging, and rarely considers tumor molecular sin-gularities. Array-based technologies enabled the analysis oflarge numbers of transcripts in multiple samples, setting thebasis for cancer classification based on expression pat-terns.87 These molecular classes reflected different biologicbackgrounds with potential implications in patient selectionfor therapies and prediction of clinical outcomes. Forexample, gene signatures in breast cancer identified thosewomen who would likely to benefit from adjuvant chemo-therapy after resection.88 This could provide clinicians withadditional tools to prevent patient’s exposure to unnec-essary drug toxicity.89 There are 2 layers of informationderived from molecular classification in HCC: first, its role asa prognostic or predictive biomarker, and second, its abilityto increase our understanding of the molecular pathogen-esis of the disease.

In HCC, many groups have reported molecular classifi-cation based on genomic profiling (Table 2).62,90–92 Most ofthese studies were conducted on resection specimens frompatients with different etiologies for their underlying liverdisease. Remarkably, despite that marker genes definingeach subclass were different across studies, integrativeanalysis showed that most of them identified commongenomic signals reflecting background biology.93 In otherwords, different investigators reported molecular subclassesthat identified similar patients, probably by spotting coregenomic alterations. Figure 4 summarizes our cumulativeunderstanding of gene expression�based HCC subclassesand its integration with additional levels of molecular in-formation. These studies along with a meta-analysis showedthat, regardless of the specific nomenclature used for each

class, HCC can be roughly divided into 2 major molecularsubtypes.94 One is broadly characterized by an enrichmentof signals related to cell proliferation and progression in thecell cycle (proliferation class) and is generally associatedwith a more aggressive phenotype, and the second classgenerally retains molecular features resembling normal he-patic physiology (nonproliferation class) (Figure 4).

Proliferation SubclassThe main traits of this class, which account for around

50% of patients, are related to activation of signaling cas-cades involved in cell proliferation/survival, enrichment ofsignatures of poor prognosis, and association with clinicalcharacteristics of aggressive tumors and poor outcomes.Activation of signaling pathway is remarkably heteroge-neous at the genomic and phenotypic levels. These includeAKT/MTOR,82 MET,95 TGFB,96 IGF,83 RAS/mitogen-activated protein kinase,97 among others. Interestingly,this class is also enriched in genomic signals that identifymarkers of progenitor cells, such as epithelial cell adhesionmolecule92,98 or hepatoblastoma-like.91 The cell of origin inHCC is a highly controversial issue, with recent experi-mental data supporting oncogenic reprogramming ofdifferent cellular lineages into cancer stem cells.99 A signa-ture derived from hepatoblastomas, the most frequent pri-mary pediatric liver tumor, as well as 2 signatures derivedfrom cholangiocarcinoma, are also enriched in thisclass,100–102 as well as Notch,103 which acts as a masterregulator of biliary differentiation. Altogether, these datafurther reinforce the notion of increased cellular plasticity inthese tumors. Different factors might explain increasedheterogeneity of the proliferation class, including higherrates of chromosomal instability or enrichment in aberrantepigenetic changes. For example, high-level DNA amplifica-tions of chromosome 11q13 (locus for FGF19, CCND1,ORAOV1, FGF4) are enriched in this class, in addition to aDNA methylation-based prognostic signature that mainlycorrelates with signatures of progenitor cells.104 MicroRNAderegulation is also prominent in this class, with a clusteredenrichment of a primate-specific family of miRNA located inchromosome 19 and microRNA/small nucleolar RNA onchromosome 14.105

From a clinical standpoint, patients in the proliferationclass have aggressive tumors, with higher a-fetoproteinlevels, moderately/poor cell differentiation on histology,and frequent vascular invasion.81,93 HBV-related HCC arepredominantly tumors belonging to this class. As predicted,patients with tumors from the proliferation subclass presenthigher risk of recurrence after resection and lower survivalrates.93,106

Nonproliferation ClassThis molecular subclass contains at least 2 critical fea-

tures: molecularly, it is dominated by activation of Wntsignaling in up to 25% of cases, and others are characterizedby immune response; and from a functional perspective, thetranscriptome of tumors resemble normal hepatic

Table 2.Prognostic Signatures in Hepatocellular Carcinoma

Predominantetiology

Clinical end point(time-to-event)

REMARKrecommendations First author, year

Advanced developmental stagemRNA-based5-gene signature Alcohol,

HCV, HBVSurvival OK Nault, 201351

EpCAM-signature HBV Survival OK Yamashita, 200892

186-gene signature (adjacent tissue) HCV Survival, HCCdevelopment

OK Hoshida, 2008118

miRNA-basedDown-regulation miR-26a HBV Survival OK Ji, 2009117

Pending further validationmRNA-based5-gene risk signature HCV Survival OK Villa, 2015152

65-gene signature HBV Survival OK Kim, 2012153

233-gene signature (adjacent tissue) HBV Late recurrence,recurrence-freesurvival

OK Kim, 2014120

G3 subclass HCV, alcohol Survival OK Boyault, 200670

Cluster A/hepatoblast-like HBV Survival OK Lee, 200691

17-gene metastasis predictor(adjacent tissue)

HBV Survival OK Budhu, 2006119

Metastasis-related gene signature HBV Survival OK Roessler, 2010154

Cholangiocarcinoma-like HBV Survival OK Woo, 2010101

617-gene SP-ZEB signature HBV Survival — Marquardt, 2011155

miRNA-based20-miRNA signature HBV Venous metastasis,

survivalOK Budhu, 2008116

Down-regulation miR-122 HBV, HCV,alcohol

Survival OK Coulouarn, 2009156

20-miRNA signature HBV Survival OK Wei, 2013157

C19MC HCV Survival — Augello, 2012158

Down-regulation Let-7 members HCV Early recurrence — Viswanathan, 2009159

Up-regulation miR-125a HBV Survival — Li, 2008160

19-miRNA signature HCV, alcohol Survival — Jiang, 2008161

Up-regulation miR-221 HCV Recurrence — Gramantieri, 2009162

DNA methylation36 CpG DNA methylation signature HCV Survival OK Villanueva, 2015104

Methylation signature HBV, HCV Survival — Hernandez-Vargas, 2010163

Genome wide hypomethylation HBV Survival — Calvisi, 2006164

Hypermethylation of E-cadherin orGSTP1

HBV Survival — Lee, 2003165

Hypermethylation of E-cadherin HBV Recurrence — Kwon, 2005166

EpCAM, epithelial cell adhesion molecule; REMARK, Reporting Recommendations for Tumor Marker Prognostic Studies.


GENETICSAND

GENETICTESTING

physiology. In terms of aberrant signaling, genomic datasuggest a dual modulation of WNT signaling in HCC,71 withdifferential representation in the proliferation and nonpro-liferation subclasses. Classical WNT signaling, as defined byup-regulation of well-known target genes, such as GLUL orLGR5, is significantly enriched in this class.70 These data arefurther confirmed when analyzing CTNNB1 mutations andnuclear translocation of b-catenin. A subset of tumorswithin this class is characterized by broad gains in chro-mosome 7, associated with overexpression of EGF receptorand male sex predominance.107 There are also data sug-gestive of immune signals within this class.81 Interestingly, asimilar predominance in inflammation-related genomictraits was also found in the less aggressive molecular

subclass of intrahepatic cholangiocarcinoma.102 From theclinical standpoint, tumors in this class show less aggressivephenotype, including better histologic differentiation, lowera-fetoprotein, and lack of enrichment in poor prognosissignatures. Regarding etiologic factors, HCV and alcohol-related HCC are more prevalent in this class.

The backbone of HCC classification relies on geneexpression profiling. However, there have been recent at-tempts to enhance cancer classification with combined ge-netic and epigenetic features. Despite not yet beingcomprehensively applied in HCC, a curated selection ofapproximately 500 features, including copy number gainsand losses, mutations, and methylation-enabled accuratecancer classification across 12 tumor types.108 Interestingly,

Figure 4. Summary of mo-lecular classification ofHCC. Major classes (prolif-eration and nonprolifera-tion) are depicted based onmessenger RNA expres-sion profiling. Additionalmolecular features affectingDNA structure, pathwayderegulation and epige-netics are overlapped.


GENE

TICS

AND

GENE

TICTESTING

at the top of this hierarchical classification, 2 main tumorsubclasses with significant differential prevalence of muta-tions (ie, M) and DNA copy number alterations (ie, C), whichsuggests that either mutations or chromosomal instabilitycan act as drivers of progression in different tumors.

Genetic Alterations as BiomarkersPrognosis

HCC prognosis prediction is currently assessed using theBarcelona Clinic Liver Cancer algorithm, as endorsed by theAmerican and European Associations for the Study of LiverDiseases.109,110 Barcelona Clinic Liver Cancer relies on acomposite of tumor burden, degree of liver damage, andcancer-related symptoms. In terms of molecular-guidedprognosis prediction, >40 prognostic gene signatures have

been described,111 although none has become a tangible toolin clinical decision making. Many factors could contribute tothis, including that molecular classes have been generatedmostly from surgical specimens, limiting its theoreticalapplicability to patients at earlier stages. There are recur-rent discussions on the impact of tumor heterogeneity, bothintra- and internodular, on molecular class predictionsbased on single biopsies.112 Despite this, molecular signa-tures are slowly permeating clinical practice guidelines.109 Aset of criteria has been proposed to determine formal in-clusion of molecular-based biomarkers in GPC.109 Also, theuse of archived specimens collected in the context of high-end clinical trials has been suggested as a legitimatesource for biomarker development.113

Most prognostic signatures were generated usingtranscriptome data, such as the G3,70,93 5-gene,114 epithe-lial cell adhesion molecule,92,115 cluster A,106 or


GENETICSAND

GENETICTESTING

hepatoblast-like.91 Others used epigenetic data to predictoutcomes, like the 36-CpG DNA methylation signature,104

the 20-miRNA signature,116 or the identification and vali-dation of miR-26 in HBV-HCC (Table 2).105,117 Most of themidentify patients within the proliferation subclass, however,the 5-gene score is also able to predict prognosis of tumorsclassified in the nonproliferative subclass (Figure 4). Inaddition to those derived from the tumor, a number ofstudies have reported signatures derived from adjacentcirrhotic tissue able to classify HCC patients based on theirprognosis. Presumably, these classes capture oncogenicsignals reflecting the so-called “field effect.” A 186-genesignature enriched in gene sets associated with inflamma-tion, including interferon signaling, activation of EGF, nu-clear factor kB, and tumor necrosis factor�a was able toidentify HCC patients, mainly HCV-related, with poor sur-vival after resection, as well was those cirrhotic patients athigher risk for HCC development.21,118 Differential inflam-mation patterns identified with genomic profiling were alsodeterminant to predict intrahepatic metastasis in a cohortof HBV-HCC.119 More recently, a 233-gene signature iden-tified in regenerating livers accurately predicts risk of denovo tumor formation (ie, late recurrence) in resectedHCC.120 Patients with poor prognosis signatures from thetumor (eg, G3, Cluster A) were not significantly enriched inthose signatures, conferring poor prognosis from theadjacent tissue (eg, 186-gene signature), as judged by anintegrative analysis in a large cohort.93 In other words, andsimilar to prognosis prediction using clinical systems,genomic data from both components (ie, tumor and adja-cent tissue) are complementary to maximize predictionaccuracy (Table 2).

Recent refinements in these studies have enabled pre-dictions based on individual risk assessments. For example,a nomogram that combines data from the 5-gene score inconjunction with Barcelona Clinic Liver Cancer stage andvascular invasion provides survival estimation on an indi-vidual basis.114 Similarly, the use of machine learningtechniques for class identification and prediction of a DNAmethylation-based signature provides a continuous gradientof risk, which ultimately might increase precision inoutcome prediction.104 Signature-based prognosis pre-dictions require tissue, which might constitute a limitationbecause HCC diagnosis can be confidently made with im-aging techniques. The development of technologies thatallow analysis of tumor byproducts in the blood, includingcell-free DNA or circulating tumor cells, might facilitate theimplementation of molecular-based biomarkers.121

TherapeuticIn oncology, survival benefits from molecular-targeted

therapies might be derived as a result of treating all pa-tients with very effective/broad therapies with manageabletoxicity, targeting a specific, generally small, subgroup ofpatients whose tumors are addicted to an oncogenic aber-ration; and targeting a broad number of patients whosetumors present a frequent mechanism of activation(signaling pathway) or immune response.

Studies in solid tumors have provided a broad picture ofthe mutational profile and identified a wide range of 5�150mutations per tumor, depending on exposure to genotoxics(non�small cell lung cancer, melanoma-high mutation rate),tumors developed in infants or hematologic tumors (low-mutation rate122). In HCC, the mean number of mutations incoding regions per tumor is 35�80, among which it hasbeen estimated that there might be 4�8 drivers of oncogeneaddiction.49 The oncogene addiction theory postulates thatsome tumors harbor certain somatic genetic alterations thatdominate the malignant phenotype.123 Translation of thisprinciple in cancer management provides the rationale forpersonalized care, commonly referred as precisionmedicine. There are many examples of survival benefitsachieved after selective inhibition of aberrant molecularsomatic events in solid tumors, such as vemurafenib inmelanoma with BRAF V600E mutations,124 or crizotinib inlung cancer patients with anaplastic lymphoma kinaserearrangements.125 Effectiveness of this approach relies onthe following factors: dominance of the candidate oncogenicaddiction loop in tumor progression, existence of abiomarker able to accurately identify patients with thisoncogenic loop, and a therapeutic agent able to selectivelyand effectively abrogate the loop. Ultimately, a well-designed, properly powered clinical trial enrolling patientswith the activated oncogenic loop and treated with itsantagonist will determine its clinical implementation.

In HCC, the only effective systemic agent is the BRAF/vascular endothelial growth factor receptor/platelet-derived growth factor receptor inhibitor sorafenib.126 Nooncogenic addiction loop has been validated yet. In addition,none of the negative phase 3 trials in advanced HCC re-ported so far used biomarker-based patient enrollment.127

It is feasible that not all tumor types can harbor oncogenicaddiction loops. In addition, inhibition of a single loop mightnot suffice to provide enough survival benefits at advancedstages, and could promote growth of resistant clones ormetastatic sites. Also, a comprehensive signaling blockademight be needed to achieve clinical responses, as has beenshown in breast cancer,128 and suggested in resistancemodels of HCC.85

Among the somatic alterations described previously andsignaling pathway activation, we can define 2 areas wherebiomarkers can be used: targeting oncogenic addictionloops based on the 3 requirements mentioned here, based inhuman data and experimental models, and targetingsignaling cascades or immune checkpoints (Tables 1 and 2).

Oncogenic loops. DNA copy number changes. VEGFA(gains in chromosome 6p21) is recurrently described,correlated with expression, and with an estimatedprevalence based on fluorescence in situ hybridization of7%�11%.81,129 Also, high VEGF plasma levels are correlatedwith poor survival130; there is solid evidence of anti-tumoreffect after VEGF inhibition in experimental models. VEGFAcould have a double pro-tumorigenic effect by promotingangiogenesis and inducing non�cell autonomous over-expression of hepatocyte growth factor.129 Early trials withVEGFA inhibitor bevacizumab in unselected populationsshow modest efficacy and raised some safety concerns.131


GENE

TICS

AND

GENE

TICTESTING

FGF19, CCND1 (gains in chromosome 11q13) aredescribed in 5%�14% of HCC,45,76,81 and are also corre-lated with high expression levels and poor prognosis.49

FGF19 blockade with monoclonal antibodies inhibits clonalgrowth and tumorigenicity in vivo.76,132 Strikingly, trans-genic mouse overexpressing FGF19, which binds to FGFR4in skeletal muscle, develop HCC,133 which points toward adual role of FGF19 in both tumor progression and initiation.In terms of FGF blockade, brivanib, a highly active FGFR1-3inhibitor, failed to improve survival in unselected advancedHCC populations.134,135 CCND1 shares genomic locus withFGF19, and it is also frequently amplified in cancer. CDK4/6dual inhibitors, such as palbociclib or abemaciclib, haveshown anti-tumoral activity in experimental models,although data from breast does not confirm a particularbenefit of CDK4/6 inhibition in CCND1 amplified tumors.136

FAK, MYC (gains in chromosome 8q24) have broad gainsdescribed in 4% of HCC, with a single study reporting26%.65,66 FAK inhibition induces tumor responses inexperimental models of HCC,137 and selective inhibitors ofFAK are currently under clinical development. MYC trans-genic mice develop liver cancer and it has a central roleduring hepatocarcinogenesis.138 However, the developmentof effective pharmacologic strategies to inhibit MYC hasbeen challenging.

Gene mutations. CTNNB1 is highly prevalent acrossstudies (approximately 25%).45,49,64 Similar to MYC, diffi-culties developing selective and nontoxic WNT inhibitorshave limited its clinical development,139 but clearly,CTNNB1 mutation is an appealing candidate for selectivetargeting.

TSC1/TSC2 are both negative modulators of MTORcascade, and their inactivation promotes MTOR signaling.140

Mutations, as well as DNA copy number alterations aredescribed in 2%�14% of HCC.45,49 Pathway-based analysisunderscores the potential driving role of this cascade inHCC,82 with a significant enrichment in Asian populations.45

In terms of MTOR inhibition, a phase 1 trial testing ever-olimus in combination with sorafenib show subtherapeuticconcentrations for everolimus and precluded furtherdevelopment of this combination.141 When tested in secondline, everolimus as a single agent failed to significantlyimprove survival in a phase 3 trial without biomarker-selected patients.142 Subgroup analyses suggested anincreased survival in HBV patients, maybe in relation to ahigher AKT/MTOR pathway activation in this population.

NRAS/KRAS/HRAS are mutated in <3% of HCC.49,65,66

Preliminary data from a phase 2 trial (ie, BASIL [(Assess-ing BAY86-9766 Plus Sorafenib for the Treatment of LiverCancer) NCT01204177]) in patients with advanced HCCreceiving a combination of MEK inhibitor refametinib(BAY86-9766) plus sorafenib found mutations in approxi-mately 6% of patients (4 of 69)143; analysis was conductedon circulating DNA. Three of these four patients hadconfirmed partial responses. Inactivating mutations in RASinhibitor RP6SKA3 are more prevalent (2%�9%)49,62,65 andcould also predict response to RAS inhibition. A phase2 study testing the combination of refametinib and sor-afenib in RAS mutant HCCs is ongoing. JAK1 mutation

frequency has been described up to 9% based on whole-genome sequencing data in HBV-HCC.66 This has not beenconfirmed in subsequent studies.45,49,64,65 Experimentalevidence suggests that its blockage could have antitumoractivity.66

Signaling Pathways or Immune CheckpointsMET. Overexpression at the messenger RNA and pro-

tein levels has been described in at least 40%�50% of HCCsamples.144 A subanalysis of the phase 2b trial testingtivantinib vs placebo in second line suggested that the drugwas particularly effective in those patients with highexpression of MET on immunohistochemistry.145 There hasbeen some concerns about tivantinib’s specificity as an METinhibitor,146 but a phase 3 RCT enrolling patients with highMET is currently evaluating its role in second line(NCT02029157).

IGF2. Enrichment of high transcript levels of IGF2(>20-fold) in approximately 10% of HCC samples suggestthat it could act as a tumor driver.83,147 DNA methylation isa reported mechanism for IGF2 deregulation in HCC,104 andit has been described as an oncogenic partner of Notch-induced HCC.103 Functional validation indicate that IGFpathway inhibition could be an effective strategy in HCC, butresults in early clinical trails using IGF-IR antagonists haveshown limited efficacy and significant toxicity.148

PD1/PD-L1. Both involved in the inhibitory signals toT cells, leading to suppression of anti-tumoral immuneresponse. Data in other solid tumors show remarkable re-sponses after selective inhibition with nivolumab,149 with apotential role of PD1/PD-L1 as a biomarkers of response.150

In HCC, intratumoral increased expression of PD1/PD-L1correlates with poor outcomes, and its inhibition in ortho-topic HCC models induces tumor responses.151 This repre-sents a novel immune-base approach. Studies withcheckpoint inhibitors are currently ongoing in phase 2 inHCC.

In conclusion, the studies reported during the last 10years have been able to delineate the landscape of muta-tions and genomic alterations occurring in HCC develop-ment and progression. This has certainly changed ourunderstanding of the disease. Nonetheless, we are at a pointthat there is a clear unmet need to translate this knowledgein actual advantages for patients, either in the arena ofrefining at-risk populations, patient prognosis, and re-sponses to therapies. A huge translational effort needsattention. The research community in HCC has to link dis-coveries with actual survival benefits and, therefore, it isnow about time to demonstrate that biomarkers definingsubclasses of patients ultimately might lead to better ther-apies that change the decision-making process in this com-plex and heterogeneous disease.

Supplementary MaterialNote: The first 50 references associated with this article areavailable below in print. The remaining references accom-panying this article are available online only with the elec-tronic version of this article. To access the supplementary


GENETICSAND

GENETICTESTING

material accompanying this article, visit the online versionof Gastroenterology at www.gastrojournal.org, and at http://dx.doi.org/10.1053/j.gastro.2015.05.061.

References
1. Tommaso AD. Risk of HCC: genetic heterogeneity and
complex genetics. J Hepatol 2010;52:252–257.2. Li M, Gerber DA, Koruda M, et al. Hepatocelluar carci-

noma associated with attenuated familial adenomatouspolyposis: a case report and review of the literature. ClinColorectal Cancer 2012;11:77–81.

3. Willson JSB, Godwin TD, Wiggins GAR, et al. Primaryhepatocellular neoplasms in a MODY3 family with a novelHNF1A germline mutation. J Hepatol 2013;59:904–907.

4. Bacq Y, Jacquemin E, Balabaud C, et al. Familia liveradenomatosis associated with hepatocyte nuclear factor1a inactivation. Gastroenterology 2003;125:1470–1475.

5. Calderaro J, Labrune P, Morcrette G, et al. Molecularcharacterization of hepatocellular adenomas developedin patients with glycogen storage disease type I.J Hepatol 2013;58:350–357.

6. Nahon P, Zucman-Rossi J. Single nucleotide poly-morphisms and risk of hepatocellular carcinoma incirrhosis. J Hepatol 2012;57:663–674.

7. Tarhuni A, Rufat P, Sutton A, et al. Impact of cytokinegene variants on the prediction and prognosis of hepa-tocellular carcinoma in patients with cirrhosis. J Hepatol2014;61:342–350.

8. Wang Y, Kato N, Hoshida Y, et al. Interleukin-1beta genepolymorphisms associated with hepatocellular carci-noma in hepatitis C virus infection. Hepatology 2003;37:65–71.

9. Migita K, Miyazoe S, Maeda Y, et al. Cytokine genepolymorphisms in Japanese patients with hepatitis Bvirus infection–association between TGF-beta1 poly-morphisms and hepatocellular carcinoma. J Hepatol2005;42:505–510.

10. Nahon P, Sutton A, Rufat P, et al. Myeloperoxidase andsuperoxide dismutase 2 polymorphisms comodulate therisk of hepatocellular carcinoma and death in alcoholiccirrhosis. Hepatology 2009;50:1484–1493.

11. Nahon P, Sutton A, Rufat P, et al. A variant in myelo-peroxidase promoter hastens the emergence of hepa-tocellular carcinoma in patients with HCV-relatedcirrhosis. J Hepatol 2012;56:426–432.

12. Nahon P, Sutton A, Rufat P, et al. Liver iron, HFE genemutations, and hepatocellular carcinoma occurrence inpatients with cirrhosis. Gastroenterology 2008;134:102–110.

13. Lauret E, Rodríguez M, González S, et al. HFE genemutations in alcoholic and virus-related cirrhotic patientswith hepatocellular carcinoma. Am J Gastroenterol 2002;97:1016–1021.

14. Long XD, Ma Y, Qu DY, et al. The polymorphism ofXRCC3 codon 241 and AFB1-related hepatocellularcarcinoma in Guangxi population, China. Ann Epidemiol2008;18:572–578.

15. Yuan J-M, Lu SC, Van Den Berg D, et al. Genetic poly-morphisms in the methylenetetrahydrofolate reductase

and thymidylate synthase genes and risk of hepatocel-lular carcinoma. Hepatology 2007;46:749–758.

16. Dharel N, Kato N, Muroyama R, et al. MDM2 promoterSNP309 is associated with the risk of hepatocellularcarcinoma in patients with chronic hepatitis C. ClinCancer Res 2006;12:4867–4871.

17. Yoon YJ, Chang HY, Ahn SH, et al. MDM2 and p53polymorphisms are associated with the development ofhepatocellular carcinoma in patients with chronic hepa-titis B virus infection. Carcinogenesis 2008;29:1192–1196.

18. Dayyeh BKA, Yang M, Fuchs BC, et al. A functionalpolymorphism in the epidermal growth factor gene isassociated with risk for hepatocellular carcinoma.Gastroenterology 2011;141:141–149.

19. Pinyol R, Nault J, Quetglas I, et al. Molecular profiling ofliver tumors: classification and clinical translation fordecision making. Semin Liver Dis 2014;34:363–375.

20. Schiffer E, Housset C, Cacheux W, et al. Gefitinib, anEGFR inhibitor, prevents hepatocellular carcinomadevelopment in the rat liver with cirrhosis. Hepatology2005;41:307–314.

21. Hoshida Y, Villanueva A, Sangiovanni A, et al. Prognosticgene expression signature for patients with hepatitis C-related early-stage cirrhosis. Gastroenterology 2013;144:1024–1030.

22. Bruix J, Takayama T, Mazzaferro V, et al. STORM: aphase III randomized, double-blind, placebo-controlledtrial of adjuvant sorafenib after resection or ablation toprevent recurrence of hepatocellular carcinoma. J ClinOncol 2014;32:5s; 2014 (Suppl; Abstr 4006).

23. Bressac B, Kew M, Wands J, et al. Selective G to Tmutations of p53 gene in hepatocellular carcinoma fromsouthern Africa. Nature 1991;350:429–431.

24. Wang B, Huang G, Wang D, et al. Null genotypes ofGSTM1 and GSTT1 contribute to hepatocellular carci-noma risk: evidence from an updated meta-analysis.J Hepatol 2010;53:508–518.

25. Kozlitina J, Xing C, Pertsemlidis A, et al. Genetic vari-ation in PNPLA3 confers susceptibility to nonalcoholicfatty liver disease. Nat Genet 2008;40:1461–1465.

26. Liu YL, Patman GL, Leathart JBS, et al. Carriage of thePNPLA3 rs738409 C >G polymorphism confers anincreased risk of non-alcoholic fatty liver disease asso-ciated hepatocellular carcinoma. J Hepatol 2014;61:75–81.

27. Nahon P, Valenti L, Henrion J, et al. Association betweenthe PNPLA3 (rs738409 C>G) variant and hepatocellularcarcinoma: Evidence from a meta-analysis of individualparticipant data. Hepatology 2014;59:2170–2177.

28. Singal AG, Manjunath H, Yopp AC, et al. The effect ofPNPLA3 on fibrosis progression and development ofhepatocellular carcinoma: a meta-analysis. Am J Gas-troenterol 2014;109:325–334.

29. Guyot E, Sutton A, Rufat P, et al. PNPLA3 rs738409,hepatocellular carcinoma occurrence and risk modelprediction in patients with cirrhosis. J Hepatol 2013;58:312–318.

30. Kumar V, Kato N, Urabe Y, et al. Genome-wide asso-ciation study identifies a susceptibility locus for

http://www.gastrojournal.org



http://refhub.elsevier.com/S0016-5085(15)00869-0/sref1


















































































































































GENE

TICS

AND

GENE

TICTESTING

HCV-induced hepatocellular carcinoma. Nat Genet 2011;43:455–458.

31. Miki D, Ochi H, Hayes CN, et al. Variation in the DEPDC5locus is associated with progression to hepatocellularcarcinoma in chronic hepatitis C virus carriers. Nat Genet2011;43:797–800.

32. Jiang D-K, Sun J, Cao G, et al. Genetic variants inSTAT4 and HLA-DQ genes confer risk of hepatitis Bvirus-related hepatocellular carcinoma. Nat Genet 2013;45:72–75.

33. Clifford RJ, Zhang J, Meerzaman DM, et al. Geneticvariations at loci involved in the immune response arerisk factors for hepatocellular carcinoma. Hepatology2010;52:2034–2043.

34. Chan KYK, Wong C-M, Kwan JS-H, et al. Genome-wide association study of hepatocellular carcinoma inSouthern Chinese patients with chronic hepatitis B virusinfection. PLoS ONE 2011;6:e28798.

35. Zhang H, Zhai Y, Hu Z, et al. Genome-wide associationstudy identifies 1p36.22 as a new susceptibility locus forhepatocellular carcinoma in chronic hepatitis B viruscarriers. Nat Genet 2010;42:767–770.

36. Farazi PA, Depinho RA. Hepatocellular carcinoma path-ogenesis: from genes to environment. Nat Rev Cancer2006;6:674–687.

37. International Consensus Group for HepatocellularNeoplasia. Pathologic diagnosis of early hepatocellularcarcinoma: a report of the international ConsensusGroup for Hepatocellular Neoplasia. Hepatology 2009;49:658–664.

38. Satyanarayana A, Manns MP, Rudolph KL. Telomeresand telomerase: a dual role in hepatocarcinogenesis.Hepatology 2004;40:276–283.

39. Calado RT, Young NS. Telomere diseases. N Engl J Med2009;361:2353–2365.

40. Günes C, Rudolph KL. The role of telomeres in stem cellsand cancer. Cell 2013;152:390–393.

41. Hartmann D, Srivastava U, Thaler M, et al. Telomerasegene mutations are associated with cirrhosis formation.Hepatology 2011;53:1608–1617.

42. Rudolph KL, Chang S, Millard M, et al. Inhibition ofexperimental liver cirrhosis in mice by telomerase genedelivery. Science 2000;287:1253–1258.

43. Farazi PA, Glickman J, Jiang S, et al. Differential impact oftelomere dysfunction on initiation and progression of he-patocellular carcinoma. Cancer Res 2003;63:5021–5027.

44. Nault JC, Mallet M, Pilati C, et al. High frequencyof telomerase reverse-transcriptasepromoter somatic

mutations in hepatocellularcarcinoma and preneoplasticlesions. Nat Commun 2013;4:1–6.

45. Totoki Y, Tatsuno K, Covington KR, et al. Trans-ancestry mutational landscape of hepatocellular carci-noma genomes. Nat Genet 2014;46:1267–1273.

46. Sung W-K, Zheng H, Li S, et al. Genome-wide survey ofrecurrent HBV integration in hepatocellular carcinoma.Nat Genet 2012;44:765–769.

47. Paterlini-Bréchot P, Saigo K, Murakami Y, et al. HepatitisB virus-related insertional mutagenesis occurs frequentlyin human liver cancers and recurrently targets humantelomerase gene. Oncogene 2003;22:3911–3916.

48. Nault JC, Calderaro J, Di Tommaso L, et al. Telomerasereverse transcriptase promoter mutation is an early so-matic genetic alteration in the transformation of prema-lignant nodules in hepatocellular carcinoma on cirrhosis.Hepatology 2014;60:1983–1992.

49. Schulze K, Imbeaud S, Letouze E, et al. Exomesequencing of hepatocellular carcinoma identifies newmutations signatures and potential therapeutic targets.Nat Genet 2015;47:505–511.

50. Di Tommaso L, Sangiovanni A, Borzio M, et al. Advancedprecancerous lesions in the liver. Best Pract Res ClinGastroenterol 2013;27:269–284.

Author names in bold designate shared co-first authorship.

Received March 19, 2015. Accepted May 20, 2015.

Reprint requestsAddress requests for reprints to: Josep M. Llovet, MD, PhD, Liver CancerTranslational Research Laboratory, Barcelona Clinic Liver Cancer Group,Liver Unit, CIBERehd, Hospital Clínic, IDIBAPS, University of Barcelona,Villarroel 170, 08036 Barcelona, Catalonia, Spain. e-mail: [email protected];fax: þ34 93 3129406; or Jessica Zucman-Rossi, MD, PhD, INSERM U 1162,Génomique Fonctionnelle des Tumeurs Solides, Paris, France. e-mail:[email protected]; fax: þ33 01 53725192.

AcknowledgmentsAuthor contributions: JZ-R, AV, J-CN, and JML participated in all stages ofmanuscript production, design, figures, tables, writing, and review of finalversion.

Conflicts of interestThese authors disclose the following: Josep M. Llovet is consultant for Bayer,BMS, Lilly, Blueprint, Celsion, GSK, and Boehringer-Ingelheim, and hasreceived research grants from Bayer, BMS, Blueprint, and Boehringer-Ingelheim. Jessica Zucman-Rossi is consultant for IntraGen. The remainingauthors disclose no conflicts.

FundingJessica Zucman-Rossi has grant from the Ligue Contre le Cancer and the INCawith the ICGC project. Josep M. Llovet received grants from the EuropeanCommission Framework Programme 7 (Heptromic, proposal number259744), Samuel Waxman Cancer Research Foundation, Grant IþD Program(SAF2013-41027), and the Asociación Española Contra el Cáncer (AECC).





























































































mailto:[email protected]

mailto:[email protected]

Supplementary References51. Nault JC, Bioulac-Sage P, Zucman-Rossi J. Hepatocel-

lular benign tumors-from molecular classification topersonalized clinical care. Gastroenterology 2013;144:888–902.

52. Bluteau O, Jeannot E, Bioulac-Sage P, et al. Bi-allelicinactivation of TCF1 in hepatic adenomas. Nat Genet2002;32:312–315.

53. Rebouissou S, Amessou M, Couchy G, et al. Frequentin-frame somatic deletions activate gp130 in inflam-matory hepatocellular tumours. Nature 2009;457:200–204.

54. Zucman-Rossi J, Jeannot E, Nhieu JTV, et al. Genotype-phenotype correlation in hepatocellular adenoma: newclassification and relationship with HCC. Hepatology2006;43:515–524.

55. Pilati C, Nault JC, Imbeaud S, et al. Genomic profiling ofhepatocellular adenomas reveals recurrent FRK-activating mutations and the mechanisms of malignanttransformation. Cancer Cell 2014;25:428–441.

56. Neuveut C, Wei Y, Buendia MA. Mechanisms of HBV-related hepatocarcinogenesis. J Hepatol 2010;52:594–604.

57. Brechot C, Pourcel C, Louise A, et al. Presence of inte-grated hepatitis B virus DNA sequences in cellular DNAof human hepatocellular carcinoma. Nature 1980;286:533–535.

58. Nault JC, Datta S, Imbeaud S, et al. Recurrent AAV2-related insertional mutagenesis in human hepatocellularcarcinomas. Nat Genet 2015 Aug 24. http://dx.doi.org/10.1038/ng.3389. [Epub ahead of print].

59. Helleday T, Eshtad S, Nik-Zainal S. Mechanisms un-derlying mutational signatures in human cancers. NatRev Genet 2014;15:585–598.

60. Poon SL, Pang S-T, McPherson JR, et al. Genome-wide mutational signatures of aristolochic acid and itsapplication as a screening tool. Sci Transl Med 2013;5:197–201.

61. Alexandrov LB, Nik-Zainal S, Wedge DC, et al. Signa-tures of mutational processes in human cancer. Nature2013;500:415–421.

62. Guichard C, Amaddeo G, Imbeaud S, et al. Inte-grated analysis of somatic mutations and focal copy-number changes identifies key genes and pathwaysin hepatocellular carcinoma. Nat Genet 2012;44:694–698.

63. Huang J, Deng Q, Wang Q, et al. Exome sequencing ofhepatitis B virus-associated hepatocellular carcinoma.Nat Genet 2012;44:1117–1121.

64. Cleary SP, Jeck WR, Zhao X, et al. Identification of drivergenes in hepatocellular carcinoma by exomesequencing. Hepatology 2013;58:1693–1702.

65. Ahn S-M, Jang SJ, Shim JH, et al. Genomic portrait ofresectable hepatocellular carcinomas: implications ofRB1 and FGF19 aberrations for patient stratification.Hepatology 2014;60:1972–1982.

66. Kan Z, Zheng H, Liu X, et al. Whole-genomesequencing identifies recurrent mutations in hepatocel-lular carcinoma. Genome Res 2013;23:1422–1433.

67. Stratton MR, Campbell PJ, Futreal PA. The cancergenome. Nature 2009;458:719–724.

68. La Coste de A, Romagnolo B, Billuart P, et al. Somaticmutations of the beta-catenin gene are frequent inmouse and human hepatocellular carcinomas. Proc NatlAcad Sci U S A 1998;95:8847–8851.

69. Satoh S, Daigo Y, Furukawa Y, et al. AXIN1 mutations inhepatocellular carcinomas, and growth suppression incancer cells by virus-mediated transfer of AXIN1. NatGenet 2000;24:245–250.

70. Boyault S, Rickman DS, De Reyniès A, et al. Tran-scriptome classification of HCC is related to gene alter-ations and to new therapeutic targets. Hepatology 2006;45:42–52.

71. Lachenmayer A, Alsinet C, Savic R, et al. Wnt-pathwayactivation in two molecular classes of hepatocellularcarcinoma and experimental modulation by sorafenib.Clin Cancer Res 2012;18:4997–5007.

72. Audard V, Grimber G, Elie C, et al. Cholestasis is amarker for hepatocellular carcinomas displaying beta-catenin mutations. J Pathol 2007;212:345–352. Avail-able at: http://doi.wiley.com/10.1002/path.2169.

73. Fujimoto A, Furuta M, Shiraishi Y, et al. Whole-genomemutational landscape of liver cancers displaying biliaryphenotype reveals hepatitis impact and molecular di-versity. Nat Commun 2015;6:1–8.

74. Hsu IC, Metcalf RA, Sun T, et al. Mutational hotspot inthe p53 gene in human hepatocellular carcinomas. Na-ture 1991;350:427–428.

75. Fujimoto A, Totoki Y, Abe T, et al. Whole-genomesequencing of liver cancers identifies etiological in-fluences on mutation patterns and recurrent mutations inchromatin regulators. Nat Genet 2012;44:760–764.

76. Sawey ET, Chanrion M, Cai C, et al. Identification of atherapeutic strategy targeting amplified FGF19 in livercancer by oncogenomic screening. Cancer Cell 2011;19:347–358.

77. Wang K, Lim HY, Shi S, et al. Genomic landscape ofcopy number aberrations enables the identification ofoncogenic drivers in hepatocellular carcinoma. Hep-atology 2013;58:706–717.

78. Li M, Zhao H, Zhang X, et al. Inactivating mutations ofthe chromatin remodeling gene ARID2 in hepatocellularcarcinoma. Nat Genet 2011;43:828–829.

79. Sporn MB, Liby KT. NRF2 and cancer: the good, thebad and the importance of context. Nature 2012;12:564–571.

80. DeNicola GM, Karreth FA, Humpton TJ, et al. Oncogene-induced Nrf2 transcription promotes ROS detoxificationand tumorigenesis. Nature 2011;475:106–109.

81. Chiang DY, Villanueva A, Hoshida Y, et al. Focal gains ofVEGFA and molecular classification of hepatocellularcarcinoma. Cancer Res 2008;68:6779–6788.

82. Villanueva A, Chiang DY, Newell P, et al. Pivotal role ofmTOR signaling in hepatocellular carcinoma. Gastroen-terology 2008;135:1972–1983.

83. Tovar V, Alsinet C, Villanueva A, et al. IGF activation in amolecular subclass of hepatocellular carcinoma and pre-clinical efficacy of IGF-1R blockage. J Hepatol 2010;52:550–559.

1239.e1 Zucman-Rossi et al Gastroenterology Vol. 149, No. 5


































http://dx.doi.org/10.1038/ng.3389

http://dx.doi.org/10.1038/ng.3389




























































http://doi.wiley.com/10.1002/path.2169


















































84. Douville E, Downward J. EGF induced SOS phosphory-lation in PC12 cells involves P90 RSK-2. Oncogene1997;15:373–383.

85. Rudalska R, Dauch D, Longerich T, et al. RNAiscreening identifies a mechanism of sorafenib resistancein liver cancer. Nat Med 2014:1–11.

86. Golub TR. Molecular classification of cancer: class dis-covery and class prediction by gene expression moni-toring. Science 1999;286:531–537.

87. DeRisi J, Penland L, Brown PO, et al. Use of a cDNAmicroarray to analyse gene expression patterns in hu-man cancer. Nat Genet 1996;14:457–460.

88. van de Vijver MJ, He YD, Van’t Veer LJ, et al. A gene-expression signature as a predictor of survival in breastcancer. N Engl J Med 2002;347:1999–2009.

89. Solin LJ, Gray R, Baehner FL, et al. A multigeneexpression assay to predict local recurrence risk forductal carcinoma in situ of the breast. J Natl Cancer Inst2013;105:701–710.

90. Hoshida Y, Toffanin S, Lachenmayer A, et al. Molecularclassification and novel targets in hepatocellular carci-noma: recent advancements. Semin Liver Dis 2010;30:35–51.

91. Lee J-S, Heo J, Libbrecht L, et al. A novel prognosticsubtype of human hepatocellular carcinoma derived fromhepatic progenitor cells. Nat Med 2006;12:410–416.

92. Yamashita T, Forgues M, Wang W, et al. EpCAM andalpha-fetoprotein expression defines novel prognosticsubtypes of hepatocellular carcinoma. Cancer Res 2008;68:1451–1461.

93. Villanueva A, Hoshida Y, Battiston C, et al. Combiningclinical, pathology, and gene expression data to predictrecurrence of hepatocellular carcinoma. Gastroenter-ology 2011;140:1501–1512.

94. Hoshida Y, Nijman SMB, Kobayashi M, et al. Integrativetranscriptome analysis reveals common molecular sub-classes of human hepatocellular carcinoma. Cancer Res2009;69:7385–7392.

95. Kaposi-Novak P. Met-regulated expression signaturedefines a subset of human hepatocellular carcinomaswith poor prognosis and aggressive phenotype. J ClinInvest 2006;116:1582–1595.

96. Coulouarn C, Factor VM, Thorgeirsson SS. Transforminggrowth factor-b gene expression signature in mousehepatocytes predicts clinical outcome in human cancer.Hepatology 2008;47:2059–2067.

97. Newell P, Toffanin S, Villanueva A, et al. Ras pathwayactivation in hepatocellular carcinoma and anti-tumoraleffect of combined sorafenib and rapamycin in vivo.J Hepatol 2009;51:725–733.

98. Yamashita T, Wang XW. Cancer stem cells in thedevelopment of liver cancer. J Clin Invest 2013;123:1911–1918.

99. Holczbauer Á, Factor VM, Andersen JB, et al. Modelingpathogenesis of primary liver cancer in lineage-specificmouse cell types. Gastroenterology 2013;145:221–231.

100.Cairo S, Armengol C, De Reyniès A, et al. Hepatic stem-like phenotype and interplay of Wnt/beta-catenin andMyc signaling in aggressive childhood liver cancer.Cancer Cell 2008;14:471–484.

101.Woo HG, Lee JH, Yoon JH, et al. Identification of acholangiocarcinoma-like gene expression trait in hepa-tocellular carcinoma. Cancer Res 2010;70:3034–3041.

102.Sia D, Hoshida Y, Villanueva A, et al. Integrative molec-ular analysis of intrahepatic cholangiocarcinoma reveals2 classes that have different outcomes. Gastroenter-ology 2013;144:829–840.

103.Villanueva A, Alsinet C, Yanger K, et al. Notch signalingis activated in human hepatocellular carcinoma and in-duces tumor formation in mice. Gastroenterology 2012;143:1660–1669.

104.Villanueva A, Portela A, Sayols S, et al. DNAMethylation-based prognosis and epidrivers in hepato-cellular carcinoma. Hepatology 2015;61:1945–1956.

105.Toffanin S, Hoshida Y, Lachenmayer A, et al. MicroRNA-based classification of hepatocellular carcinoma andoncogenic role of miR-517a. Gastroenterology 2011;140:1618–1628.

106.Lee J-S, Chu I-S, Heo J, et al. Classification and pre-diction of survival in hepatocellular carcinoma by geneexpression profiling. Hepatology 2004;40:667–676.

107.Keng VW, Sia D, Sarver AL, et al. Sex bias occurrence ofhepatocellular carcinoma in Poly7 molecular subclass isassociated with EGFR. Hepatology 2012;57:120–130.

108.Ciriello G, Miller ML, Aksoy BA, et al. Emerging land-scape of oncogenic signatures across human cancers.Nat Genet 2013;45:1127–1133.

109.European Association for the Study of the Liver, EuropeanOrganisation for Research and Treatment of Cancer.EASL-EORTCClinicalPracticeGuidelines:managementofhepatocellular carcinoma. J Hepatol 2012;56:908–943.

110.Bruix J, Sherman M; American Association for the Studyof Liver Diseases. Management of hepatocellular carci-noma: an update. Hepatology 2011;53:1020–1022.

111.Hoshida Y, Moeini A, Alsinet C, et al. Gene signatures inthe management of hepatocellular carcinoma. SeminOncol 2012;39:473–485.

112.Friemel J, Rechsteiner MP, Frick L, et al. Intratumorheterogeneity in hepatocellular carcinoma. Clin CancerRes 2014;21:1951–1961.

113.Simon RM, Paik S, Hayes DF. Use of archived speci-mens in evaluation of prognostic and predictive bio-markers. J Natl Cancer Inst 2009;101:1446–1452.

114.Nault JC, De Reyniès A, Villanueva A, et al.A hepatocellular carcinoma 5-gene score associatedwith survival of patients after liver resection. Gastroen-terology 2013;145:176–187.

115.Yamashita T, Ji J, Budhu A, et al. EpCAM-positive he-patocellular carcinoma cells are tumor-initiating cellswith stem/progenitor cell features. Gastroenterology2009;136:1012–1124.

116.Budhu A, Jia HL, Forgues M, et al. Identification ofmetastasis-related microRNAs in hepatocellular carci-noma. Hepatology 2008;47:897–907.

117.Ji J, Shi J, Budhu A, et al. MicroRNA expression, sur-vival, and response to interferon in liver cancer. N Engl JMed 2009;361:1437–1447.

118.Hoshida Y, Villanueva A, Kobayashi M, et al. Geneexpression in fixed tissues and outcome in hepatocel-lular carcinoma. N Engl J Med 2008;359:1995–2004.

October 2015 Genetic Landscape in HCC 1239.e2




























































































































































119.Budhu A, Forgues M, Ye Q, et al. Prediction ofvenous metastases, recurrence, and prognosis in hepa-tocellular carcinoma based on a unique immune responsesignature of the liver microenvironment. Cancer Cell 2006;10:99–111.

120.Kim JH, Sohn BH, Lee H-S, et al. Genomic predictors forrecurrence patterns of hepatocellular carcinoma: modelderivation and validation. PLoS Med 2014;11:e1001770.

121.Jiang P, Chan CWM, Chan KCA, et al. Lengthening andshortening of plasma DNA in hepatocellular carcinomapatients. Proc Natl Acad Sci U S A 2015:201500076.

122.Vogelstein B, Zhou S. Cancer genome landscapes.Science 2013;339:1546–1558.

123.Weinstein IB. Cancer. Addiction to oncogenes—theAchilles heal of cancer. Science 2002;297:63–64.

124.Chapman PB, Hauschild A, Robert C, et al. Improvedsurvival with vemurafenib in melanoma with BRAF V600Emutation. N Engl J Med 2011;364:2507–2516.

125.Shaw AT, Kim D-W, Nakagawa K, et al. Crizotinibversus chemotherapy in advanced ALK-positive lungcancer. N Engl J Med 2013;368:2385–2394.

126.Llovet JM, Ricci S, Mazzaferro V, et al. Sorafenib inadvanced hepatocellular carcinoma. N Engl J Med 2008;359:378–390.

127.Llovet JM, Villanueva A, Lachenmayer A, et al. Moleculartargeted therapies in hepatocellular carcinoma. Nat RevClin Oncol 2015;12:408–424.

128.Swain SM, Baselga J, Kim S-B, et al. Pertuzumab,trastuzumab, and docetaxel in HER2-Positive metastaticbreast cancer. N Engl J Med 2015;372:724–734.

129.Horwitz E, Stein I, Andreozzi M, et al. Human and mouseVEGFA-amplified hepatocellular carcinomas are highlysensitive to sorafenib treatment. Cancer Discov 2014;4:730–743.

130.Llovet JM, Pena CEA, Lathia CD, et al. Plasma biomarkersas predictors of outcome in patients with advanced he-patocellular carcinoma. Clin Cancer Res 2012;18:2290–2300.

131.Siegel AB, Cohen EI, Ocean A, et al. Phase II trial eval-uating the clinical and biologic effects of bevacizumab inunresectable hepatocellular carcinoma. J Clin Oncol2008;26:2992–2998.

132.Desnoyers LR, Pai R, Ferrando RE, et al. TargetingFGF19 inhibits tumor growth in colon cancer xenograftand FGF19 transgenic hepatocellular carcinoma models.Oncogene 2008;27:85–97.

133.Nicholes K, Guillet S, Tomlinson E, et al. A mouse modelof hepatocellular carcinoma: ectopic expression offibroblast growth factor 19 in skeletal muscle of trans-genic mice. Am J Pathol 2002;160:2295–2307.

134.Llovet JM, Decaens T, Raoul JL, et al. Brivanib in pa-tients with advanced hepatocellular carcinoma who wereintolerant to sorafenib or for whom sorafenib failed: re-sults from the randomized phase III BRISK-PS study.J Clin Oncol 2013;31:3509–3516.

135.Johnson PJ, Qin S, Park JW, et al. Brivanib versus sor-afenib as first-line therapy in patients with unresectable,advanced hepatocellular carcinoma: results from therandomized phase III BRISK-FL study. J Clin Oncol2013;31:3517–3524.

136.Finn RS, Crown JP, Lang I, et al. The cyclin-dependentkinase 4/6 inhibitor palbociclib in combination withletrozole versus letrozole alone as first-line treatment ofoestrogen receptor-positive, HER2-negative, advancedbreast cancer (PALOMA-1/TRIO-18): a randomisedphase 2 study. Lancet Oncol 2015;16:25–35.

137.Shang N, Arteaga M, Zaidi A, et al. FAK is required for c-Met/b-catenin-driven hepatocarcinogenesis. Hepatology2015;61:214–226.

138.Kaposi-Novak P, Libbrecht L, Woo HG, et al. Central roleof c-Myc during malignant conversion in human hep-atocarcinogenesis. Cancer Res 2009;69:2775–2782.

139.Anastas JN, Moon RT. WNT signalling pathways astherapeutic targets in cancer. Nat Rev Cancer 2013;13:1–16.

140.Inoki K, Corradetti MN, Guan K-L. Dysregulation of theTSC-mTOR pathway in human disease. Nat Genet 2005;37:19–24.

141.Finn RS, Poon RTP, Yau T, et al. Phase I study investi-gating everolimus combined with sorafenib in patientswith advanced hepatocellular carcinoma. J Hepatol2013;59:1271–1277.

142.Zhu AX, Kudo M, Assenat E, et al. Effect of everolimus onsurvival in advanced hepatocellular carcinoma after fail-ure of sorafenib. JAMA 2014;312:57.

143.Lim HY, Heo J, Choi HJ, et al. A phase II study of theefficacy and safety of the combination therapy of theMEK inhibitor refametinib (BAY 86-9766) plus sorafenibfor Asian patients with unresectable hepatocellular car-cinoma. Clin Cancer Res 2014;20:5976–5985.

144.Goyal L, Muzumdar MD, Zhu AX. Targeting the HGF/c-MET pathway in hepatocellular carcinoma. Clin CancerRes 2013;19:2310–2318.

145.Santoro A, Rimassa L, Borbath I, et al. Tivantinib forsecond-line treatment of advanced hepatocellular car-cinoma: a randomised, placebo-controlled phase 2study. Lancet Oncol 2013;14:55–63.

146.Michieli P, Di Nicolantonio F. Targeted therapies:Tivantinib—a cytotoxic drug in MET inhibitor’s clothes?Nat Rev Clin Oncol 2013;10:372–374.

147.Breuhahn K, Vreden S, Haddad R, et al. Molecularprofiling of human hepatocellular carcinoma definesmutually exclusive interferon regulation and insulin-likegrowth factor II overexpression. Cancer Res 2004;64:6058–6064.

148.Abou-Alfa GK, Capanu M, O’Reilly EM, et al. A phase IIstudy of cixutumumab (IMC-A12, NSC742460) inadvanced hepatocellular carcinoma. J Hepatol 2014;60:319–324.

149.Robert C, Long GV, Brady B, et al. Nivolumab in previ-ously untreated melanoma without BRAFMutation.N Engl J Med 2015;372:320–330.

150.Shi F, Shi M, Zeng Z, et al. PD-1 and PD-L1 upregulationpromotes CD8þ T-cell apoptosis and postoperativerecurrence in hepatocellular carcinoma patients. Int JCancer 2010;128:887–896.

151.Chen Y, Ramjiawan RR, Reiberger T, et al. CXCR4inhibition in tumor microenvironment facilitates anti-PD-1 immunotherapy in sorafenib-treated HCC in mice.Hepatology 2015;61:1591–1602.

1239.e3 Zucman-Rossi et al Gastroenterology Vol. 149, No. 5


























































































































































152.Villa E, Critelli R, Lei B, et al. Neoangiogenesis-relatedgenes are hallmarks of fast-growing hepatocellular car-cinomas and worst survival. Results from a prospectivestudy. Gut 2015 (in press, doi: 10.1136/gutjnl-2014-308483).

153.Kim JH, Sohn BH, Lee HS, et al. Genomic predictors forrecurrence patterns of hepatocellular carcinoma: modelderivation and validation. PLoS Med 2015 (in press, doi:10.1371/journal.pmed.1001770).

154.Roessler S, Budhu A, Forgues M, et al. A uniquemetastasis gene signature enables prediction of tumorrelapse in early-stage hepatocellular carcinoma patients.Cancer Res 2010;70:10202–10212.

155.Marquardt JU, Raggi C, Andersen JB. Human hepaticcancer stem cells are characterized by common stem-ness traits and diverse oncogenic pathways. Hepatology2011;54:1031–1042.

156.Coulouarn C, Factor VM, Andersen JB, et al. Loss ofmiR-122 expression in liver cancer correlates with sup-pression of the hepatic phenotype and gain of metastaticproperties. Oncogene 2009;28:3526–3536.

157.Wei R, Huang GL, Zhang MY, et al. Clinical Significanceand prognostic value of microRNA expression signaturesin hepatocellular carcinoma. Clin Cancer Res 2013;19:4780–4791.

158.Augello C, Vaira V, Caruso L, et al. MicroRNA profiling ofhepatocarcinogenesis identifies C19MC cluster as anovel prognostic biomarker in hepatocellular carcinoma.Liver Int 2012;32:772–782.

159.Viswanathan SR, Powers JT, Einhorn W, et al. Lin28 pro-motes transformation and is associated with advancedhuman malignancies. Nat Genet 2009;41:843–847.

160.Li W, Xie L, He X, et al. Diagnostic and prognostic im-plications of microRNAs in human hepatocellular carci-noma. Int J Cancer 2008;123:1616–1622.

161.Jiang J, Gusev Y, Aderca I, et al. Association of Micro-RNA expression in hepatocellular carcinomas with hep-atitis infection, cirrhosis, and patient survival. ClinCancer Res 2008;14:419–427.

162.Gramantieri L, Fornari F, Ferracin M, et al. MicroRNA-221Targets Bmf in Hepatocellular Carcinoma and Correlateswith Tumor Multifocality. Clin Cancer Res 2009;15:5073–5081.

163.Hernandez-Vargas H, Lambert MP, Le Calvez-Kelm F,et al. Hepatocellular carcinoma displays distinct DNAmethylation signatures with potential as clinical pre-dictors. PLoS ONE 2010;5:e9749.

164.Calvisi D, Ladu S, Gorden A, et al. Ubiquitous activationof Ras and Jak/Stat pathways in human HCC. Gastro-enterology 2006;130:1117–1128.

165.Lee S, Lee HJ, Kim J-H, et al. Aberrant CpG islandhypermethylation along multistep hepatocarcinogenesis.Am J Pathol 2003;163:1371–1378.

166.Kwon GY, Yoo BC, Koh, et al. Promoter methylation ofE-cadherin in hepatocellular carcinomas and dysplasticnodules. J Korean Med Sci 2005;20:242–247.

Author names in bold designate shared co-first authorship.

October 2015 Genetic Landscape in HCC 1239.e4

http://10.1136/gutjnl-2014-308483

http://10.1136/gutjnl-2014-308483

http://10.1371/journal.pmed.1001770




























































genetic landscape and biomarkers of hepatocellular …genetic landscape and biomarkers of...

Documents