Nras in melanoma: Targeting the undruggable target

Nras

in

melanoma:

Targeting

the

undruggable

target

Mario

Mandalà

,

Barbara

Merelli

,

Daniela

Massi

a_{UnitofMedicalOncology,DepartmentofOncologyandHematology,PapaGiovanniXXIIIHospital,Bergamo,Italy} b_{DivisionofPathologicalAnatomy,DepartmentofSurgeryandTranslationalMedicine,UniversityofFlorence,Italy}

Accepted9May2014

Contents

1. HistoricalperspectiveonNRASincancerandafocusonmelanoma... 108

2. BiologicalfunctionsofNRAS... 109

2.1. Cellproliferation... 109

2.2. Suppressionofapoptosis ... 111

2.3. Metabolism... 111

2.4. Remodelingthemicroenvironment ... 111

2.5. Evasionoftheimmuneresponse... 111

2.6. Metastasis... 111

3. NRASinmelanocyticcellneoplasms ... 112

4. NRAS:prognosticorpredictivebiomarkerinmelanoma?acriticalanalysisofcurrentliterature... 113

4.1. IsNRASaprognosticbiomarkerinmelanoma?... 113

4.2. IsNRASapredictivebiomarkerinmelanoma?... 115

5. NRASasamechanismofresistancetoBRAFinhibitorsinmelanoma... 116

6. Futuredirections... 118

Reviewers... 119

References... 119

Biographies ... 122

Abstract

RASbelongstotheguanosine5-triphosphate(GTP)-bindingproteins’family,andoncogenicmutationsincodons12,13,or61ofRAS familyoccurinapproximatelyonethirdofallhumancancerswithN-RASmutationsfoundinabout15–20%ofmelanomas.Theimportance ofRASsignalingasapotentialtargetincancerisemphasizednotonlybytheprevalenceofRASmutations,butalsobythehighnumber ofRASactivatorsandeffectorsidentifiedinmammaliancellsthatplacestheRASproteinsatthecrossroadsofseveral,importantsignaling networks.Rasproteinsarecrucialcrossroadsofsignalingpathwaysthatlinktheactivationofcellsurfacereceptorswithawidevarietyof cellularprocessesleadingtothecontrolofproliferation,apoptosisanddifferentiation.Furthermore,oncogenicrasproteinsinterferewith metabolismoftumorcells,microenvironment’sremodeling,evasionoftheimmuneresponse,andfinallycontributestothemetastaticprocess. After40yearsofbasic,translationalandclinicalresearch,muchisnowknownaboutthemolecularmechanismsbywhichthesemonomeric guanosinetriphosphatase-bindingproteinspromotecellularmalignancy,anditisclearthattheyregulatesignalingpathwaysinvolvedinthe controlofcellproliferation,survival,andinvasiveness.InthisreviewwesummarizethebiologicalroleofRASincancerbyfocusingour attentiononthebiologicalrationalandstrategiestotargetRASinmelanoma.

©2014ElsevierIrelandLtd.Allrightsreserved.

Keywords:NRAS;Melanoma;Prognostic;Predictive;Resistance

∗_{Correspondingauthorat:UnitofMedicalOncology,DepartmentofOncologyandHaematology,PapaGiovanniXXIIIHospital,PiazzaOMS1,Bergamo}

24100,Italy.Tel.:+390352673687;fax:+390352674985.

E-mailaddress:mariomandala@tin.it(M.Mandalà).

http://dx.doi.org/10.1016/j.critrevonc.2014.05.005

Ras proteins are crucial crossroads of signaling

path-ways that link the activation of cell surface receptors

with a wide variety of cellular processes leading to the

control of proliferation, apoptosis anddifferentiation.

Fur-thermore,oncogenicrasproteinsinterferewithmetabolism

oftumorcells,microenvironment’sremodeling,evasionof

theimmuneresponse,andfinallycontributestothemetastatic process.

TheimportanceofRASsignalingasapotentialtargetin

cancer is emphasized not onlyby the prevalence of RAS

mutations,butalso bythe highnumberof RASactivators

andeffectorsidentifiedinmammaliancellsthat placesthe

RASproteinsatthecrossroadsofseveral,importantsignaling networks.

After40yearsofbasic,translationalandclinicalresearch,

much is now known about the molecularmechanisms by

which these monomeric guanosine triphosphatase-binding

proteinspromotecellularmalignancy,anditisclearthatthey

regulatesignalingpathways involvedinthe controlof cell

proliferation, survival,andinvasiveness. In thisreviewwe

summarizethebiologicalroleofRASincancerbyfocusing

ourattentiononthebiologicalrationalandstrategiestotarget

RASinmelanoma.

1. HistoricalperspectiveonNRASincanceranda focusonmelanoma

Thirty yearsago a pioneering study demonstrated that

small fragmentsofDNA fromhumancancer-derivedcells

couldinducemalignantcharacteristicsinmousefibroblasts

[1].ThecellularhomologofanoncogenefoundintheHarvey

ratsarcomaretrovirus(H-RAS)wasidentifiedas theDNA

sequenceresponsibleforsuch malignanttransformation.A

newstepintumorbiologyhadbeenputinplace:thiswasthe

first demonstrationthat humantumors containedactivated

oncogenes,relatedtothosepickedupbyretrovirusesfrom

theirhostgenomes[2,3].Genesequencingrevealedthatthe

differencebetweenthewild-type(wt)humanH-RASgene

andtheoncogenicformfoundintumorswasasinglepoint

mutation.Subsequently, threeRAS genesand

correspond-ingproteinsweredescribed:N-RAS(neuroblastoma-RAS),

H-RASandK-RAS(Kirsten-RAS)[4–6].

RAS belongs to the guanosine 5-triphosphate

(GTP)-binding proteins’ family. When acted upon by specific

factors,suchasextracellularligandsthatbindspecific

mem-brane receptors,theseproteins cycle between anactivated

and inactivated form, RAS-GTP and RAS-GDP,

respec-tively[7].Activationrequiresdissociationofproteinbound

GDP, aprocess that is accelerated byguanine

nucleotide-exchangefactors(GEFs).Thisswitch-onprocessinvolvesthe

reversibleexchangeofGDPforGTP.Theswitch-offprocess

isentirelydifferentandinvolveshydrolysisofGTPtoGDP,

the guanosine triphosphatase (GTPase) reaction, which is

basicallyirreversible.ThisprocessisacceleratedbyGTPase activatingproteins(GAPs)(Fig.1a).

Inphysiologicalconditions,RASproteinsaretetheredto

the innercellmembrane,coupling growthfactorreceptors

to downstream signalingpathways andregulate important

cellularfunctionssuchascellgrowth,proliferation,and

sur-vival.Muchisnowknownaboutthemolecularmechanisms

bywhichthesemonomericguanosinetriphosphatase-binding

proteinspromotecellularmalignancy,anditisclearthatthey

regulate signalingpathwaysinvolved inthecontrol ofcell

proliferation,survival,andinvasiveness.

Mutationsat positions12, 13,or 61of the H-RAS,

N-RAS,andK-RASimpairtheGTPaseactivityofthecarrier

RAS proteinsandlockthemintoaconstitutivelyactivated

stateinwhichtheyelicitdownstreameffectors,eveninthe absenceofligandsthatbindspecificmembranereceptors[8]. Thispeculiaroncogenicactivation–disablingtheenzymatic

activity –differentiatesRASfromotheroncogenickinases

such as EGFR or B-RAF, whichare typically mutated to

produceahyperactiveenzyme.

TheimportanceofRASsignalingasapotentialtargetin

cancer is emphasizednot only by the prevalence of RAS

mutations, butalsoby thehighnumberof RAS activators

andeffectorsidentified inmammaliancells thatplaces the

RASproteinsatthecrossroadsofseveral,importantsignaling networks(Fig.1b).

The first RAS effector identified is the RAF

ser-ine/threonine kinase [9–12].Activation of RAFinitiates a

phosphorylationcascadethat progressesthrough MEKand

ERK(p42/p44MAPK),andultimatelyleadstofine

adjust-mentsindownstreamtargetsthatregulatecellproliferation,

survival, and differentiation [13]. A second RAS effector

isthe p110catalyticsubunitof thephosphatidylinositol 3-kinase(PI3K)[14].Phosphorylationofphosphatidylinositol byPI3KbringstheAKTserine/threoninekinasetotheplasma

membrane,whereitbecomesactivatedandtransmits

down-streamsignalstoregulatecellsurvival,proteinsynthesis,and

metabolism[15].RAFandPI3Karealsocommonlymutated

in melanoma, suggesting that theseproteins might be the

primaryoncogeniceffectorsofRASsignaling[16].

Interestingly,whilemutationsinRAFandRASare

gen-erally mutually exclusive, this is not the case for PI3K

mutation.Thesebiologicaldifferencessuggestthat

endoge-nouslevelsofactivatedRASdonotefficientlyactivatePI3K

signaling,whileRASandRAFmutationsappearfunctionally

equivalent.AnotherexplanationisthattheRAS/RAFdouble

mutationislethalforthecellwhereasRAS/PI3Karenot.

Since oncogenic mutations incodons 12, 13, or 61 of

RASfamilyoccurinapproximatelyonethirdofallhuman

cancerswithN-RASmutationsfoundinabout 15–20%of

melanomas,RASandthesignalingpathwaysunderits

con-trol have been kept firmly in focus as therapeutic targets

(Fig.2).However,after40yearsofresearch,manyproblems

remain open.First,whathaspreventedthedevelopmentof

drugsagainstRAS?

Several factors have hampered the development of

therapies that are able to inhibit RAS in a specific and

Fig.1.(a)MechanismofRASactivation.Receptortyrosinekinase(RTK)-mediatedactivationrequiresdissociationofproteinboundGDP,aprocessthatis

acceleratedbyguaninenucleotide-exchangefactors(GEFs).Thisswitch-onprocessinvolvesthereversibleexchangeofGDPforGTP.Theswitch-offprocess

isentirelydifferentandinvolveshydrolysisofGTPtoGDP,theguanosinetriphosphatase(GTPase)reaction,whichisbasicallyirreversible.Thisprocess

isacceleratedbyGTPaseactivatingproteins(GAPs).(b)EffectorsidentifiedinmammaliancellsthatplacetheRASproteinsatthecrossroadsofseveral,

importantsignalingnetworks.(TIAM1:T-celllymphomainvasionandmetastasis1;PI3K:phosphoinositide3-kinase;PDK1:phosphoinositide-dependent

kinase-1;ERK:extracellularregulatedkinase;RALGDS:RALguaninenucleotidedissociationstimulator;PLD:phospholipaseD;PLC␧:phospholipaseC␧;

PKC:proteinkinaseC).

highintracellularconcentrationsofGTP;(3)theattemptto

inhibit farnesylation, a key posttranslational modification

step of RAS that is essential for RAS function, through

thefarnesyltransferaseinhibitors(FTIs),was ineffectivein

clinical trials;(4) targetingmutant N-RASwith siRNAis

still limited to preclinical models because of the

signifi-cant challenge in delivering antisense oligonucleotides in

vivo.

In this review we summarize the biological role of

RASincancerbyfocusing ourattention onthe biological

rationalandstrategiestotargetRASinmelanoma.Forthis

purpose,weperformedanextensive“Medline”andCancerlit

literature review (1995–2012). Various combinations of

searchterms were used depending onthe requirements of

thedatabasebeingsearched.Thesetermsincluded“RAS”,

“MAPK”, “target therapy”, “MEK” in combination with

“cancer patients”, “melanoma”, “incidence”,

“pathogen-esis”, “management”, “cancer”, “tumors”, “resistance”,

“trials”,“prospective”,“phase”,“retrospective”.Inaddition,

we manually researched all relevant review articles and

the references of the retrieved papers. Finally, trials were excludedifrelevantdatacouldnotbeextracted.

2. BiologicalfunctionsofNRAS

Ras proteins are crucial crossroads of signaling

path-waysthat linkthe activationof cellsurfacereceptorswith awidevarietyofcellularprocessesleadingtothecontrolof proliferation,apoptosisanddifferentiation(Fig.3).

Further-more, oncogenic ras proteinsinterferewithmetabolismof

tumorcells,microenvironment’sremodeling,evasionofthe

immune response,andfinallycontributes tothemetastatic

process.

2.1. Cellproliferation

Three decades ago Feramisco et al. demonstrated that

Fig.2.Potentialtherapeuticstargetsinmelanoma.

Fig.3.Rasproteinsarecrucialcrossroadsofsignalingpathwaysthatlinktheactivationofcellsurfacereceptorswithawidevarietyofcellularprocesses

bymicroinjectionintoavarietyof somaticcells determine

dramaticmorphologicalchangesfollowedbytransient cell

proliferation [17]. Proliferation is a check and balances

process,being the result of different stimuli, that elicit or inhibitcellcycle[18].OncogenicRASfuelscell

prolifera-tionthroughfourdistinctbiologicalmechanismsthat carry

thebalanceofdifferentstimulitohangonthesideofthecell

cycle:upregulationofgrowthfactors,expressionofgrowth

factorreceptors,upregulationofintegrinsthatpromote pro-liferation anddownregulation of anti-proliferative signals. Thesecomplexandstillunclarifiedmechanismsleadto

acti-vationofseveraltranscriptionfactorssuch serumresponse

factor(SRF),JUN,activatingtranscriptionfactor2(ATF2)

andnuclearfactor-␬B(NF-␬B)[19,20].Inturn,thesefactors triggertheexpressionofcyclinD1[21].Theexpressionofthe

G1cyclinseemsacrucialdeterminantofRAS-induced

trans-formation.IthasbeenreportedthatcyclinD1-deficientmice areresistanttodevelopingepithelialtumorsthatareinduced

bytheHRASoncogene.Pharmacologicalinterferencewith

cyclinD1orcyclin-dependentkinaseinhibitors(CKIs),such

asp27andp21,whichwouldotherwiseassociatewithand

inhibitcyclin-dependentkinases(CDKs),couldbean

excit-ingavenueofcancerresearchinthecomingyears.

2.2. Suppressionofapoptosis

OncogenicRASmayhavebothpro-apoptoticand

anti-apoptoticfunctions.Theanti-apoptoticfunctionofoncogenic

RAS is mediated by several effector pathways, including

theRAS–PI3KandtheRAS–RAFpathway.Bothpathways

have been implicated in phosphorylating and inactivating

thepro-apoptoticprotein BCL-2-associatedagonist of cell

death(BAD).There isevidencethat RASisimplicated in

both the development and maintenance of melanoma. In

experimental models, melanoma genesis and maintenance

arestrictlydependentuponexpressionofHRasV12Gandon

theoppositeHRasV12Gdown-regulationresultsinclinical

andhistologicalregressionofprimaryandexplantedtumors

[22].Theinitialstagesofregressioninvolvedmarked apopto-sisinthetumorcellsandhost-derivedendothelialcells.These

dataclearlysupport thehypothesis of anoncogenic

RAS-drivenerosionoftheapoptoticpathwaysanditscontribution

tomelanomadevelopment.

2.3. Metabolism

RAS-drivenactivationofMAPKandPI3Keffector

path-waysstimulatemTORactivitywhich,inturn,up-regulates

thehypoxia-induciblefactor1␣(HIF1␣),whichiswell

rec-ognizedforitsabilitytostimulateaglycolyticshift[23].RAS

dependentupregulationofHIF1␣enhancesthetranscription

oftheglucosetransporterGLUT1,thusconferringcellswith anincreasedcapacitytotakeupglucose.Inaddition, onco-genicRASleadstoanincreaseinthelevelsofkeyglycolytic

enzymes[24].Thus,oncogenicRASdirectlycontributesto

metabolicreactionsthatstimulatetheuseof glucoseasan

anabolicsubstrateinproducingbuildingmaterialforcellular

growth.OncogenicRASinterfaceswithcellularmetabolism

andthisinteractionincreasesultimatelythe glycolyticrate andcellularviability,supportingtumorgrowthinvivo[25].

2.4. Remodelingthemicroenvironment

RASactivationsustainspro-angiogenicprocessesthrough

modulation of endothelial growth factors levels,

enhance-ment of local inflammation and stromal remodeling [26].

RAS upregulates VEGFA via multiple effectors,

includ-ing,HIF1␣,cyclooxygenase2(COX2)andprostaglandins’

production[27].Furthermore,RAS-mediatedproductionof

pro-inflammatory cytokines, such as IL-6 and IL-8, has

emergedasanothercontributortotheinductionof angiogen-esis[28].Finally,upregulationofmatrixmetalloproteinase2

(MMP2),MMP9andurokinase-typeplasminogenactivator

(uPA)hasbeendescribed[29].

2.5. Evasionoftheimmuneresponse

Oncogenic RAS can disrupt antitumor immunity by

essentially two mechanisms: first,by reducing the surface

expression of antigen-presenting major histocompatibility

complexes (MHC) on tumor cells, resulting in decreased

immunogenicityoftheRAS-transformedcells[30].Second,

byovercominghost-protecting adaptiveimmuneresponses

[31].Upon oncogenic RAS expression, the recruitmentof

immunosuppressiveregulatoryTcellsandmyeloid-derived

suppressor cells attumorsitemay leadtoa compromised

antitumorimmuneresponse[32].

2.6. Metastasis

Metastasisisamulti-stageprocessinvolvingamultitude of cellular activities such as cancer cell motility, intrava-sation, transit intheblood orlymph vessels,extravasation

andgrowthatanewsite.RASpromotestheseprocessesby

engaging adiverseandbroad platform of effector

mecha-nisms.OncogenicRASinducesalterationsincell–celland

cell–matrix interactionsandtheacquisitionof amigratory

phenotypeultimatelycontributingtothemetastaticprocess.

OncogenicRASreducesE-cadherinlevelsandinducesthe

destabilization of E-cadherin – ␤-catenin complexes and

the ␤ catenin relocalization [33]. In addition, oncogenic

RAScontributestotheenhancedmotilityoftumorcellsby

affecting changes in the polymerization, organization and

contraction of actin;the polymerization and/or stabilityof microtubules;andthetranscriptionalregulationofmitogenic geneproducts[34].OncogenicRASprotectstumorcellsfrom matrixdeprivation-inducedapoptosis,oranoikisthereby con-tributingtotheircapacityofmigrationthroughthecirculatory system[33–35].

3. NRASinmelanocyticcellneoplasms

Oneof the unresolvedissues concerningthe oncogenic

activationofRASpertainstowhetherspecificoncogenic

out-putsare drivenbymutations inaparticular RASisoform.

This hypothesisis supported by the well-recognized

non-random distribution pattern of activated isoforms of RAS

amongdifferentcancertypes.

NRAS mutations have been found in approximately

15–20% of human melanomas while HRAS and KRAS

mutationsarerare(1%)[36].Arationalexplanationforthe greateroccurrenceofNRASmutationsreliesondistinct

dif-ferences between the signaling capabilities of NRAS and

KRASinmelanocytes [37].Whenthe transformation

effi-ciencies of mutant NRAS and KRAS were compared in

immortal,non-transformedInk4a/Arf-deficientmelanocytes,

it was shown that in contrast to KRAS mutation, NRAS

mutationleadstoincreasedcellularproliferationandismore

potently tumorigenic [37]. Furthermore, NRAS mediates

activation of bothMAPK andPI3K/AKT/MYCsignaling.

Specifically, although both NRAS and KRAS efficiently

activate the classical MAPK pathway, only NRAS

effec-tivelypreventsglycogensynthasekinase3(GSK3)-mediated

phosphorylation of Myc via PI3K/AKT, which results in

enhancedactivityofendogenousMycprotein[37].In

con-trasttoKRAS,NRASandHRASalsoshowamorepotent

activation of PI3K/AKT likely due to the fact that both

NRASandHRAScolocalizetolipidrafts,whereasKRAS

isexcludedfromlipidraftsandlocalizes tothedisordered plasmamembrane[38],resultinginalessefficientactivation oralimitedaccesstoadefinedsubsetofdownstreameffector proteins.

ThereisagreatdebatewhetherspecificRASisoforms dic-tatespecificclinico-pathologicalmelanocyticcellneoplasms.

AnextraordinarilyhighNRASmutationfrequencyseemsto

becharacteristicofmedium-sized(≥1.5cm)andlarge-giant

congenital nevi whereas common acquired nevi and Spitz

nevihaverareNRASmutations(4.6%and4%,respectively)

[39].

ThefrequencyofNRASmutationsinmedium-sized

con-genital nevi is 64–70% [39–41] and raises to 94.7% in

large-giant congenitalnevi whereithasbeen recently

rec-ognizedas thesolerecurrentsomatic mutation[42].It has

beensuggestedthatNRASmutationsexertstrongergrowth

signals,resultingintheformationoflargernevi thanthose linkedtoBRAFmutations[43].Incontrast,smallcongenital nevi(<1.5cm)aregeneticallysimilartocommonacquired

neviandtendtoshowalowerincidenceofNRASmutations

andhigherincidenceofBRAFmutations[40].Inaddition,

ithasbeen reportedthatnevi thatdisplayhistological

fea-tures frequently found in nevi present at birth (so-called

“congenital pattern nevi”) but lack a definitive history of

presenceatbirthshowedonly25%ofNRASmutationsand

71%of BRAFmutations[44].NRASmutationswerealso

foundin48%to70%ofproliferativenodulesthatdeveloped

withincongenitalneviearlyinlife,butthepresenceofsuch

mutationsdoesnotseemtoconferanincreasedriskof malig-nanttransformation[44,45].

Recently, differentstudieshavedemonstratedthat early

embryonic/post zygotic somatic mutations in the NRAS

gene areimplicated inthe developmentofneurocutaneous

melanocytosis,ararecongenitaldisorder,inwhichaffected patientshaveanincreasednumberofmelanocytesinthe lep-tomeningesandtheskin,withalargecongenitalmelanocytic

nevus usually associated with so-called “satellites” in the

vicinity, and childhood melanoma of the central nervous

system[46–48].Inlinewiththeseobservations,recentlyit

hasbeenshownthatprimarymelanomaoftheCNSin

chil-drencarriesoncogenicmutationsinNRAS,unlikeprimary

melanomaofthecentralnervoussysteminadults,inwhich

NRASisnotacommondriveroncogene[46].

So-called “dysplasticnevi”donotseemtocarryNRAS

mutations[49–51].However,inanotherstudy5/7

“dysplas-tic nevi”from individualswithahereditarypredisposition

to melanoma (whocarried germline CDKN2Amutations)

were reported to be NRAS mutated and it was suggested

thatNRASmutationsareimplicatedduringearlymelanoma

development[52].Overall,giventhelimitednumberofcases

analyzed and the lack of interobserver agreement for the

morphology-based diagnosis of “dysplastic nevi” it is too

early to draw significant conclusions. A recent study has

shownthatnevus-associatedmelanomasshowasimilar

fre-quencyofBRAFV600-andNRASQ61-mutationscompared

topublishedreportsofmelanomasoftheskiningeneral[53].

SuchresultsdonotsupporttheconceptthatoncogenicBRAF

orNRASmutationsplayamajorroleinthedevelopmentof

melanomafromnevianddonotsustainthemultisteptheory

ofmelanomaprogressionfromabenignmelanocyticnevus

through“dysplasticnevus”andeventuallytomelanoma[53].

RAShasbeenextensivelyinvestigatedinmelanomaand

severalstudieshaveassessedwhetherspecificRASisoforms

correlatewithrace,patternofsunexposure,clinical

presen-tation,andconventionalmorphologicalfeatures,whichare

commonlyreportedinhistopathologicalreports.

NRASismutatedinapproximately15–20%of primary

cutaneousmelanomasinCaucasianpatients[54–58].Inblack

Africans andAsianpopulationsthereisalowerfrequency

(12%and7.2%,respectively)[59,60].Patientswith

NRAS-mutatedmelanomaswerereportedtobeolderincomparison

withindividualswithBRAF-mutatedtumors[61]althoughin

arecentmeta-analysison31studiesinvolving1972patients,

noassociationbetweenageandNRASmutationswasfound

[55].Similarly,nocorrelationwasfoundbetweengenderand

NRASmutations[58].

In moststudies,NRASmutationwassignificantlymore

frequentinmelanomasarisinginchronicsun-damagedskin

[55,62].TheincidenceoftheNRASmutationaccordingto tumorsitewashighestintheextremities(25%),followedby

thefaceorscalp(18%)andtrunk(18%)[55,61,63].NRAS

mutationshavealsobeenfoundinconjunctivalmelanomas

(18% frequency) [63], sinonasal melanomas (22%) [65],

1, which is a rare mutation site for cutaneous melanoma

[66]. Interestingly, melanoma of unknown primary sites

showedNRAS mutationsin32% of casesassociated with

highsomaticmutationrates,highratiosofC>T/G>A

transi-tions,anda45%ofBRAFmutations,collectivelyindicating

a mutation profile consistent with cutaneous sun-exposed

melanomas[67].

NRASmutationsareoverallmorefrequentlyevidentin

patients with nodular melanoma [55]. From 25% to 31%

of NRAS mutations occurred in this melanoma subtype

[55,59,68].AhigherincidenceofNRASmutationswasfound

in non-acral fast growing melanomas in comparison with

non-fastgrowingmelanomas(26.5versus12.1%)[69].

WhileinsomestudiesNRASmutatedmelanomaswere

reportedtobesignificantlythickerandhigherClark’slevel

thanwttumors [61,62,64,68]otherreports could not

con-firm any association between NRAS mutation and tumor

thickness[70,71]. Ulceration was reported to be lower in

NRAS-mutated tumors in comparison to BRAF mutated

tumors(9.7versus22.4%,respectively)[63]butnoobvious

effectofmutationalstatusonthepresenceofulcerationwas

reportedbyothers[68].MelanomasharboringNRAS

muta-tionshave showngreater mitotic rates thanBRAF mutant

melanomas[63,68].

Inconclusion,theretrospectivenatureofthestudiesand

theheterogeneity of patients’ populationsmayexplainthe

differentresultsobtainedso far,anditshouldbe

acknowl-edgedthatphenotypic-genotypic correlationsinmelanoma

isstillaworkinprogress.

4. NRAS:prognosticorpredictivebiomarkerin melanoma?acriticalanalysisofcurrentliterature

Theprognostic andpredictive significanceof NRAS in

melanomaisstillamatterofintensedebate.

Abiomarker is,by definition, an objectively measured

andevaluated parameter that provides information on the

naturalhistoryofaspecificdisease,itspathogenicprocess

oronpharmacologicalresponsestoaspecifiedtherapeutic

intervention.Aprognosticbiomarkerprovides information

onoverallcanceroutcome,regardlessoftherapy.Inthe med-icalliteraturetwotypesofprognosticbiomarkershavebeen

reported:biomarkersthatgiveinformationonrecurrencein

patientswhoreceivecurativetreatment andthosethat

cor-relatewiththemedianoverallsurvival(OS)inpatientswith

metastaticdisease.AccordingtoaNIHConsensus

Confer-ence,aclinicalusefulprognosticmarkermustbeaproven

independent,significantfactor,thatiseasytodetermineand interpretandhastherapeuticconsequences[72].

Prognosticbiomarkersthatprovideinformationontherisk ofrelapseareimportantnotonlytobetterstratifypatientsin clinicaltrialsbutalsotosparemanypatientsthe

treatment-relatedtoxicitywithoutcompromisingsurvival.Abiomarker

withpredictive value gives information on the effectof a

therapeuticinterventioninapatient.Twotypesofpredictive

biomarkershavebeenreported:(1)upfrontand(2)early pre-dictive markers.The firstcanbe usedfor patientselection

andthe second provides informationearly duringtherapy.

Thelatterbiomarkerislessusefulthantheformerbecause

doesnotprovidereliableandusefulinformationtoselectthe beststrategytobeadoptedbeforestartingtherapy.

4.1. IsNRASaprognosticbiomarkerinmelanoma?

Severalstudieshavebeencarriedouttoexaminewhether

mutationsinNRASconferdifferentpathologicalfeaturesand clinicalbehavior. Theeffectof thesemutationsonclinical

outcome remainsuncertain[59,61,73,74].Table1

summa-rizesmostimportantstudiesontheprognosticroleofNRAS

inmelanoma[63,68,74–79,61,80–83].Themajorityofthese

studieshavebeenretrospectiveinnature,andmostofthem

includedpatientswithrecurrentormetastaticdisease.

WhenOSwasmeasuredfromthetimeofprimarytumor,

NRAS mutations were found to have no impact on OS

[59,63,61]. Akslen et al. evaluated 51 primary nodular

melanomas.InthisretrospectivestudyNRASmutationwas

foundin27%ofpatients[82].RASmutationwasnot associ-atedwithtumorcellproliferationbyKi-67expression,tumor thickness,microvesseldensity,orvascularinvasion,andthere werenodifferencesinpatientsurvival[82].

In anattempttocorrelateBRAFandNRASmutational

status with features known to influence tumor behavior,

including age, gender,Breslow depth, Clark level, mitotic

rate, the presence of ulceration, and AJCCstaging,

Eller-horstetal.performedastudyon223microdissectedprimary

melanomas[63].Patientswhosetumorscarriedeither

muta-tion presented with more advanced stages compared to

patientswithwttumors,andspecifically,weremorelikelyto

haveStageIIIdiseaseatdiagnosis.BRAFandNRAS

muta-tionsdid notinfluencesurvival.Furthermore, inthisstudy survivaldidnotdifferbetweenStageIIIpatientswhose

pri-marytumorsdoordonotcarrymutations,eventhoughthe

mutatedtumorstendedtoproducelargervolumenodal

dis-ease[63].

Recently, Devitt et al. reported data obtained from a

prospectivecohortof249patients[67].Whencomparedtowt

NRASpatients,multivariateanalysisofmelanoma-specific

survival identified NRASmutationsas anadverse

progno-sticfactor.Howeverinthemultivariate analysis,therewas

noevidencethatNRASmutationwasneitheranindependent

predictorofrelapsefreesurvival(RFS)norofOS[68].

However,intwostudieswhereOSwasmeasuredfromthe

timeofbiopsyofadvanceddisease,NRASmutationswere

associatedwithimprovedOSwhencomparedtotumorswith

BRAFmutationsorbothBRAF/NRASwttumors[73,74].

Mann et al. performed a comprehensive

clinico-pathologicalassessmentoffresh-frozenmacroscopicnodal

metastases and the preceding primary melanoma, somatic

mutation profiling, and gene expression profiling to

iden-tifydeterminantsofoutcomein79melanomapatients[81].

Table1

SummaryofmostsignificantstudiesaddressingtheprognosticsignificanceofNRASmutationsinmelanoma.

RASmutationandmelanomaprognosis

Author Patientsno. Stage Siteofprimary

melanoma

Genes Exons PFS OS

Demunter(2001)[75] 81 Allstages Skin NRAS 1 p=0.0130 –

Omholt(2002)[78] 72 Allstages Skin NRAS 2

3

– NS

Houben(2004)[76] 174 Allstages Skin BRAF 15

11

NS NS

p=0.02a

NRAS 1

2

Akslen(2005)[81] 57 Allstages Skin BRAF 15

11 – NS NRAS 2 1 Edlundh-Rose(2006) [79] 219 NA Skin BRAF 15 11 – NS NRAS 2 Ugurel(2007)[73] 109 III IV Skin Mucosa Occult NA BRAF 15 11 – p=0.006 NRAS 2 1

Ellerhorst(2010)[62] 223 I–III Skin BRAF 15 – NS

NRAS 2

Devitt(2011)[67] 244 I–III Skin BRAF 15 – p=0.04

(MSS)

NRAS 3

Jakob(2012)[77] 667 Allstages Skin

Mucosa Uvea Occult BRAF 15 – p=0.004 NRAS 1 2

Mann(2012)[80] 79 III Skin BRAF 15 – NS

NRAS 2 Bucheit(2013)[61] 438 IV Skin Mucosa Softparts Occult BRAF 15 – NS NRAS 1 2 Birkeland(2013)[74] 85 III IV Skin Mucosa Uvea Occult NRAS 3 p<0.01 p<0.001

Ekedahl(2013)[82] 203 IV Skin BRAF 15 – p=0.25

NRAS 2

OS:overallsurvival;PFS:progressionfreesurvival;MSS:melanoma-specificsurvival;NS:notsignificant.

a_{OSfrommetastasectomy.}

NRAS mutation was independently associated with better

survival.Furthermore,a46-geneexpressionsignaturewith

strongoverrepresentationofimmuneresponsegeneswas

pre-dictiveofbettersurvival;inthefullcohort,mediansurvival

was>100monthsinthosewiththesignature,but10months inthosewithout.

Recently,inaretrospectivestudy,Jacobetal.testedfor

Table2

StudiesreportingonRASmutationsaspredictivebiomarkersinmelanoma.

RASaspredictivebiomarkerinmelanoma

Author Patients

no.

Stage Siteof

primary melanoma

Genes Mutations Drug(s) OS PFS TTP CCR/CB

Banerji(2008)

[83]

6 NR NR BRAF V600E 17-AAG NR – NR NR

NRAS G13D 17-AAG Joseph(2012) [84] 208 IIIc IV NR BRAF V600 HDIL2 NS NS NS NS – – p=0.05 NRAS G12 G13 Q61 HDIL2 Birkeland(2013) [74] 85 III IV Skin Mucosa Uvea Occult NRAS G12 G13 DTIC p<0.001 NS – NS Patelet(2013)[85] 18 III IV

Skin BRAF V600E

R603 S.+/−DTIC, TXT,E.orT. NR – NS NS NRAS Q61R Q61K G12S S.+/−DTIC, TXT,E.orT.

17-AAG:HSP90inhibitor17-allylamino-17-demethoxygeldanamycin;DTIC:dacarbazine;CB:clinicalbenefit:objectiveresponseorstablediseaserecorded

3monthsafterDTICtreatment;HDIL2:high-doseinterleukin2;CRR:clinicalresponserate;TTP:timetoprogression;NR:notreported;S.:selumetinib;

TXT.:docetaxel;E.:erlotinib;T.:temsirolimus.

significantassociationsofmutationwithtumorandpatient

characteristicsandwithsurvivalfromthediagnosisofstage

IVdisease[78].Tumormutationstatuswasassociatedwith

theriskofcentralnervoussysteminvolvementatthe

diag-nosis.PatientswithNRASmutationshadamediansurvival

of8.2monthsfrom stageIVdiagnosis, whichwasshorter

thanthe median survival of wtpatients (15.1 months).At

multivariateanalysis,afteradjustingforage,sex,metastatic

site,serumlactatedehydrogenaselevel,NRASmutationwas

independentlyassociatedwithdecreasedOS.

Overall, the results published so far are heterogeneous

in terms of patients’ selection criteria and methodology.

Specifically,difficultiesincomparingresultsarisefromthe followingconsiderations: (i)mostof theavailabledataare retrospective;(ii)patientswithdifferenttumorstageshave

beenevaluated; (iii) primary or metastaticsiteshavebeen

tested;(iv)differenttumorhistotypeshavebeenincluded.

Hence,thereisnodefinitiveevidencethatNRAS

muta-tionisprognosticinpatientswithlimitedradicallyresected disease(stagesI–III) orinmetastaticsetting.Furthermore,

mostoftheobservationshavebeenconductedinCaucasian

populationswithscarcityofdatafromothergeographicareas (e.g.Asian).

4.2. IsNRASapredictivebiomarkerinmelanoma?

TheRASmutationalstatusdoesnotgiveinformationon

theeffectofatherapeuticinterventioninapatient,henceitis notapredictivemarkereitherupfrontorasearlypredictive

marker. Table 2 includes studiesaddressing the predictive

significanceofNRASmutationsinmelanoma[75,84–86].

So far, several different strategies of directly targeting RAShavenotresultedineffectivetherapeutics.Thereis

evi-dence that someNRAS-mutated celllines are sensitiveto

MEK inhibition invitro[87].However, inthismodel, the

sensitivitytotheMEKinhibitorofN-RASmutatedcellswas

significantlylowerthanthoseharboringBRAFmutation.

TheloweractivityofMEKinhibitorsinN-RAS-mutated

in comparison with BRAFV600-mutated melanoma cells

maybeexplainedbythecomplexityofpathwayswithwhich

RASinteractswithinthecell.

Itis wellknown thatRAS familymembershave

multi-ple other targets, such as PI(3)K andRalGDS; thesemay

exert more prominent oncogenic effects in certain tumor

subtypes,therebyreducingtherequirementforMAPK

acti-vation.Hence,single-agenttherapeuticstrategiesmayprove

insufficient in RAS mutant tumors. Instead, direct RAS

inhibitorsorcombinatorialstrategiesmayberequired.

Recently, anoral MEK inhibitor (MEK162)was tested

in patients with metastaticmelanoma harboring BRAF or

NRASmutationswithencouragingresultsinNRASmutated

patients [88]. In preclinical models MEK162 inhibited

growth of NRAS-mutated and Val600Glu BRAF-mutated

melanoma instudiesthat usedinvitroandinvivomodels

[89].

However,theresponseratewasreportedinonly20%of

patients and only in 10% of this population the response

was confirmed. Furthermore, the median progression-free

responsewas7.6weeks[88].Thesedataclearlyindicatethat

mostofthepatients rapidlydevelopresistancetotheMEK

inhibitor.

Atwo-arm,randomized, prospective,open-label,

multi-center,phaseIIIstudytocomparetheefficacyandsafetyof

MEK162(45mgbisindie)versusdacarbazine(1000mg/m2

IVevery 3 weeks) inpatients withadvanced (Stage IIIC)

unresectableormetastatic(StageIV)NRASQ61

mutation-positive cutaneous melanoma is currently underway. The

primaryendpointofthestudyisprogression-freesurvival,

whilesecondaryendpointisoverallsurvival(“NEMOtrial”

NCT01763164).

Anothersecond generationMEKinhibitor,selumetinib,

demonstratedmarkedinhibitionofpERK,eitherincelllines

harboring BRAF mutations as well as in thoseharboring

NRASmutations[90].

A randomized phase II study comparing the MEK

inhibitor Pimasertib (AS703026) with dacarbazine in

pre-viously untreated subjects with N-Ras mutated locally

advanced or metastatic malignant cutaneous melanoma is

currentlyunderway(NCT01693068).

Atthetimeofthepublicationofthismanuscriptthereare

norandomizedclinicaltrialscomparingMEK162withother

MEKinhibitorsinNRASmutatedmelanomapatients.

Recently thedevelopment of small moleculesthat

irre-versiblybindtoacommononcogenicmutant,K-Ras(G12C)

has been reported [91]. These compounds rely on the

mutantcysteineforbindingandthereforedonotaffectthe

wt protein. These inhibitors to K-Ras(G12C) subvert the

native nucleotide preference tofavor GDP over GTP and

impairingbindingtoRaf.Thesefindingsarerelevantsince

they reveal, for the first time, a new allosteric regulatory

site on Ras that is targetable in a mutant-specific

man-ner.

AsubgroupofmelanomaswithRASdependenceisthose

withlow-activity.

BRAFmutations,such asthosefoundatpositions466,

464and597.Celllineswithlow-activity BRAFmutations

showanimpairedactivationofMAPKsignalinginisolated

kinaseassaysandoftenharborconcurrentNRASmutations

at positions 12and 13. It cannot be excludedthat NRAS

melanomacellswithlowactivitymutantBRAFmaypartially

explainthesensitivityofasubgroupofNRASmelanomacells

toMEKinhibitors.

Inaccordance withthishypothesis,Dahlmanetal.

per-formedananalysisofBRAFexon15in49tumorswithlack

ofBRAFV600 mutationandshowedthat 2(4%)harbored

L597mutationsandother2 BRAFD594 andK601

muta-tions[92].InvitrosignalinginducedbyL597mutantswas

suppressedbyMEKinhibition.ApatientwithBRAFL597S

mutantmetastaticmelanomarespondedsignificantlyto

treat-mentwiththeMEKinhibitor,TAK-733.Collectively,these

data show clinical significance response to BRAF(L597)

mutationsinmelanoma.

ThefocusofindirectRASinhibitionhasthenshiftedto

interferewiththecomplexnetworkofactivateddownstream

cascades such as the MAPK, phosphoinositol 3-kinase

(PI3K),phospholipidC(PLC),RalGEF.

Posch et al. evaluated the sensitivity of RAS mutated

melanomacellsandxenograftstoMEKandPI3Kinhibitors

[93]. NRAS mutated cells were more sensitive to MEK

inhibition compared withthe PI3K/mTORcascade

inhibi-tion.CombinedtargetingofMEKandPI3Kwassuperiorto

MEKandmTORinhibitioninallNRASmutantmelanoma

celllines,suggestingthatPI3Ksignalingismoreimportant

for cell survival in NRAS mutant melanoma when MEK

is inhibited. However, targeting of PI3K/mTOR in

com-bination with MEK inhibitors is necessary to effectively

abolishgrowthofNRASmutantmelanomacellsinvitroand

regressxenograftedNRASmutantmelanoma.Inthismodel

MEK and PI3K/mTOR inhibition was synergistic. These

results indicate that combined targeting of the MEK/ERK

andPI3K/mTORpathwayshasantitumoractivityandcould

beavalidoptioninthetreatmentofNRASmutantmelanoma,

forwhichtherearecurrentlynoeffectivetherapies.

Finally, Johnson et al. reported that patients with

NRASmutatedmetastaticmelanomaachieveincreased

clin-ical benefit from immunotherapy compared to thosewith

BRAF/NRASwt[94].

These datasuggest that NRAS mutation statusmay be

a biomarker of response to immunotherapy in metastatic

melanomaandthatmolecularlytargetedimmunotherapymay

befeasible.Howeveralarger,prospectiveanalysisis neces-sarytovalidateandexpandontheseresults,includingthose

withBRAFmutandKITmutmetastaticmelanomatodraw

firmconclusions.

Overall, the above data suggest that:(i) a subgroup of

NRASmutatedmelanomamaybesensitivetoMEK

inhibi-tionbutinmostcasesresistancerapidlyoccur;(ii)asubgroup

of NRAS mutated melanoma harbor low activity BRAF

mutation,andthemeaningofthesemutationsshouldbe

fur-therinvestigated;(iii)single-agenttherapeuticstrategiesmay

proveinsufficientinRASmutanttumors.Instead,

combina-torialstrategiesmayberequiredtoovercomeresistance.

5. NRASasamechanismofresistancetoBRAF inhibitorsinmelanoma

AhighpercentageofpatientswithBRAFV600Emutant

melanomasrespondtoselectiveRAFinhibitorsbutresistance

eventuallyemerges.

Unlike what happens in othertumors whereadditional

mutationseventuallyoccurinthetarget(EGFRinnon-small

cell lung cancer, c-KIT in GISTs, BCR-ABL in chronic

myeloidleukemia)theearlyevidencefromdirect

sequenc-ingofBRAFexonssuggeststhatnewpointmutationsarenot

evidentandthatBRAFV600Epersists.

RAShasbeenconsistentlydescribedasamechanismof

resistancetoBRAFinhibitors.Itiswellknownthatthereisa

switchinRAFisoformusagedependingonwhetherBRAF

Table3

ResistancetoBRAFinhibitors.

ResistancetoBRAFinhibitors

Author Patientswithacquired

resistancetotherapyno.

Drug Mechanismofresistance NRASacquired

mutationspatientsno.(%)

Nazarian(2010)[95] 12 Vemurafenib NRAScodon61

mutations

PDGFRBoverexpression

1(8.3%)

Trunzer(2013)[96] 13 Vemurafenib IncreasedpERKlevels

MEK1mutations

NRAScodon61

mutations

3(23%)

McArthur(2011)[97] 11 Vemurafenib NRAScodon61

mutations

1(9.09%)

Poulikakos(2011)[98] 19 Vemurafenib IncreasedRAS-GTP

levels Increased RAS-independentRAF dimerization 6(31.6%) Wagle(2014)[99] 5 Dabrafenib Trametinib MutationinMEK2

BRAFspliceisoform

BRAFamplification

5(100%)

VanAllen(2014)[100] 30 Vemurafenib

Dabrafenib MAPKpathway Alterations MEK1Mutations MEK2Mutations MIFTAmplification 23/45(51%)

Rizos(2014)[101] 38 MutationinMEK2

MutationinMEK1

MutationinNRAS

MutationinAKT

BRAFspliceisoform

BRAFamplification

3(8%)

whichBRAFismutated,BRAFisprimarilyresponsiblefor

signalingtoMEKandERK.In presenceofRASmutation

anexcessiveERKsignalingthrough BRAFandingeneral

MAPKactivation would inducecellcycle arrest or

senes-cencethroughtranscriptionalup-regulationofproteinssuch

as p21, p27, and p16INK4A [102].To avoidthis, the cells

switch to CRAF,which provides weaker signaling andis

compatiblewithtumorprogression.

Nazarianetal.demonstratedthathighlevelsofactivated

N-RASresultingfrommutationsleadtosignificantMAPK

pathwayreactivationuponBRAFinhibitortreatment[95].In

aseriesofelegantexperiments,knockdownofNRASreduced

growth of the respective BRAF inhibitors resistant cells.

Ontheopposite,overexpressionofN-RASconferredBRAF

inhibitorresistancetoBRAFinhibitorsensitiveparentalcell lines.

Recently,Suetal.usedcelllinestoestablishBRAFV600E

melanoma clones with acquired resistance to a BRAF

inhibitor [103].The authors confirmedthat nosecond-site

mutationscouldbeidentifiedintheBRAFcodingsequence.

Inthismodel,resistancecorrelatedwithincreasedlevelsof

RAS-GTP,andsequencingofRASgenesrevealedarare

acti-vatingmutationinKRAS,resultinginaK117Nchange in

theKRASprotein.Elevatedlevels ofCRAFand

phospho-rylatedAKTwerealsoobserved.Interestingly,combination

treatmentwithBRAFinhibitorandeitheraMEKinhibitor

oranAKTinhibitorsynergisticallyinhibitedproliferationof resistantcells.Thesedatasupportclinicalstudiesinwhich

combination therapy with other targeted agents are being

strategizedtoovercomeresistance.

Trunzer et al. [96] evaluated serial biopsies to study

changes in mitogen-activated protein kinase (MAPK)

signaling,cell-cycleprogression,andfactorscausing

intrin-sicor acquired resistancebyimmunohistochemistry,DNA

sequencing, or somatic mutation profiling to a BRAF

inhibitorwithintheBRIM2study[104].Inthisstudy3/13

patients hadNRASQ61Kco-occurring mutationsintumor

samplestakenatprogression.Combiningthesefindingswith

thosepreviously reportedby Nazarian[95] andMcArthur

[97],among36patientsanalyzed,fivepatients(14%)hadan

NRASmutationinaprogressivelesion.Thisfurthersupports thehypothesisbyNazarianetal.[95]thattheNRASmutation

isonemechanismofescapefromvemurafenibtherapy.

Overall,theabovereporteddatasuggestthat:(1)A

con-comitantbaselinemutationintheupstreamNRASoncogene

israrebutmayresultinearlylackofclinicalbenefittoBRAFi;

(2) RAS mutation is a common mechanism of acquired

resistance; (3) whether a combination therapy with other

targetedagentscouldovercomeresistanceremainstobe elu-cidated.

Table4

RAStargetinlocallyadvancedormetastaticmelanoma:ongoingclinicaltrials(www.clinicaltrials.govaccessedJanuary26,2014).

Drug Phase Trial Disease(s) Primaryoutcomemeasures

Monotherapynon-randomized

MEK162 II NCT01320085 BRAForNRAS

Mutatedmelanoma ORR RAF265 II NCT00304525 Melanoma MTD DLT Associationmutationsin NRAS/clinicalresponse(◦)

Selumetinib(AZD6244) II NCT00866177 BRAForNRAS

Mutatedmelanoma

Anti-tumorresponse

Monotherapyrandomized

Pimasertibversusdacarbazina II NCT01693068 NRASmutatedmelanoma PFS

MEK162versusdacarbazine III NCT01763164 NRASmutatedmelanoma PFS

AZD6244versustemozolamide II NCT00338130 Melanoma PFS

ORR* TTD

Durationofresponse

Assessmentoftheefficacyof

AZD6244versus

temozolomideBRAFor

NRASMMpatients(◦)

Combinationtherapynon-randomized

BKM120+MEK162 I NCT01363232 EGFRmutantNSCLCinPD

onEGFRinhibitors

Triplenegativebreastcancer

Pancreaticcancer

CRC Melanoma

NSCLC,withKRAS,NRAS,

and/orBRAFmutations

DLT

Trametinib

(GSK1120212)+GSK2141795

II NCT01941927 BRAFwtmelanoma ORR*inpatientswitheither

mutatedNRASorwt NRAS/wtBRAFM RAFinhibitor (BMS-908662)+immunotherapy (ipilimumab) I NCT01245556 Melanoma Toxicity PDwillbeassessedby evaluatingmarkersof

RAS/RAFpathwayactivity

(◦)

PI3K/mTORinhibitor

BEZ235+MEK1/2inhibitor

MEK162

Ib NCT01337765 EGFRmutantNSCLCinPD

onEGFRinhibitors

Triplenegativebreastcancer

Pancreaticcancer

Colorectalcancer

Melanoma NSCLC

Otheradvancedsolidtumors

withKRAS,NRAS,and/or

BRAFmutations

IncidenceofDLT

LEE011+MEK162 Ib NCT01781572 NRASmutatedmelanoma IncidenceofDLT

ORR*

MTD:maximumtolerateddose;EAS:ectopicACTHsecreting;wt:wildtype;NR:notreported;ORR:overallresponserate;PFS:progressionfreesurvival;

TTD:timetodeath;DLT:doselimitingtoxicity;ORR*:objectiveresponserate;CRC:colorectalcancer;(◦):secondaryoutcomemeasures.

The importance of RASin melanoma deserves clinical

andbiologicalinvestigationtooptimizetreatmentoflocally

advancedandmetastaticmelanoma.Although,inthelasttwo

decades,progresshasbeenslow,therearenowavarietyof

therapeuticstrategiesthatareprimedfor clinical

investiga-tion.Table4summarizesongoingtrialsinwhichRAS,and

preferentiallyNRAS,hasbeenselectedasatarget.

6. Futuredirections

Thirty yearsof basic, clinicalandtranslationalresearch

have producedalargeamount of knowledgepertainingto

theRASoncogenefamily(Fig.4).TheprevalenceofRAS

mutations, but also the high number of RAS activators

Fig.4. TimelineofkeyadvancesinNRASclinicalandtranslationalresearch.(a)Harveyetal.(1964)[97];(b)Kirstenetal.(1970)[98];(c)Shihetal.(1980)

[99];(d)Changetal.(1982)[2,3],(e)Deretal.(1982)[4],(f)Paradaetal.(1982)[5],(g)Santosetal.(1982)[100],(h)Milburnetal.(1990)[103];(i)Pai

etal.(1990)[102];(j)Boguskietal.(1993)[101];(k)Moodieetal.(1993)[9];(l)Vojteketal.(1993)[10];(m)Warneetal.(1993)[11],(n)Zhangetal.(1993)

[12];(o)Chinetal.(1999)[22];(p)Solitetal.(2006)[86];(q)Nazarianetal.(2010)[90];(r)Whitwametal.(2007)[37];(s)Asciertoetal.(2013)[87],(t)

Poschetal.(2013)[89].

proteinsatthecrossroadsofseveralsignalingnetworks.

Nev-ertheless, this extensive knowledge has not yet translated

intoclinicallyeffectivetherapiesformelanomasexpressing

mutantformsofRAS.

AsRAS is mutated in15–20% of melanomas, priority

actionsareneeded:

1. Future studies should focus on co-extinction strategies

otherthanreinforcinginhibitionofMAPKsignaling.

2. Inhibitionoftheactivateddownstreamcascadesincluding

MAPK,PI3K,PLC,RALshouldbepursuedinpreclinical

andearlyphaseclinicalstudies.

3. Mostof the downstreamtargets are not tumor specific

therapiesandbeartheriskofseveresideeffects.Hence,

well-designed clinical studies with appropriate

phar-macokineticsandpharmacodynamic endpoint between

combinationtherapiesareneeded.

4. MEKinhibitors as monotherapyshouldbe validated in

prospective,randomizedphaseIIIstudies.

Reviewers

OlivierMichielin,MD,PhD,CHUV–Multidisciplinary

OncologyCenter(CePO),UNIL–UniversityofLausanne,

Lausanne,Switzerland.

Matteo Carlino, PhD, Westmead Institute of Cancer

Research,TheUniversityofSydney,Australia.

References

[1]MalumbresM,BarbacidM.RASoncogenes:thefirst30years.Nat RevCancer2003;3:459–65.

[2]ChangEH,FurthME,ScolnickEM,LowyDR.Tumorigenic trans-formationofmammalian cellsinduced byanormalhumangene homologoustotheoncogeneofHarveymurinesarcomavirus.Nature 1982;297:479–83.

[3]ChangEH,GondaMA,EllisRW,ScolnickEM,LowyDR.Human genomecontainsfourgeneshomologoustotransforminggenesof HarveyandKirstenmurinesarcomaviruses.ProcNatlAcadSciUS A1982;79:4848–52.

[4]DerCJ,KrontirisTG,CooperGM.Transforminggenesofhuman bladderandlungcarcinomacelllinesarehomologoustotherasgenes ofHarveyandKirstensarcomaviruses.ProcNatlAcadSciUSA 1982;79:3637–40.

[5]ParadaLF,TabinCJ,ShihC,WeinbergRA.HumanEJbladder car-cinomaoncogeneishomologueofHarveysarcomavirusrasgene. Nature1982;297:474–8.

[6]ShimizuK,GoldfarbM,SuardY,etal.Threehumantransforming genesarerelatedtotheviralrasoncogenes.ProcNatlAcadSciUS A1983;80:2112–6.

[7]Henis YI, Hancock JF, Prior IA.Ras acylation, compartmental-izationandsignalingnanoclusters.MolMembrBiol2009;26:80–92 [Review].

[8]MorA,PhilipsMR.CompartmentalizedRas/MAPKsignaling.Annu RevImmunol2006;24:771–800.

[9]MoodieSA,WillumsenBM,WeberMJ,WolfmanA.Complexesof Ras.GTPwithRaf-1andmitogen-activatedproteinkinasekinase. Science1993;260:1658–61.

[10]Vojtek AB, Hollenberg SM, Cooper JA. Mammalian Ras inter-acts directlywith theserine/threonine kinaseRaf. Cell 1993;74: 205–14.

[11]WarnePH,VicianaPR,DownwardJ.DirectinteractionofRasand theamino-terminalregionofRaf-1invitro.Nature1993;364:352–5.

[12]ZhangXF,SettlemanJ,KyriakisJM,etal.Normalandoncogenic p21rasproteinsbindtotheamino-terminalregulatorydomainof c-Raf-1.Nature1993;364:308–13.

[13]Seger R, Krebs EG. The MAPK signaling cascade. FASEB J 1995;9:726–35.

[14]Rodriguez-VicianaP,WarnePH,DhandR,etal. Phosphatidylinositol-3-OHkinaseasadirecttargetofRas.Nature1994;370:527–32.

[15]Engelman JA,LuoJ,CantleyLC.Theevolutionof phosphatidyl-inositol3-kinasesasregulatorsofgrowthandmetabolism.NatRev Genet2006;7:606–19.

[16]MandalàM,VoitC.TargetingBRAFinmelanoma:biologicaland clinicalchallenges.CritRevOncolHematol2013;87:239–55.

[17]FeramiscoJR,GrossM,KamataT,RosenbergM,SweetRW. Microin-jectionoftheoncogeneformofthehumanH-ras(T-24)proteinresults inrapidproliferationofquiescentcells.Cell1984;38:109–17.

[18]HanahanD,WeinbergRA.Hallmarksofcancer:thenextgeneration. Cell2011;144:646–74.

[19]FincoTS,WestwickJK,NorrisJL,BegAA,DerCJ,BaldwinJrAS. OncogenicHa-Ras-inducedsignalingactivatesNF-kappaB transcrip-tionalactivity,whichisrequiredforcellulartransformation.JBiol Chem1997;272:24113–6.

[20]WestwickJK,CoxAD,DerCJ,etal.OncogenicRasactivatesc-Jun viaaseparatepathwayfromtheactivationofextracellular signal-regulatedkinases.ProcNatlAcadSciUSA1994;91:6030–4.

[21]WinstonJT,CoatsSR,WangYZ,PledgerWJ.Regulationofthecell cyclemachinerybyoncogenicras.Oncogene1996;12:127–34.

[22]ChinL,TamA,PomerantzJ,etal.EssentialroleforoncogenicRas intumourmaintenance.Nature1999;400:468–72.

[23]Johannessen CM, Reczek EE, James MF, Brems H, Legius E, CichowskiK.TheNF1tumorsuppressorcriticallyregulatesTSC2 andmTOR.ProcNatlAcadSciUSA2005;102:8573–8.

[24]Semenza GL. Hypoxia and cancer. Cancer Metastasis Rev 2007;26:223–4.

[25]Chiaradonna F, Sacco E, Manzoni R, Giorgio M, Vanoni M, AlberghinaL.Ras-dependentcarbonmetabolismandtransformation inmousefibroblasts.Oncogene2006;25:5391–404.

[26]KranenburgO,GebbinkMF,VoestEE.Stimulationofangiogenesis byRasproteins.BiochimBiophysActa2004;1654:23–37.

[27]TsujiiM,KawanoS,TsujiS,SawaokaH,HoriM,DuBoisRN. Cyclooxygenase regulates angiogenesis induced by coloncancer cells.Cell1998;93:705–16.

[28]Ancrile BB, O’HayerKM, Counter CM.Oncogenic ras-induced expressionofcytokines:anewtargetofanti-cancertherapeutics.Mol Interv2008;8:22–7.

[29]BlasiF,CarmelietP.uPAR:aversatilesignallingorchestrator.Nat RevMolCellBiol2002;3:932–43.

[30]SeligerB,HardersC,WollscheidU,StaegeMS,Reske-KunzAB, HuberC.SuppressionofMHCclassIantigensinoncogenic trans-formants: association with decreased recognition bycytotoxic T lymphocytes.ExpHematol1996;24:1275–9.

[31]KubuschokB,NeumannF,BreitR,etal.NaturallyoccurringT-cell responseagainstmutatedp21rasoncoproteininpancreaticcancer. ClinCancerRes2006;12:1365–72.

[32]TranThangNN,DerouaziM,PhilippinG,etal.Immuneinfiltration ofspontaneousmouseastrocytomasisdominatedby immunosup-pressivecellsfromearlystagesoftumordevelopment.CancerRes 2010;70:4829–39.

[33]FrischSM,FrancisH.Disruptionofepithelialcell–matrixinteractions inducesapoptosis.JCellBiol1994;124:619–26.

[34]CampbellPM, DerCJ.OncogenicRasanditsroleintumorcell invasionandmetastasis.SeminCancerBiol2004;14:105–14.

[35]GiehlK.OncogenicRasintumourprogressionandmetastasis.Biol Chem2005;386:193–205.

[36]ChinL,GarrawayLA,FisherDE.Malignantmelanoma:geneticsand therapeuticsinthegenomicera.GenesDev2006;20:2149–82.

[37]WhitwamT,VanbrocklinMW,RussoME,etal.Differential onco-genicpotentialofactivatedRASisoformsinmelanocytes.Oncogene 2007;26:4563–70.

[38]PriorIA,HancockJF.CompartmentalizationofRasproteins.JCell Sci2001;114(Pt9):1603–8.

[39]RossAL,SanchezMI,GrichnikJM.Molecularnevogenesis. Derma-tolResPract2011;2011:463184.

[40]Ichii-NakatoN,TakataM,TakayanagiS,etal.Highfrequencyof BRAFV600Emutationinacquiredneviandsmallcongenitalnevi, butlowfrequencyofmutationinmedium-sizedcongenitalnevi.J InvestDermatol2006;126:2111–8.

[41]DessarsB,DeRaeveLE,MorandiniR,etal.Genotypicandgene expressionstudiesincongenitalmelanocyticnevi:insightinto ini-tial steps of melanotumorigenesis. J Invest Dermatol 2009;129: 139–47.

[42]CharbelC,FontaineRH,MaloufGG,etal.NRASmutationisthe solerecurrentsomaticmutationinlargecongenitalmelanocyticnevi. JInvestDermatol2013;(October).

[43]Ichii-NakatoN1,TakataM,TakayanagiS,etal.Highfrequencyof BRAFV600Emutationinacquiredneviandsmallcongenitalnevi, butlowfrequencyofmutationinmedium-sizedcongenitalnevi.J InvestDermatol2006;126:2111–8.

[44]BauerJ,CurtinJA,PinkelD,BastianBC.Congenitalmelanocytic nevifrequentlyharborNRASmutationsbutnoBRAFmutations.J InvestDermatol2007;127:179–82.

[45]PhadkePA,RakhejaD,LeLP,etal.Proliferativenodulesarising within congenital melanocytic nevi: a histologic, immunohisto-chemical,andmolecularanalysesof43cases.AmJSurg Pathol 2011;35:656–69.

[46]Pedersen M, Küsters-Vandevelde HV, Viros A, et al. Primary melanomaoftheCNSinchildrenisdrivenbycongenitalexpression ofoncogenicNRASinmelanocytes.CancerDiscov2013;3:458–69.

[47]GessiM,HammesJ,LauriolaL,etal.GNA11andN-RASmutations: alternativesforMAPKpathwayactivatingGNAQmutationsin pri-marymelanocytictumoursofthecentralnervoussystem.Neuropathol ApplNeurobiol2013;39:417–25.

[48]Kinsler VA, Thomas AC, Ishida M, et al. Multiple congenital melanocytic nevi and neurocutaneous melanosis are caused by postzygotic mutations in codon 61of NRAS.J InvestDermatol 2013;133:2229–36.

[49]CarrJ,MackieRM.PointmutationsintheN-rasoncogenein malig-nantmelanomaandcongenitalnaevi.BrJDermatol1994;131:72–7.

[50]JafariM,PappT,KirchnerS,etal.Analysisofrasmutationsinhuman melanocyticlesions:activationoftherasgeneseemstobeassociated withthenodulartypeofhumanmalignantmelanoma.JCancerRes ClinOncol1995;121:23–30.

[51]AlbinoAP,NanusDM,MentleIR,etal.Analysisofrasoncogenesin malignantmelanomaandprecursorlesions:correlationofpoint muta-tionswithdifferentiationphenotype.Oncogene1989;4:1363–74.

[52]EskandarpourM,HashemiJ,KanterL,RingborgU,PlatzA, Hans-sonJ.FrequencyofUV-inducibleNRASmutationsinmelanomas ofpatientswithgermlineCDKN2Amutations.JNatlCancerInst 2003;95:790–8.

[53]TschandlP,BerghoffAS,PreusserM,etal.NRASandBRAF muta-tionsinmelanoma-associatednevianduninvolvednevi.PLOSONE 2013;8:e69639.

[54]CurtinJA,FridlyandJ,KageshitaT,etal.Distinctsetsofgenetic alterationsinmelanoma.NEnglJMed2005;353:2135–47.

[55]LeeJH,ChoiJW,KimYS.FrequenciesofBRAFandNRAS muta-tionsaredifferentinhistologicaltypesandsitesoforiginofcutaneous melanoma:ameta-analysis.BrJDermatol2011;164:776–84.

[56]LovlyCM,DahlmanKB,FohnLE,etal.Routinemultiplex muta-tionalprofilingofmelanomasenablesenrollmentingenotype-driven therapeutictrials.PLOSONE2012;7:e35309.

[57]HenaryH,HongDS,FalchookGS,etal.Melanomapatientsina phaseIclinic:molecularaberrations,targetedtherapyandoutcomes. AnnOncol2013;24:2158–65.

[58]Platz A, Egyhazi S, Ringborg U,Hansson J.Human cutaneous melanoma; a review of NRAS and BRAF mutationfrequencies in relation to histogenetic subclass and body site. Mol Oncol 2008;1:395–405.

[59]Akslen LA,PuntervollH, BachmannIM, etal.Mutation analy-sisoftheEGFR-NRAS-BRAFpathwayinmelanomasfromblack Africansandothersubgroupsofcutaneousmelanoma.Melanoma Res2008;18:29–35.

[60]SiL,KongY,XuX,etal.PrevalenceofBRAFV600Emutationin Chinesemelanomapatients:largescaleanalysisofBRAFandNRAS mutationsina432-casecohort.EurJCancer2012;48:94–100.

[61]Edlundh-RoseE,EgyháziS,OmholtK,etal.NRASandBRAF muta-tionsinmelanomatumoursinrelationtoclinicalcharacteristics:a studybasedonmutationscreeningbypyrosequencing.Melanoma Res2006;16:471–8.

[62]BallNJ,YohnJJ,MorelliJG,NorrisDA,GolitzLE,HoefflerJP.Ras mutationsinhumanmelanoma:amarkerofmalignantprogression.J InvestDermatol1994;102:285–90.

[63]EllerhorstJA,GreeneVR,EkmekciogluS,etal.Clinicalcorrelates ofNRASandBRAFmutationsinprimaryhumanmelanoma.Clin CancerRes2011;17:229–35.

[64]Griewank KG, Westekemper H, Murali R, et al. Conjunctival melanomasharborBRAFandNRASmutationsandcopynumber changessimilartocutaneousandmucosalmelanomas.ClinCancer Res2013;19:3143–52.

[65]Turri-ZanoniM,MedicinaD,LombardiD,etal.Sinonasalmucosal melanoma: molecularprofileand therapeuticimplications froma seriesof32cases.HeadNeck2013;35:1066–77.

[66]SekineS,NakanishiY,OgawaR,KoudaS,KanaiY.Esophageal melanomasharborfrequentNRASmutationsunlikemelanomasof othermucosalsites.VirchowsArch2009;454:513–7.

[67]Dutton-Regester K, Kakavand H,Aoude LG, et al. Melanomas of unknown primary have a mutation profile consistent with cutaneous sun-exposed melanoma. Pigment Cell Melanoma Res 2013;26:852–60.

[68]DevittB,LiuW,SalemiR,etal.Clinicaloutcomeand pathologi-calfeaturesassociatedwithNRASmutationincutaneousmelanoma. PigmentCellMelanomaRes2011;24:666–72.

[69]Nagore E,HackerE, Martorell-CalatayudA,etal.Prevalenceof BRAFandNRASmutationsinfast-growingmelanomas.Pigment CellMelanomaRes2013;26:429–31.

[70]vanElsasA,ZerpSF,vanderFlierS,etal.Relevanceof ultraviolet-inducedN-rasoncogenepointmutationsindevelopmentofprimary humancutaneousmelanoma.AmJPathol1996;149:883–93.

[71]Demunter A, Stas M,Degreef H, De Wolf-Peeters C, van den Oord JJ. Analysis of N-and K-rasmutations in the distinctive

tumorprogressionphasesofmelanoma.JInvestDermatol2001;117: 1483–9.

[72]Treatmentofearly-stagebreastcancer.NIHconsensusconference. JAMA1991;265:391–5.

[73]Omholt K, Platz A, Kanter L, Ringborg U, Hansson J. NRAS and BRAF mutations ariseearly during melanoma pathogenesis andarepreservedthroughouttumorprogression.ClinCancerRes 2003;9:6483–8.

[74]UgurelS,ThirumaranRK,BloethnerS,etal.B-RAFandN-RAS mutationsarepreservedduringshorttimeinvitropropagationand differentiallyimpactprognosis.PLoSONE2007;2:e236.

[75]BirkelandE, BuschC, BergeEO, etal.LowBRAFand NRAS expressionlevels are associatedwith clinical benefitfrom DTIC therapyandprognosisinmetastaticmelanoma.ClinExpMetastasis 2013;30:867–76.

[76]DemunterA,AhmadianMR,LibbrechtL,etal.AnovelNrasmutation inmalignantmelanomaisassociatedwithexcellentprognosis.Cancer Res2001;61:4916–22.

[77]HoubenR,BeckerJC,KappelA,etal.Constitutiveactivationofthe Ras-Rafsignalingpathwayinmetastaticmelanomaisassociatedwith poorprognosis.JCarcinog2004;3:6.

[78]JakobJA,BassettJrRL, NgCS,etal.NRASmutationstatusis anindependentprognostic factorinmetastaticmelanoma.Cancer 2012;118:4014–23.

[79]OmholtK,KarsbergS,PlatzA,KanterL,RingborgU,Hansson J.ScreeningofN-rascodon61mutations inpaired primaryand metastaticcutaneousmelanomas:mutationsoccurearlyandpersist throughouttumorprogression.ClinCancerRes2002;8:3468–74.

[80]BucheitAD,SyklawerE,JakobJA,etal.Clinicalcharacteristicsand outcomeswithspecificBRAFandNRASmutationsinpatientswith metastaticmelanoma.Cancer2013Nov1;119:3821–9.

[81]MannGJ,PupoGM,CampainAE,etal.BRAFmutation,NRAS mutation,andtheabsenceofanimmune-relatedexpressedgene pro-filepredictpooroutcomeinpatientswithstageIIImelanoma.JInvest Dermatol2013;133:509–17.

[82]AkslenLA,AngeliniS,StraumeO,etal.BRAFandNRAS muta-tions are frequent in nodular melanoma but are not associated withtumorcellproliferationorpatientsurvival.JInvestDermatol 2005;125:312–7.

[83]EkedahlH,CirenajwisH,HarbstK,etal.Theclinicalsignificanceof BRAFandNRASmutationsinaclinic-basedmetastaticmelanoma cohort.BrJDermatol2013;169:1049–55.

[84]BanerjiU,AffolterA,JudsonI,MaraisR,WorkmanP.BRAFand NRASmutations in melanoma:potentialrelationships to clinical responsetoHSP90inhibitors.MolCancerTher2008;7:737–9.

[85]JosephRW, Sullivan RJ, HarrellR, et al.Correlation of NRAS mutationswithclinicalresponsetohigh-doseIL-2inpatientswith advancedmelanoma.JImmunother2012;35:66–72.

[86]PatelSP,LazarAJ,PapadopoulosNE,etal.Clinicalresponsesto selumetinib(AZD6244;ARRY-142886)-basedcombinationtherapy stratifiedbygenemutationsinpatientswithmetastaticmelanoma. Cancer2013;119:799–805.

[87]SolitDB,GarrawayLA,PratilasCA,etal.BRAFmutationpredicts sensitivitytoMEKinhibition.Nature2006;439:358–62.

[88]AsciertoPA,SchadendorfD,BerkingC,etal.MEK162forpatients withadvancedmelanomaharbouringNRASorVal600BRAF muta-tions:anon-randomised,open-labelphase 2study.LancetOncol 2013;14:249–56.

[89]WinskiS,AndersonD,BouhanaK,etal.MEK162(ARRY-162), a novel MEK 1/2 inhibitor, inhibits tumorgrowth regardless of KRas/Rafpathwaymutations.In:Proceedingsofthe22nd EORTC-NCI-AACRsymposiumonmoleculartargetsandcancertherapeutics. 2010.

[90]YehTC, Marsh V, BernatBA, et al.Biological characterization ofARRY-142886(AZD6244),a potent,highlyselective mitogen-activated protein kinase kinase 1/2 inhibitor. Clin Cancer Res 2007;13:1576–83.

[91]OstremJM,PetersU,SosML,WellsJA,ShokatKM.K-Ras(G12C) inhibitorsallostericallycontrolGTPaffinityandeffectorinteractions. Nature2013;503:548–51.

[92]DahlmanKB,XiaJ,HutchinsonK,etal.BRAF(L597)mutationsin melanomaareassociatedwithsensitivitytoMEKinhibitors.Cancer Discov2012;2:791–7.

[93]PoschC,MoslehiH,FeeneyL,etal.CombinedtargetingofMEK andPI3K/mTOReffectorpathwaysisnecessarytoeffectivelyinhibit NRASmutantmelanomainvitroandinvivo.ProcNatlAcadSciU SA2013;110:4015–20.

[94]JohnsonDB,LovlyCM,FlavinM,etal.NRASmutation:a poten-tial biomarker ofclinical responseto immune-based therapiesin metastaticmelanoma(MM).JClinOncol2013;31(Supppl.)[abstr 9019].

[95]NazarianR,ShiH,WangQ,etal.Melanomasacquireresistanceto B-RAF(V600E)inhibitionbyRTKorN-RASupregulation.Nature 2010;468:973–7.

[96]Trunzer K, Pavlick AC, Schuchter L, et al. Pharmacodynamic effectsandmechanismsofresistancetovemurafenibinpatientswith metastaticmelanoma.JClinOncol2013;31:1767–74.

[97]McArthurG,RibasA,ChapmanPB,etal.Molecularanalysesfrom aphaseItrialofvemu-rafenibtostudymechanismofaction(MOA) and resistanceinrepeatedbiopsiesfrom BRAFmutationpositive metastatic melanomapatients(pts).JClinOncol2011;29(Suppl.) [abstr8502].

[98]Poulikakos PI, Persaud Y, Janakiraman M, et al.RAF inhibitor resistance is mediated by dimerization of aberrantly spliced BRAF(V600E).Nature2011;480:387–90.

[99]WagleN,VanAllenEM,TreacyDJ,etal.MAPkinasepathway alter-ationsinBRAF-mutantmelanomapatientswithacquiredresistance tocombinedRAF/MEKinhibition.CancerDiscov2014;4:61–8.

[100]VanAllenEM,WagleN,SuckerA,etal.Thegeneticlandscapeof

clinicalresistancetoRAFinhibitioninmetastaticmelanoma.Cancer Discov2014;4:94–109.

[101]RizosH,MenziesAM,PupoGM,etal.BRAFinhibitorresistance

mechanismsinmetastaticmelanoma:spectrumandclinicalimpact. ClinCancerRes2014;20:1965–77.

[102]DumazN,HaywardR,MartinJ,etal.Inmelanoma,RASmutations

areaccompaniedbyswitchingsignalingfromBRAFtoCRAFand disruptedcyclicAMPsignaling.CancerRes2006;66:9483–91.

[103]SuF,BradleyWD,WangQ,etal.ResistancetoselectiveBRAF

inhibitioncanbemediatedbymodestupstreampathwayactivation. CancerRes2012;72:969–78.

[104]SosmanJA,KimKB,SchuchterL,etal.SurvivalinBRAF

V600-mutantadvancedmelanomatreatedwithvemurafenib.NEnglJMed 2012;366:707–14.

Biographies

MarioMandalàiscurrentlyamedical oncologistinthe

UnitofMedicalOncology,PapaGiovanniXXIIIHospitalin

Bergamo,Italy.HeisinchargeoftheClinicaland

Transla-tionalResearchUnit.HereceivedhisMDfromtheCatholic

University inRome,Italy in1995.He hascompleted

resi-denciesinMedicalOncologyattheCatholicUniversityin

RomeaswellasatMilanUniversityandtheEuropean

Insti-tuteofOncologyinMilan,Italy.Healsoobtaineddiplomas

inMedicalOncologyandHaematologyfromMilan

Univer-sity in2000 and2005,respectively. His researchinterests

include clinical and translational research on cancer and

thrombosis.Hismainoncologicalresearchfocuseson

gas-trointestinal cancerandmelanoma.He isafullmemberof

the European Society of Medical Oncology. He has

lead-or co-authored many scientific papers and is a reviewer

for several journals, including Lancet Oncology, Cancer,

British Journal of Cancer, Annals of Oncology, Cancer

TreatmentReview,JournalofThrombosisandHaemostasis,

Thrombosis and Hemostasis, Arteriosclerosis-Thrombosis

and Vascular Biology, Critical Reviews in Oncology and

Haematology.

DanielaMassi,MD,PhD,isassociateprofessorof pathol-ogyattheUniversityofFlorenceMedicalSchool,Italy.Her

pathologytrainingincludesadermatopathologyfellowship

attheInstituteforDermatopathology,ThomasJefferson Uni-versity,Philadelphia,PA,underthedirectionofA.Bernard Ackerman.Shethenreceivedapost-doctoralresearch

fellow-shipfromtheAmerican-ItalianCancerFoundation(A.I.C.F.)

with aresearchprogram incutaneousmelanoma. She has

beenmemberoftheExecutiveCommitteeoftheInternational

SocietyofDermatopathology(1997–2003),Chairmanofthe

DermatopathologyWorkingGroupoftheEuropeanSociety

of Pathology(2007–2011),andsheiscurrentlymemberof

the EORTC Melanoma Pathology Group. She is an

asso-ciate editor of VirchowsArchiv andscientificreviewer for

severalinternationalscientificjournals.Her research

inter-ests are focused onskintumorpathology,andparticularly

receptorsignalingandmoleculargeneticsofmelanoma.She

has authored more than 200 publications and contributed

to the volume ‘Pathology & Genetics of Skin Tumours’

of theWorldHealthOrganization(WHO)Classification of