Exploring Biomarkers in Head and Neck Cancer

ExploringBiomarkersinHeadandNeckCancer

CoreyJ.Langer,MD

Personalizedmedicinebasedonpredictivemarkerslinkedtodrugresponse,itishoped,willleadtoimprovementsinoutcomesandavoidanceofunnecessarytreatmentinsquamouscellcarcinomaoftheheadandneck(SCCHN).RecentresearchhasshownthatexpressionofERCC1maypredictresistancetotreatmentwithplatinumagents.Futuretestingforthismarkermayhelpselecttheoptimaltypeofchemotherapy.Infectionwithhumanpapillomavi-rus(HPV)isassociatedwithlessaggressivediseaseandbetterprognosisinlocallyadvancedSCCHNtreatedwithchemoradiationorradiationalone;HPV-positivepatientsmayultimatelybenefitfromlessintensive,lesstoxictherapy.K-RASmutations,occurringinabout40%ofcolorectalcancersandassociatedwithlackofbenefitfromepidermalgrowthfactorreceptor(EGFR)antibodiesinthisdisease,arefoundin<5%ofSCCHNpatients,makingroutinetest-ingforK-RASmutationsunwarrantedatthistime.VirtuallyallheadandnecktumorsoverexpressEGFR,whichlimitstheusefulnessofEGFRexpressionasamarkerfortreatmentselection.AlthoughtheincidenceofEGFRtyrosinekinasedomainmutationsisveryrare,abetterunderstandingoftheroleofEGFRmutations,expression,amplification,anddownstreameffectsinSCCHNmayhelpdefinetheroleofEGFRinthissetting.TheseobservationscautionagainstextrapolatingresultsobtainedwithbiomarkersinothertypesofcancertoSCCHN.Validationofeachbio-markerinthecontextofSCCHNclinicaltrialswillberequiredbeforeaspecificmarkercanbeincorporatedintodaily

C2012AmericanCancerSociety.practice.Cancer2012;118:3882-92.V

KEYWORDS:squamouscellcarcinomaoftheheadandneck,ERCC1,RRM1,humanpapillomavirus,K-RAS,epidermal

growthfactorreceptor,tubulin,cetuximab.

UnitedStatesin2010.1MostpatientsdiagnosedwithSCCHNpresentwithlocallyadvanceddisease,forwhichacom-binedmodalitytreatmentapproachwithcurativeintentisusuallyprescribed.Curerateswithcurrentcombined-modalitystrategiesarefavorable,buttherapiesareindiscriminatelyaggressive,oftenresultinginsevereshort-termtoxicityandlong-termfunctionalimpairment.Thus,aclearroleforbetterpatientselectiontoavoidunnecessarytreatmentandminimizelong-termtoxicityiswarranted.

Bytailoringtreatmenttothespecificgeneticormolecularprofileofatumorinanindividualpatient,responseandsurvivaloutcomescouldbeimprovedthroughrefinedtreatmentdecisions,treatment-relatedcostsandcomplicationsreduced,andunnecessaryinterventionsavoided.Suchrefinementsmayalsofacilitatethedevelopmentofnewer,moreeffectivetherapies.TheunderstandingofbiologicalmarkersinSCCHNisnotaswelldefinedasinothertumortypes(Table1),2suchasbreastorlungcancer,butrecentadvanceshavebroughtusclosertoprovidingpersonalizedmedicineforthesepatients.ThefollowingreviewprovidesanoverviewofsomeofthemostrelevantbiomarkersrelatedtoSCCHN.ERCC1

TheERCC1(excisionrepaircross-complementationgroup1)proteinplaysanimportantroleinrepairingDNAdamagecausedbyplatinumagents,andmaythereforebeusefulinpredictingwhichpatientswillbenefitfromplatinum-basedtherapy.PlatinumagentsworkbycreatingDNAadductsthatinhibitDNAreplicationandtranscriptionincancercells.3Cellsthatrepairthedamageefficientlyaremorelikelytoberesistanttoplatinumtherapy.4,5Removalofcisplatin-DNAadductsismediatedbythenucleotideexcisionrepair(NER)complex,amultistepprocessinvolving>30distinctproteins

Correspondingauthor:CoreyJ.Langer,MD,DivisionofHematologyandOncology,UniversityofPennsylvania,3400CivicCenterBlvd,2PCAM,Philadelphia,PA19104;Fax:(215)615-5122;corey.langer@uphs.upenn.edu

AbramsonCancerCenter,UniversityofPennsylvania,Philadelphia,Pennsylvania

DOI:10.1002/cncr.26718,Received:November11,2011;Revised:September9,2011;Accepted:October28,2011,PublishedonlineJanuary26,2012inWileyOnlineLibrary(wileyonlinelibrary.com)

Anestimated49,260newcasesofsquamouscellcarcinomaoftheheadandneck(SCCHN)werediagnosedinthe

3882Cancer

August15,2012

Table1.MostRelevantBiologicalAlterationsinHeadandNeckCancerbyFrequency2Alteration

Rate

FGFR319%CDKN2A18%H-RAS10%PIK3CA10%K-RAS

1%-4%OthermarkersERCC1highlevels71%-73%HPV-positive

18%-38%

Abbreviation:HPV,humanpapillomavirus.

thatrecognizeDNAdamage,incisethelesion,resynthe-sizeDNA,andligatetherepairpatch;1ofthekeyenzymesinthispathwayisERCC1.6,7SeveralgroupshavelinkedlowERCC1expressiontoplatinumsensitiv-ityinvitro,8,9andsomehaveshownthatblockingtheexpressionoractivityofERCC1increasessensitivitytoplatinumagents.10-12Forexample,Selvakumaranetal11demonstratedthatbymodifyingplatinum-resistantovar-iancancercellstoconstitutivelyexpressantisenseERCC1RNA,thecellsweremadetoinhibitthecapacitytorepairDNAdamageandconsequentlyreduceIC50valuesforplatinumagents.Thestudyalsoshowedthatmicetrans-plantedwithantisensecancercellslivedlongeraftercisplatintreatmentthanthosebearingcontrolcells.

Inprimarytumorsamples,pretreatmentlevelsofERCC1havebeenshowntopredictoutcomesafterplati-numtherapyinpatientswithovariancancer,13gastriccancer,14colorectalcancer,15esophagealcancer,16andnonsmallcelllungcancer(NSCLC).17-19Inlungcancer,Olaussenetal19assessedwhetherERCC1statuspredictedoutcomeafterplatinum-basedadjuvantchemotherapyinpatientswithcompletelyresectedNSCLC.ERCC1statuswasdeterminedin761samples(ofthe1867patientsoriginallyenrolledintheInternationalAdjuvantLungCancerTrial)usingasemi-quantitativeHscorebasedonthediffusenessandintensityofERCC1stainingbyimmunohistochemistry(IHC);themedianHscorewasusedasthecutoffbetweenERCC1-positive(44%)andERCC1negative(56%)disease.Plati-num-basedadjuvantchemotherapywasfoundtobebene-ficialinpatientswithERCC1-negativetumors,butnotinthosewithERCC1-positivetumors.Inthecontrolgroup,whichdidnotreceiveadjuvantchemotherapy,5-yearsur-vivalwassignificantlygreaterinpatientswithERCC1-positivediseasethaninthosewithERCC1-negativedis-ease(46%vs39%).Hence,ERCC1wasbothprognostic

Cancer

August15,2012

BiomarkersinSCCHN/Langer

andpredictive.Subsequenttothisstudy,thereliabilityandspecificityofthekeyreagentusedtomeasureERCC1expression,theantibody8.F1,hasbeencalledintoques-tion,whichwouldcastdoubtsovertherobustnessoftheseresults.20Coboetal21conducted1ofthefirstprospectiverandomizedtrialstoselecttreatmentbasedonERCC1.Inthisstudy,patientswithstageIVNSCLCwererandom-izedtoacontrolgroup,whichreceivedstandardtherapywithcisplatin/docetaxel,oranexperimentalgenotypicgroup,inwhichpatientswithlowERCC1expression,thistimemeasuredbymRNAexpression,weretreatedwithcisplatin/docetaxelandthosewithhighERCC1expressionreceivedanonplatinum-containingregimen(docetaxel/gemcitabine).Amongthe346evaluablepatients,theresponseratewasmarkedlyincreasedinthegenotypicgroup(50.7vs39.3%inthecontrolgroup;P¼.02);withinthegenotypicgroup,responseratewas53.2%forpatientswithlowERCC1and47.2%forthehighERCC1subgroup.However,nodifferenceinsur-vivalwasobservedbetweenthe2treatmentgroups,althoughthemedianfollow-uptimewasshort(9months).Althoughtheseresultsareencouragingandsug-gestthepotentialtobasetreatmentdecisionsonERCC1expressionlevelsinthefuture,thehighdropoutrate(17.6%),dueprimarilytoinsufficienttissueavailableforERCC1mRNAanalysis,suggeststhatbetterprotocolsforERCC1testingareneeded.

ERCC1inheadandneckcancer

PreliminarystudiessuggestERCC1hasaroleinpredictingpatientresponsetocisplatin-basedtherapyinSCCHN.22,23Handra-Lucaetal22assessedtheimpactofERCC1expressionlevel(measuredbyIHCwiththeaforementioned8.F1reagent)onoutcomesaftercispla-tin-basedchemotherapyinpatientswithlocallyadvanceddisease.Ofthe96pretreatmenttissuesamplesanalyzed,68(71%)hadhighexpressionofERCC1.Patientswithlowlevelsofexpressionhada4-foldgreaterchanceofachievinganobjectiveresponsetocisplatin-basedchemo-therapycomparedwiththosewithhighlevelsofexpres-sion.Afteradjustingforage,diseasestage,tumordifferentiation,andlocation,lowERCC1expressionwasasignificantpredictorofprolongedcancer-specificsurvival.ThesefindingssuggestthatpretreatmentERCC1expressionlevelsareinverselycorrelatedwithresponseandsurvivalaftercisplatin-basedtherapy.

Junetal23evaluatedwhetherpretreatmentlevelsofERCC1expression(alsoby8.F1-basedimmunostaining)

3883

ReviewArticle

predictedsurvivalinpatentswithlocallyadvancedSCCHNundergoingcisplatin-basedconcurrentchemo-radiotherapy.ERCC1expressionwasmeasuredusingIHCinformalin-fixed,paraffin-embeddedtissueblocksfrom45patients.Mosttumors(73%)hadhighERCC1expression.Withamedianfollow-upof53.6months,theestimated3-yearprogression-freesurvival(PFS)wassig-nificantlylongerinthosewithlowERCC1expressioncomparedwiththosewithhighexpression(83.3%vs49.4%).The3-yearoverallsurvival(OS)wasalsosignifi-cantlygreaterinthosewithlowERCC1expression(91.7%vs45.5%).Inamultivariateanalysis,lowERCC1expressionwasasignificantandindependentpredictorofprolongedsurvival.

Notably,theprevalenceofERCC1overexpressioninbothSCCHNstudieswassimilar(71%-73%)andcon-siderablyhigherthanthatreportedinNSCLC(approxi-mately40%)19usingasimilarscoringtechnique.ThissuggeststhattheproportionandpatternofERCC1expressionmayvarybytumortype.Moreover,theresponseratesobservedinthesestudiessuggestthatapro-portionofpatientswithhighERCC1expressionmaystillrespondtoplatinumtherapy.Thus,thesensitivityofERCC1asapredictorofplatinumresponseinSCCHNappearstoberelativelylow,andotherfactors—withintheNERpathwayorbeyond—mayalsoinfluencetumorresponse.Althoughtheargumentcouldbemadethatinthesestudiescisplatinwascombinedwith5-fluorouracilortaxanes,andthereforethemagnitudeoftheERCC1effectcouldbeatleastpartiallyconfoundedbytheuseofmultiplechemotherapeuticagents,thedifferencesinout-comesbetweentheERCC1groupsintheJunstudywereconsistentregardlessofthechemotherapeuticagentpart-neredwithcisplatin.22Fromapracticalperspective,itappearsthathighERCC1expressioninSCCHNtumorsmaysuggestalowprobabilityofbenefitingfromplatinumtherapy.Patientswiththismarkermaybecandidatesforalternativeapproaches,includingnonplatinumtherapy.However,furthervalidationisneeded,includingprospectivetrialsthatevaluatestrategiesforassigningtreatmentinSCCHNbasedonERCC1expressionstatus.ThemostcommonmethodforevaluatingERCC1mRNAexpressionisquantitativereal-timereversetranscriptionpolymerasechainreaction(PCR),whichrequiresthecollectionoffreshformalin-fixed,paraffin-embeddedtumorsam-ples.13-15,17CommercialkitsareavailablefortestingERCC1expression(viaIHCorPCR-basedtechniques)inothertumortypes,suchaslungandcoloncancer,but

3884notinSCCHN.Samplesmustbetakenfromtumorsthathavenotbeenirradiated(becausefibrosisorradiation-inducednecrosismayinterferewithinterpretation),arequirementthatmayhinderthepotentialapplicationofERCC1asabiomarkerinSCCHN.

InadditiontotheeffectofERCC1levelsonresponsetoplatinumtherapy,certainpolymorphismsontheERCC1genemayinfluenceresponseofpatientstoradiotherapy.Inastudyof108patientswithstageIISCCHNtreatedwithradiotherapyonly,Carlesetal24identifiedthehomozygousgenotypeThr259Thrassignif-icantlyassociatedwithworseoutcomes,forbothtimetoprogression(median,11.6monthsvs>85monthsforge-notypesLys259ThrorLys259Lys;P¼.00005)andOS(median,27.9monthsvs>88months;P¼.00).ThissamestudyevaluatedothergenesencodingDNArepair-relatedproteins,suchasERCC5orXPA.ApartfromERCC1,only1oftheERCC5polymorphismsshowedasignificantcorrelationwithoutcome;patientswithhomo-zygousT/Tnucleotideatcodonposition46hadshortertimetoprogression(median,56.7monthsvs>81monthsforthosewithC/TorC/C;P¼.049)andworsesurvival(median,61.0monthsvs>87months;P¼.0066).Ifvalidated,thesefindingsmayproveparticularlyrelevanttoclinicaldecisionmakinginasettingwheretheremaybeachoicebetweenradiationandsurgicalmanagement.

RRM1

TheRRM1(ribonucleotidereductaseM1)geneencodesaregulatorysubunitoftheenzymeribonucleotidereductasethatisthekeymoleculartargetofgemcitabine.18,25,26Pre-clinicalstudieshaveshownthathighlevelsorspecificmutationsonRRM1areassociatedwithgemcitabineresistance.3,27-30Clinically,lowRRM1expressionorsinglenucleo-tidepolymorphismshavebeenfoundtocorrelatewithbetteroutcomesinpatientstreatedwithgemcitabine-basedchemotherapyinseveralsolidtumors.18,29-31TheroleofRRM1indeterminingchemosensitivityinSCCHNisunclear.However,itscorrelationwithERCC1expressionandplatinumsensitivity(althoughweakerthanitsassociationwithgemcitabinesensitivity)suggestthattheroleofthisbiomarkerinSCCHNshouldbeexploredfurther.StudiesevaluatingnotonlyRRM1butmultiplerelatedgeneticmarkers,suchasRRM2andBRCA1,mayalsoprovideamorecomprehen-sivepictureoftheroleofRRM1inpredictingtreatmentoutcomes.32Cancer

August15,2012

b-TubulinIsotypes

Taxanesexerttheiranticanceractivitybybindingtob-tubulinpolymersandinhibitingmicrotubuledepoly-merizationandmitoticprogression.Inpreclinicalmodels,aberrantexpressionlevelsoftheisotypeIIIofb-tubulinhavebeenassociatedwithresistancetopaclitaxelinseveralcelltypes.33,34Intheclinic,thisrelationshiphasbeenexploredmostlyinNSCLC.Monzoetal35showedtheassociationbetweentubulinmutationsandlackofresponseandinfe-rioroutcomesin49patientswithadvancedNSCLCtreatedwithpaclitaxel.Morerecently,Seveetal36andDumontetetal37haveprovidedconsistentresultsin2groupsofpatients(91patientsand19patients,respec-tively)treatedwithtaxane-basedtherapy(paclitaxelordocetaxelincombinationwithplatinum,ordocetaxelasasingleagent).Patientswereclassifiedasloworhighb-tubulin-IIIexpressing,accordingtowhethertheirlevelsofexpression(byIHC)werebeloworabovethemedianvalueintheentirepopulation37orwhetherthenumberofstainedcellsinthesamplewas50%.36Regardlessofthenuancesofthescoringsystem,thelow-expressingpatientsexhibitedhigherresponseratesandlongerPFSandOS.

InSCCHN,docetaxelhasbecomepartofthestand-ardinductionregimenincombinationwithcisplatinand

5-fluorouracil.AcorrelativestudytoTaxotereVR(TAX)324,38thephase3trialcomparinginductionwith5-fluo-rouraciltoinductionwithcisplatinand5-fluorouracilaloneindicatedthatexpressionofb-tubulinIImeasuredbyIHC(sampleswerescoredbypercentageofstainedcellsandintensityofstaining;positiveresultswerethoseabovethemedianscoreandnegativethosebelow)maybeassociatedwithoutcomes.Inthesubsetofevaluablepatients(265of501patientsenrolled),patientswithlowexpressionhadbetterPFSandOS,regardlessoftreatmentwithorwithoutdocetaxel(OShazardratio[HR],2.39;P¼.0001),suggestingapotentialprognosticeffect.More-over,forthissamesubgroupwithlowexpression,theben-efitfromthespecificadditionofdocetaxeltotheinductiontherapyregimenseemedtobegreater,whereasinthehigh-expressiongroup,thereweremodestornodif-ferencesinoutcomesbetweenthe2inductionregimens.38AstheuseoftaxanesinSCCHNgrows,thismarkermayassumeagreaterroleintherapeuticdecisionmaking.Todate,therearenocommercialassaykitstodetermineexpressionofb-tubulinisotypes,andasdescribedabove,thereissomevariationinthescoringsystemsreportedintheliterature,indicatingthatapotentialpracticalapplica-

Cancer

August15,2012

BiomarkersinSCCHN/Langer

tionmayhavetofirstovercomesomestandardizationissues.

HumanPapillomavirus

Virtuallyallcasesofcervicalcancer,andasubsetofSCCHNtumors,arelinkedtohumanpapillomavirus(HPV)infection.39,40HPV-positivetumorsinSCCHNarealmostalwaysoropharyngeal,andthereportedHPVinfectionprevalencevarieswidely,dependingonthemethodofdetectionused.About35%ofoverallheadandnecktumorshavebeenreportedtobeHPVpositive,asdetectedbyPCR.40Specifically,HPVhasbeendetectedmostoftenincancersofthetonsils(43.6%)andthebaseofthetongue(38.4%).41OneofthelargestandmostrecentanalysesofpatientswithoropharynxcancerfromRTOG(RadiationTherapyOncologyGroup)0129demonstrateda%incidenceofHPV-positivetumorsusinginsituhybridizationtechniques,suggestingthatthecorrelationofthisinfectionandoropharynxcancermaybegreaterthanpreviouslyappreciated.42TheprevalenceofHPV-relatedSCCHNmayalsovarybygeographicregion.AglobalsurveyoforopharynxcancerfoundthattheprevalenceofHPVinfectionwas18%43;anotherreportindicatedtheprevalencewas38%andnotablyhigherinNorthAmerica(47%)thaninEurope(28%).43Furthermore,theprevalenceoftypesofSCCHNpotentiallyrelatedtoHPVinfectionappearstobeincreasingintheUnitedStates.40TheincreaseinHPV-relatedtumors(eg,baseoftongue,tonsil,orophar-ynx)hasbeennotedinrecentdecades,particularlyamongyounger-agecohorts,whereastheincidenceofcancerstypicallyunrelatedtoHPV(eg,tongue,gum,lip,otheroralcancers)hasdecreased.Thesetrendsmaybeexplainedinpartbychangesinbothsexualbehaviorandsmoking.

OfthoseheadandnecktumorsthatcontainHPVDNA,upto90%presentwiththeoncogenicvariantHPV-16.44-46Comparedwithtumorsoftheoropharynxandoralcavity,tonsillarsquamouscellcarcinomahasbeenmoststronglyandconsistentlyassociatedwithHPV-16infection;inthesetumors,molecularanalyseshaverevealedvariableHPVcopynumbers,comparedwithlowcopynumbersinothersites.46Hence,acausativerelation-shipbetweenHPVandSCCHNisplausible,ifnotlikely.HPV-16hasbeenshowntoimmortalizeepithelialcellsofbothcervicalandoralorigininvitro.47,48Thiseffectmaybemediatedbyinactivationofthetumorsuppressorpro-teinsp53andpRb.493885

ReviewArticle

Table2.SummaryofDataonthePrognosticEffectofHPVonDifferentPhase3TrialsinLocallyAdvancedSquamousCellCarcinomaoftheHeadandNeck42,52-StudyTreatmentSetting

HPVorp16Positive,%

%

OutcomeEffect,HPVPositivevsHPVNegative

2-yearOS<87.9%vs65.8%;tobaccouseinterferedwiththeprognosticeffectofHPVOSHR¼0.27;P¼.0075-yearOS,49%vs19.6%

RTOG0129

CisplatinplusstandardfractionationRTvsacceleratedfractionationRTCisplatin/RTÆtirapazamine

StandardfractionationRTvshyperfractionationvsacceleratedfractionationwithsplitvsacceleratedfractionationwithconcomitantboostRT

Inductiontherapycisplatin/5-fluorouracilÆdocetaxel

TROG02.02RTOG9003p16positive,58%;HPVpositive,28%39.5%

TAX,Taxotere32450%$82-monthOS,79%vs31%

Abbreviations:HPV,humanpapillomavirus;HR,hazardratio;OS,overallsurvival;RT,radiotherapy;RTOG,RadiationTherapyOncologyGroup;TROG¼TransTasmanOncologyGroup.

NumerousstudieshavenowshownthatHPVinfec-tionisanindependentriskfactorforSCCHN.Inanepi-demiologicalstudyofnearly900,000serumsamplesfromhealthyindividualsinNorway,Sweden,andFinland,292casesofSCCHNweredocumented,occurringafterame-dianof9.4years.50TheprevalenceofseropositivityoftheoncogenicHPV-16wasnearlytwiceashighinSCCHNpatientsthanincontrols(12%vs7%).Inanestedcase-controlstudyinvolving130consecutivepatientswithoro-pharynxcancer,theriskoforopharynxcancerwaslinkedtosexualbehavior,lifetimeexposuretoHPV-16(deter-minedbyseropositivityforHPV-16LIcapsidprotein),andthepresenceofHPVinfection(anyHPVorHPV-16).44Notably,therelationshipbetweenHPV-16andcancerriskwasindependentoftobaccoandalcoholuse.Afteradjustingforage,sex,andalcoholuse,HPV-16seropositivitywasfoundtobeanindependentpredictoroforopharynxcancerandaccountedfor55%ofallcases.

BeyonditsroleasariskfactorfordevelopingSCCHN,HPVstatusappearstodefineadistinctsubtypeofSCCHNtumors,andultimatelymayaidinshapingtreatmentchoices.RetrospectivestudieshaveconsistentlyshownthatHPV-positivetumorshavebetterresponseandsurvivalaftertreatment,possiblybecauseofthepro-pensityofthesetumorstomaintainanapoptoticresponsetoradiationandchemotherapy.51Alternatively,thesepatientstendtobeyoungerandhealthiercomparedwithotherheadandneckcancerpatients,withfewersmoking-relatedcomorbidities;however,itisimportanttonotethatinmultivariateanalysesofclinicaltrials,theprognos-ticeffectofHPVisnotdrivenbydemographicsalone.

TheimpactofHPVstatusonresponseandsurvivalwasevaluatedbytheEasternCooperativeOncologyGroup(ECOG)inaprospectivemulticenterphase2studyofchemoradiationasorgan-preservingtherapyin96patientswithresectablestageIII/IVlaryngealororo-pharynxcancer.Byusinginsituhybridizationtechniques,HPVDNAwasdetectedin40%oftumorsamples;63%oforopharyngealtumorswereHPVpositive,whereasnoneofthelaryngealtumorsevidencedHPVDNA.MostpositivecasesinvolvedHPV-16(95%).PatientswithHPV-positivetumorshadahigherresponserateafterinductionchemotherapy(82%vs55%;P¼.01)andafterchemoradiation(84%vs57%;P¼.007).OSwassignifi-cantlylongerinHPV-positivepatientsthaninHPV-neg-ativepatients(aliveat2years:95%vs62%).ThelowerdeathandprogressionriskforHPV-positivepatientswasstillpresentafteradjustingforage,diseasestage,andper-formancestatus.45Recentretrospectivestudiesoflargephase3trialshavesolidifiedtheclinicalimportanceofHPVasaprog-nosticmarkerforSCCHN(Table2).Thephase3TROG(TransTasmanOncologyGroup)02.02(HeadSTART)trialrandomizedpatientswithstageIII/IVSCCHNtoreceivecisplatin/radiotherapyÆtirapazamine;465patientshadoropharynxcancer.Oftheevaluablesamples,of195(28%)wereHPVpositive,and107of186(58%)werep16positive.Atthe2-yearmark,HPV-posi-tivetumorswereassociatedwithbetterOSversusHPV-negativetumors(94%vs77%;HR,0.27;P¼.007)aswellasimprovedfailure-freesurvival(FFS;85%vs75%;HR,0.43;P¼.035).Inaddition,patientswithp16-posi-tivetumorshadimproved2-yearOS(92%vs74%;HR,0.35;P¼.004)andimproved2-yearFFS(87%vs72%;HR,0.38;P¼.003).PatientswithbothHPVandp16-positivetumorshadincreasedOSandFFSratescomparedwithdouble-negativepatients(OS:95%vs71%;P¼.003;FFS:%vs69%;P¼.002).523886Cancer

August15,2012

Similarly,analysisfromtheRTOG0129study,whichrandomizedpatientstoreceivestandardfractionation70grays(Gy)þcisplatinoracceleratedfractionationwithconcomitantboost72Gyþcisplatin,correlatedHPVstatuswithpatientprognosis.Oftheevaluablesamples,206of323(%)wereHPVpositive,ofwhich96%wereHPV-16positive.OS,PFS,andlocoregionalfailurewereallsignificantlysuperioramongHPV-positivepatientsat2years(OS:87.9%vs65.8%;P<.001;PFS:71.8%vs50.4%;P<.001;locoregionalfailure:13.6%vs24.8%;P¼.004).Interestingly,heavysmokingstatus(!20pack-years)seemedtonegatepartiallythebenefitcon-ferredbyanHPV-positivetumorstatus;comparedwithHPV-positive/<20pack-yearpatients,HPV-positive/!20pack-yearpatientshadanOSHRof1.91(95%con-fidenceinterval[CI],1.20-3.05);forHPV-negative/<20pack-yearsubjectsOSHRwas2.25(95%CI,1.44-3.50),andforHPV-negative/!20pack-yearsubjectsOSHRwas4.30(95%CI,2.40-7.71).41Morerecently,Gillisonandcolleagues53haveper-formedanidenticalanalysisofanolderstudy,RTOG9003,arandomizedphase3trialcomparingvariousradia-tiondeliveryschedulesinlocallyadvancedSCCHN.Atotalof1068patientswererandomizedtostandardfrac-tionation,hyperfractionation,acceleratedfractionationwithsplit,oracceleratedfractionationwithconcomitant-boostradiotherapy,allwithoutchemotherapy.ThefinalanalysisdemonstratednodifferencesinPFS,locoregionalcontrol,andOSamongthetrialarms.Consistentwiththeobservationsfromchemoradiotherapytrials,theretro-spectiveanalysisof6evaluablesamplesshowedthatHPV-positivepatients(identifiedbythesurrogatemarkerp16)hada5-yearsurvivalrateof49%comparedwith19.6%forthosewhowereHPVnegative(P<.0001).

Finally,theobservationsfromthesubsetofevalu-ablepatients(111ofthe2oropharynxcancerpatientsenrolled)fromthephase3trialTAX324thatcompared2inductionchemotherapyregimens(cisplatinand5-fluo-rouracilwithorwithoutdocetaxeltotalN¼501)beforedefinitivechemoradiotherapyalsowereconsistent,show-ingastrongprognosticcorrelationwithHPVstatus.TherateofHPV-positivetumorswas50%.Atamedianfol-low-upof83months,79%ofHPV-positivepatientswerealive,versus31%ofthosewithHPV-negativetumors(P<.0001).Together,thesestudieshavedefinitivelyestablishedHPVasameaningfulprognosticmarkerfororopharyn-gealSCCHN,andaresufficienttowarrantaccountingforHPVstatusinthedesignandanalysisoffutureclinical

Cancer

August15,2012

BiomarkersinSCCHN/Langer

trialsusingcombinedorsequentialchemotherapyandradiation.Wehavenosuchprognosticdatainthesettingofsurgeryonly.

ImplicationsinSCCHN

TheprevailingconsensusatthistimeisthatpatientswithHPV-relateddiseaseshouldprobablybeconsideredadistinctentitycomparedwithpatientswithHPV-unre-lateddisease.WhethertostratifyfuturetrialsonthebasisofHPVstatusoractuallyinstituteseparatetrialsforHPV-positiveandHPV-negativepatientsremainsunclear,althoughconsensusisclearlymovingtowardseparatetrials,particularlybecausetheclinicalgoalsoftherapyaredifferent.InpatientswhoareHPVpositive,withlessaggressivecancerandreasonablygoodsurvival,thegoalistodeintensifytherapyandtherebyreducetoxicity.InHPV-negativepatients,whoseprognosisisdecidedlyworse,alternativeapproachesfocusingonnewtargetedapproachesandtolerabletherapeuticintensifica-tionaredesired.

PossiblestrategiesfordeintensificationinHPV-pos-itivediseaseincludetheadministrationofradiationorplatinumtherapyatlowerdoseswithorwithouttaxanesonaweeklybasis,orthereplacementofplatinumtherapywithcetuximab.Forinstance,ECOGiscompletingaccrualtoaphase2studyofinductiontherapyfollowedbyconcurrentcetuximabandeitherlow-doseradiationorstandardradiation(intensity-modulatedradiotherapy[IMRT])inHPV-positive,resectableoropharynxcancer(seewww.clinicaltrials.govNCT01084083).AsofSep-tember2011,79of86patientstargetedforaccrualhavebeenenrolled.ThephaseIIIintergrouptrialRTOG1016(seewww.clinicaltrials.govNCT01302834),dedicatedtopatientswithHPV-positiveoropharyngealtumors,willcomparecombinationradiationandcetuximabtocombi-nationradiationandcisplatin.PatientsareeligibleiftheirtumorsareHPVpositiveasdeterminedbyp16status,andtheradiationmodalityusedisIMRT.Thetrialincludesanextensivepanelofqualityoflife-relatedmea-suresassecondaryendpointstoassessanypotentialdiffer-enceoradvantageinfunctionaloutcomeswithcetuximabversuscisplatin,becausethegoalofthisstudyistopursuebetterfunctionaloutcomeswithoutjeopardizingefficacy.

AstheprevalenceofHPVinfectionandHPV-relatedSCCHNincreases,40,44,55questionsregardingscreeningandpreventionareraised.Screeningserumsam-plesforHPV-16isreliableandrequiresonlybasiclabora-torytechniques.56Butthevalueofscreeninghigh-riskpatientsforHPVinfectionhasnotbeendemonstrated,

3887

ReviewArticle

andthebestapproachtotreatingHPV-relatedpremalig-nantlesionsdetectedbyscreeningisunknown.Moreover,whetheridentifyingpatientsathighriskofcontractingHPVmayimproveearlydetectionofSCCHNbyraisingpatientandphysicianawarenessandincreasingmonitor-ingisunclear;therisingincidenceofHPVinfectionraisesotherissues,includingthepotentialblanketinstitutionofvaccinationstrategies.Toaddresstheseissues,anewmul-tiprongedapproachtoresearchislikelyneeded,involvingcontinuedretrospectiveanalysesandepidemiologicalstudies,prospectivedatacollectionandvalidation,andtheinitiationofthefirsttrialstostratifySCCHNpatientsaccordingtoHPVstatus,asdescribedabove.CooperativetrialsthatarecollectingsamplesforHPVassessment,suchastheRTOG0522trialthatcomparedchemoradiationwithorwithoutcetuximab,showingnodifferencesoverallwiththeadditionofcetuximab,57willhelptodeterminetheimpactofHPVstatusonoutcomesfollowingcurrentstandardsofcare(seewww.clinicaltrials.govNCT00265941).

K-RAS

Therasfamilyofgenes,includingH-RAS,K-RAS,andN-RAS,encodeaproteinthatresidesonthecytoplasmicsideoftheplasmamembrane.Theproteintransmitsmi-togenicsignalsinresponsetoavarietyofphysiologicalstimuli.58Currently,thereisconsiderableinterestintheroleofactivatedmutantformsofK-RASincarcinogenesis.Mostoftheattentionhasfocusedoncolorectalcancer,wheretheprevalenceofK-RASmutationsishigh(about40%).59-61Incolorectalcancer,certainK-RASmutationsarepredictiveoflackofbenefitfromepidermalgrowthfactorreceptor(EGFR)-targetingagents,andtestingforK-RASmutationstatusisnowrecommendedbeforetreat-mentwithEGFRtherapyinthissetting.62,63AssaysfordetectingK-RASmutationsincludereal-timePCRanddirectsequencinganalysis62;commercialkitsareavailableforassayingK-RASmutationstatusincolorectalcancer.

WhetherK-RASmutationstatuspredictsresponsetoEGFR-targetedtherapyinothertypesofcancerisunknown.InSCCHN,theprevalenceofK-RASmuta-tionsislow.Onestudyreportedaprevalenceof8%,butmostreportsaregenerally<5%,oraslowas2%.65,66AlthoughsomeevidencesuggeststhatK-RASmutationsmayindicateenhancedproliferationandanaggressivedis-easecourseinSCCHN,65,66theirextremelylowpreva-lencewilllikelyprecludeamajorroleinhelpingtodefinetreatment.Atthistime,thereisnoclinicalbenefittotest-ingforK-RASmutationstatusbeforeadministering

3888EGFR-targetedtherapytopatientswithSCCHN.Fur-therevaluationofK-RAS,includingtheimplicationsofoverexpressionofwild-typeK-RAS,67isneededtobetterdefinethepotentialprognosticandpredictiverole,ifany,ofthisfactorinSCCHN.

EGFRandEGFRvIII

TheEGFRisatransmembraneproteinthatgeneratesanintracellularsignalingcascadethatresultsincellprolifera-tioninresponsetoextracellularbindingwithitsnaturalligands,suchasepidermalgrowthfactorandtransforminggrowthfactor(TGF)-a.CellsthatacquiretheabilitytooverproducetheseligandsorincreasethenumberofEGFRsontheirsurfacecancreateanautocrinegrowthpathway,resultinginuncontrolledgrowth.68,69VirtuallyallSCCHNcasesexpressEGFRaccordingtocurrentlyacceptedassays,69,70usuallyathighlevels(2þ,3þ).Indeed,mRNAexpressionofEGFRanditsligandTGF-awasincreasedin24tumorsamplescom-paredwithnormalmucosasamplestakenfromnonaf-fectedindividuals.69TherewasnoincreaseinEGFRgenecopynumber.Notably,levelsofEGFRexpressionwerealsoupregulatedinnormalmucosafromSCCHNpatients,suggestingthatincreasedEGFRexpressionisanearlyeventinthedevelopmentofSCCHN,andmayhelptoexplainthehighincidenceofsynchronousandmeta-chronousdiseaseobservedinthistypeofcancer.

OverexpressionofEGFRisanegativeprognosticfactorassociatedwithpoorlocalcontrolandsurvivalinpatientswithSCCHN.68,70,71Inarandomizedtrialeval-uatingcisplatinwithorwithoutcetuximabinpatientswithadvancedSCCHN,thehighestlevelsofEGFRexpressionbyIHCwereassociatedwithdiminishedresponsivenesstocetuximab.72Itwasalsonotedthatcetuximab-relatedskinrashtendedtobelessfrequentinthosewiththehighestlevelsofEGFRexpression.Thesefindingssuggestthat,inpatientswithveryhighlevelsofEGFR,standarddosesofcetuximabmaybeinsufficienttosaturatethetargetreceptor,resultinginlessefficacyandlesstoxicity.Thismaysetthestagefordose-escalationstudiesofcetuximabusingeitherdirectmeasurementofEGFRexpressionortheclinicalsurrogatemarkerofskinrashtogaugebiologicaltargeteffects.72TheveryhighfrequencyofEGFRexpressioninSCCHNgreatlylimitstheusefulnessofclassifyingtumorsassimplypositiveornegativeasamarkerforselectingtreatment.TheabilitytoquantifytheEGFRexpressionlevelsmoreprecisely,however,mayenhancetheuseful-nessofthisbiomarker.Themostcommonlyusedmethod

Cancer

August15,2012

BiomarkersinSCCHN/Langer

Table3.SummaryofKeyImplicationsAboutCandidateBiomarkersinSCCHN

Marker

ERCC1expressionERCC1

polymorphismsRRM1b-TubulinHPV

K-RASmutationsEGFR

EGFRvIIIEGFRkinase

domainmutations

Implications

Expressionmayberelevantforresponsetoplatinumtherapy,needsfurthervalidation.Mayberelevanttooutcomesafterradiotherapytreatment,needsfurthervalidation.

Expressionmayberelevanttoresponsetogemcitabine.ItmayalsocorrelatewithERCC1expression,andotherDNArepair-relatedmarkers,unclearrelevance.

Expressionofcertainisotypesmayinfluenceresponsetotaxanes,needsfurthervalidation.Strongprognosticfactor,warrantsdedicatedtrialdesigns.NospecifictreatmentsfortheHPV-positivepopulationyet.

Lowprevalence,nopredictivevaluedocumented.

ExpressionisuniversalinSCCHN;overexpressionisanegativeprognosticfactorafterRT.Genecopyamplification(byFISH)isnotpredictiveofoutcomesaftercetuximabtreatment.Mayaffectsensitivitytocetuximab,notyetvalidatedintheclinic.Lowprevalence,unclearrelevance.

Abbreviations:FISH,fluorescentinsituhybridization;HPV,humanpapillomavirus;RT,radiotherapy;SCCHN,squamouscellcarcinomaoftheheadandneck.

forassessingEGFRexpressionlevelsisIHCofparaffin-embeddedtumorsamples,althoughresultscanvarydependingondifferencesintechnique,thetypeofanti-bodyorfixativeused,andthestoragetimeofthesam-ple.73FluorescentinsituhybridizationcanbeusedtodeterminetheEGFRgenecopynumber,butincreasedEGFRgenecopynumberappearstobeneitherthemaincauseofincreasedEGFRexpressionnorpredictiveofout-comesinpatientswithSCCHN.69,74AnewtechniqueforassessingproteinexpressionknownastheautomatedquantitativeanalysisofproteinexpressionmayprovideamoreaccuratemeasurementoftotalandphosphorylatedEGFRexpressionintumorsamples.75Useofthisassay,however,isinitsfledglingstagesandhasyettobethor-oughlyevaluatedforitspredictiveability.

EGFRvariants

EGFR.77,78However,thesemutationsarepresentinjust7.3%ofAsianpatientswithSCCHNwhohavebeenevaluatedforthemutationandinonly1%ofevaluatedCaucasianpatientswithSCCHN,indicatingthatitisarelativelyraremutationforthistumortype.77,79AlthoughthepracticalapplicationsofEGFRmutationstatusarecurrentlylimited,furtherassessmentofEGFRvIIIandothermutationsinSCCHNiswarranted.

SerumProfiling

Serum-baseddiagnosticsofferadifferenttypeofapproach,lessinvasiveandnotdependentonthecollec-tionoftumorspecimens.Markersinserumprovideinfor-mationonhostfactorsthatmaydetermineresponsetotherapy.

Serumsamplescanbeanalyzedbymatrix-assistedlaserdesorption/ionizationtime-of-flightmassspectrom-etry.Thisanalysisgeneratesaproteomicpeakprofile,becausemostcommonlythespeciesdetectedinthesam-pleareproteinsorproteinfragments.InNSCLC,aprofileofthiskindhasbeencharacterizedthatmaybeprognosticand/orpredictiveofbenefitwithEGFRtyrosinekinaseinhibitor(TKI)therapy.80Aretrospectivestudyin108patientswithrecurrent/metastaticSCCHNtreatedwithcetuximaborEGFRTKIsrevealedthatpatientswiththesameproteomicprofile(66%ofthesample)derivedgreaterbenefitfromanti-EGFRtherapythanthosenega-tivefortheprofile(HRforOS,0.20to0.41,dependingontheagent).Therewasalsoasmallnumericdifferencebetweenpositiveandnegativepatientstreatedwithdoce-taxelchemotherapy(n¼34),whichcomplicatestheinterpretationofthesedata,suggestingthatthismarker

AmutantformofEGFRknownasEGFRvIIIhasbeendetectedinupto40%ofSCCHNcases.76Thistruncated,constitutivelyactivereceptorisligandinde-pendentanddoesnotbindwithantibodiesthattargettheextracellulardomainofwild-typeEGFR.Invitro,cellsthatexpressEGFRvIIIhavebeenshowntobelesssensi-tivetothegrowth-inhibitingeffectsofcetuximab.Nota-bly,EGFRvIIIistypicallyonlyseenincellsthatoverexpresswild-typeEGFR,whichhasledtothehypoth-esisthatEGFRvIIImutationsarealatestageeventcausedbytherapidproliferationinducedbywild-typeEGFRoverexpression.76OtheralterationsinEGFRincludeactivatingmuta-tions,whichhavebeendetectedin10%to30%ofpatientswithNSCLCandpredictincreasedsensitivitytosmallmoleculesthattargettheintracellularportionof

Cancer

August15,2012

38

ReviewArticle

possiblycouldhaveaprognosticcomponent,inadditiontoapredictiveeffect.81Adifferentstrategyistheanalysisofadefinedpanelofbiomarkersinserum,ratherthanaproteomicapproach.Byersetal82andFerrisetal83haveanalyzeddefinedpanelsofcytokinesandangiogenicfactorsbyenzyme-linkedimmunosorbentassayonserumfrompatientswithlocallyadvancedSCCHNtreatedwithinduction5-fluorouracilordocetaxel/cisplatinpluscetux-imab.Inbothstudies,markersignaturesincludingvascu-larendothelialgrowthfactorandinterleukin6wereidentifiedthatappearedtocorrelatewithprogressivedis-easeorrecurrence.Follow-upstudiescouldconfirmifthiseffectallowsforprognosticstratification,orispredictiveofspecifictreatmentbenefits.

Conclusions

ThereareagrowingnumberofmolecularmarkersthatmaypotentiallybeusedaseitherprognosticorpredictivetoolsinthetreatmentofSCCHN(Table3).Inparticular,2factors—ERCC1andHPV—havebeenstudiedexten-sivelyandhaveproducedconsistentresultsinbothSCCHNandothertypesofcancer.Pendingfurtherclini-calstudy,thesefactorsarelikelytosoonalterhowweselecttreatmentforpatientswithSCCHN.

Withregardtootherpotentialmarkers,however,itisbecomingincreasinglyapparentthatwhatisrelevantin1typeofcancermaynotnecessarilyapplytoothers.De-spitetheirimportanceincolorectalcancer,K-RASmuta-tions,forexample,donotappeartoplayanyroleinSCCHN,androutinetestingforK-RASmutationsinSCCHNpatientsisnotwarrantedatthistime.EGFRunquestionablyplaysanimportantroleinthedevelop-mentofSCCHN,butitsusefulnessasabiomarkerinthissettingmaydependonfurtherrefinementoftestingmeth-odsandbetterdefinitionofcandidateabnormalities.ThecooperativegroupswillcontinuetostudytheseandothermarkerstobetterselectthemostappropriatetreatmentforpatientswithSCCHN.

FUNDINGSOURCES

C.J.L.isfullyresponsibleforthecontentofthisarticle,butwishesInc.toacknowledgeresearchingeditorialreferences,team,withthepreparingfundingassistancefromoftheClinicalInsights,tables,Bristol-MyerseditingtheSquibb,draft,andinformattingitforsubmission.

CONFLICTOFINTERESTDISCLOSURES

Speaker’sbureaus:BMS,ImClone,Lilly,Genentech/OSI(allcurtailedasofDecember2010);advisoryboards:BMS,

30ImClone,Lilly,Genentech/OSI,Sanofi-Aventis,Novartis,Biode-six,ringer-Ingelheim,Clarient,Astra-Zeneca,Lilly,Genentech/OSI,Bayer-Onyx;Celgene,Abraxis,Allos,Pfizer,Boeh-Celgene,researchAbraxis,funding:Pfizer,BMS,Boehringer-ImClone,Ingelheim,GSK.

REFERENCES

1.JemalA,SiegelR,XuJ,etal.Cancerstatistics,2010.CACancerJClin.2010;60:277-300.

2.TheCatalogueofSomaticMutationsinCancer(COSMIC)data-base.Availableat:http://www.sanger.ac.uk/cosmicAccessedAugust18,2009.

3.KwonH-C,RohMS,OhSY,etal.PrognosticvalueofexpressionofERCC1,thymidylatesynthase,andglutathioneS-transferaseP1for5-fluoruracil/oxaliplatinchemotherapyinadvancedgastriccan-cer.AnnOncol.2007;18:504-509.

4.JohnsonSW,PerezRP,GodwinAK,etal.Roleofplatinum-DNAadductformationandremovalincisplatinresistanceinhumanovariancancercelllines.BiochemPharmacol.1994;47:6-697.

5.JohnsonSW,SwiggardPA,HandelLM,etal.Relationshipbetweenplatinum-DNAadductformationandremovalandcisplatincyto-toxicityincisplatin-sensitiveand-resistanthumanovariancancercells.CancerRes.1994;:5911-5916.

6.ReedE.Platinum-DNAadduct,nucleotideexcisionrepairandplat-inumbasedanti-cancerchemotherapy.CancerTreatRev.1998;24:331-344.

7.ReardonJT,VaismanA,ChaneySG,etal.Efficientnucleotideexci-sionrepairofcisplatin,oxaliplatin,andbis-aceto-ammine-dichloro-cy-clohexylamine-platinum(IV)(JM216)platinumintrastrandDNAdiadducts.CancerRes.1999;59:3968-3971.

8.ZhenW,LinkCJJr,O’ConnorPM,etal.Increasedgene-specificrepairofcisplatininterstrandcross-linksincisplatin-resistanthumanovariancancercelllines.MolCellBiol.1992;12:36-3698.

9.BramsonJ,PanasciL.EffectofERCC-1overexpressiononsensitiv-ityofChinesehamsterovarycellstoDNAdamagingagents.CancerRes.1993;53:3227-3240.

10.JiangH,YangL-Y.CellcyclecheckpointabrogatorUCN-01inhib-itsDNArepair:associationwithattenuationoftheinteractionofXPAandERCC1nucleotideexcisionrepairprotein.CancerRes.1999;59:4529-4534.

11.SelvakumaranM,PisarcikDA,BaoR,etal.Enhancedcisplatincy-totoxicitybydisturbingthenucleotideexcisionrepairpathwayinovariancancercelllines.CancerRes.2003;63:1311-1316.

12.ChangI-Y,KimM-H,KimHB,etal.SmallinterferingRNA-inducedsuppressionofERCC1enhancessensitivityofhumancancercellstocisplatin.BiochemBiophysResComm.2005;327:225-233.13.DabholkarM,VionnetJ,Bostick-BrutonF,etal.MessengerRNA

levelsofXPACandERCC1inovariancancertissuecorrelatewithresponsetoplatinum-basedchemotherapy.JClinInvest.1994;94:703-708.

14.MetzgerR,LeichmanCG,DanenbergKD,etal.ERCC1mRNA

levelscomplementthymidylatesynthasemRNAlevelsinpredictingresponseandsurvivalforgastriccancerpatientsreceivingcombina-tioncisplatinandfluorouracilchemotherapy.JClinOncol.1998;16:309-316.

15.ShirotaY,StoehlmacherJ,BrabenderJ,etal.ERCC1andthymidy-latesynthasemRNAlevelspredictsurvivalforcolorectalcancerpatientsreceivingcombinationoxaliplatinandfluorouracilchemo-therapy.JClinOncol.2001;19:4298-4304.

16.Moore-JoshiM-B,ShirotaY,DanenbergKD,etal.Highgene

expressionofTS1,GSTP1,andERCC1areriskfactorsforsurvivalinpatientstreatedwithtrimodalitytherapyforesophagealcancer.ClinCancerRes.2005;11:2215-2221.

17.LordRV,BrabenderJ,GandaraD,etal.LowERCC1expression

correlateswithprolongedsurvivalaftercisplatinplusgemcitabinechemotherapyinnon-smallcelllungcancer.ClinCancerRes.2002;8:2286-2291.

Cancer

August15,2012

18.CeppiP,VolanteM,NovelloS,etal.ERCC1andRRM1gene

expressionbutnotEGFRarepredictiveofshortersurvivalinadvancednon-small-celllungcancertreatedwithcisplatinandgem-citabine.AnnOncol.2006;17:1818-1825.

19.OlaussenKA,DunantA,FouretP,etal.DNArepairbyERCC1

innon-small-celllungcancerandcisplatin-basedadjuvantchemo-therapy.NEnglJMed.2006;355:3-1.

20.BhagwatNR,RoginskayaVY,AcquafondataMB,etal.Immunode-tectionofDNArepairendonucleaseERCC1-XPFinhumantissue.CancerRes.2009;69:6831-6838.

21.CoboM,IslaD,MassutiB,etal.Customizingcisplatinbasedon

quantitativeexcisionrepaircross-complementing1mRNAexpres-sion:aphaseIIItrialinnon-small-celllungcancer.JClinOncol.2007;25:2747-27.

22.Handra-LucaA,HernandezJ,MountziosG,etal.Excisionrepair

crosscomplementationgroup1immunohistochemicalexpressionpredictsobjectiveresponseandcancer-specificsurvivalinpatientstreatedwithcisplatin-basedinductionchemotherapyforlocallyadvancedheadandnecksquamouscellcarcinoma.ClinCancerRes.2007;13:3855-3859.

23.JunHJ,AhnMJ,KimHS,etal.ERCC1expressionasapredictive

markerofsquamouscellcarcinomaoftheheadandnecktreatedwithcisplatin-basedconcurrentchemoradiation.BrJCancer.2008;99:167-172.

24.CarlesJ,MonzoM,AmatM,etal.Single-nucleotidepolymor-phismsinbaseexcisionrepair,nucleotideexcisionrepair,anddou-blestrandbreakgenesasmarkersforresponsetoradiotherapyinpatientswithstageItoIIhead-and-neckcancer.IntJRadiatOncolBiolPhys.2006;66:1022-1030.

25.KwonWS,RhaSY,ChoiYH,etal.RibonucleotidereductaseM1

(RRM1)24G>Apolymorphismshowsanassociationwithgem-citabinechemosensitivityincancercelllines.PharmacogenetGenomics.2006;16:429-438.

26.DavidsonJD,MaL,FlagellaM,etal.Anincreaseintheexpression

ofribonucleotidereductaselargesubunit1isassociatedwithgemci-tabineresistanceinnon-smallcelllungcancercelllines.CancerRes.2004;:3761-3766.

27.GoanY-G,ZhouB,HuE,etal.Overexpressionofribonucleotide

reductaseasamechanismofresistanceto2,2-difluorodeoxycyti-dineinthehumanKBcancercellline.CancerRes.1999;59:4204-4207.

28.DumontetC,Fabianowska-MajewskaK,MantincicD,etal.Com-monresistancemechanismstodeoxynucleosideanaloguesinvariantsofthehumanerythroleukaemiclineK562.BrJHaematol.1999;106:78-85.

29.BeplerG,KusmartsevaI,SharmaS,etal.RRM1modulatedin

vitroandinvivoefficacyofgemcitabineandplatinuminnon-small-celllungcancer.JClinOncol.2006;24:4731-4737.

30.YeoW,SoongRC,ChuahBY,etal.CorrelationofRRM1pro-moterregionssinglenucleotidepolymorphisms(SNPs)withresponseandoutcomeinbreastcancerpatientstreatedwithgemci-tabine-basedchemotherapy[abstract].JClinOncol.2008;26(suppl).Abstract14513.

31.RosellR,DanenbergKD,AlberolaV,etal.Ribonucleotidereduc-tasemessengerRNAexpressionandsurvivalingemcitabine/cispla-tin-treatedadvancednon-smallcelllungcancerpatients.ClinCancerRes.2004;10:1318-1325.

32.PapadakiC,TrypakiM,KoutsopoulosA,etal.Associationof

tumoralBRCA1,RRM1,andRRM2mRNAexpressionlevelswithclinicalresponsetofront-linedocetaxel/gemcitabine(DG)chemo-therapyinpatientswithnon-smallcelllungcancer(NSCLC)[abstract].JClinOncol.2008;26(suppl).Abstract8112.

33.KavallarisM,KuoDY,BurkhartCA,etal.Taxol-resistantepithelial

ovariantumorsareassociatedwithalteredexpressionofspecificbeta-tubulinisotypes.JClinInvest.1997;100:1282-1293.

34.RanganathanS,DexterDW,BenetatosCA,HudesGR.Cloning

andsequencingofhumanbetaIII-tubulincDNA:inductionofbetaIIIisotypeinhumanprostatecarcinomacellsbyacuteexposuretoantimicrotubuleagents.BiochimBiophysActa.1998;1395:237-245.Cancer

August15,2012

BiomarkersinSCCHN/Langer

35.MonzoM,RosellR,SanchezJJ,etal.Paclitaxelresistanceinnon-small-celllungcancerassociatedwithbeta-tubulingenemutations.JClinOncol.1999;17:1786-1793.

36.SeveP,MackeyJ,IsaacS,etal.ClassIIIbeta-tubulinexpressionin

tumorcellspredictsresponseandoutcomeinpatientswithnon-smallcelllungcancerreceivingpaclitaxel.MolCancerTher.2005;4:2001-2007.

37.DumontetC,IsaacS,SouquetPJ,etal.ExpressionofclassIIIbeta

tubulininnon-smallcelllungcanceriscorrelatedwithresistancetotaxanechemotherapy.BullCancer.2005;92:E25-E30.

38.CullenKJ,SchumakerL,NikitakisN,etal.Beta-tubulin-II

expressionstronglypredictsoutcomeinpatientsreceivinginduc-tionchemotherapyforlocallyadvancedsquamouscarcinomaoftheheadandneck:acompanionanalysisoftheTAX324trial.JClinOncol.2009;27:6222-6228.

39.WalboomersJM,JabobsMV,ManosMM,etal.Humanpapillo-mavirusisanecessarycauseofinvasivecervicalcancerworldwide.JPathol.1999;1:12-19.

40.ChaturvediA,EngelsE,AndersonW,etal.Incidencetrendsfor

humanpapillomavirus-related(HPV-R)andunrelated(HPV-U)headandnecksquamouscellcarcinomas(HNSCC)intheUnitedStates(US)[abstract].JClinOncol.2007;25(suppl).Abstract6001.41.RyersonAB,PetersES,CoughlinSS,etal.Burdenofpotentially

humanpapillomavirus-associatedcancersoftheoropharynxandoralcavityintheUS,1998-2003.Cancer.2008;113(10suppl):2901-2909.

42.AngKK,HarrisJ,WheelerR,etal.Humanpapillomavirusand

survivalofpatientswithoropharyngealcancer.NEnglJMed.2010;363:24-35.

43.HerreroR,CastellsagueX,PawlitaM,etal.Humanpapillomavirus

andoralcancer:theInternationalAgencyforResearchonCancerMulticenterStudy.JNatlCancerInst.2003;95:1772-1783.

44.D’SouzaG,KreimerAR,ViscidiR,etal.Case-controlstudyof

humanpapillomavirusandoropharyngealcancer.NEnglJMed.2007;356:1944-1956.

45.FakhryC,WestraWH,LiS,etal.Improvedsurvivalofpatient

withhumanpapillomavirus-positiveheadandnecksquamouscellcarcinomainaprospectiveclinicaltrial.JNatlCancerInst.2008;1000:261-269.

46.SyrjanenS.Humanpapillomavirus(HPV)inheadandneckcancer.

JClinVirol.2005;32S:S59-S66.

47.PecoraroG,MorganD,DefendiV.Differentialeffectsofhuman

papillomavirustype6,16,and18DNAsonimmortalizationandtransformationofhumancervicalepithelialcells.ProcNatlAcadSciUSA.19;86:563-567.

48.ParkNH,MinBM,LiSL,etal.Immortalizationofnormalhuman

oralkeratinocyteswithtype16humanpapillomavirus.Carcinogene-sis.1991;12:1627-1631.

49.AndlT,KahnT,PfuhlA,etal.Etiologicalinvolvementofonco-genichumanpapillomavirusintonsillarsquamouscellcarcinomaslackingretinoblastomacellcyclecontrol.CancerRes.1998;58:5-13.

50.MorkJ,LieAK,GlattreE,etal.Humanpapillomavirusinfection

asariskfactorforsquamous-cellcarcinomaoftheheadandneck.NEnglJMed.2001;344:1125-1131.

51.GillisonML,KochWM,CaponeRB,etal.Evidenceforacausal

associationbetweenhumanpapillomavirusandasubsetofheadandneckcancers.JNatlCancerInst.2000;92:709-720.

52.RischinD,YoungR,FisherR,etal.Prognosticsignificanceof

p16INK4Aandhumanpapillomavirusinpatientswithoropharyn-gealcancertreatedonTROG02.02phaseIIItrial.JClinOncol.2010;28:4142-4148.

53.GillisonML,ZhangQ,AngK,etal.Analysisoftheeffectofp16

andtobaccopack-years(p-y)onoverall(OS)andprogression-freesurvival(PFS)forpatientswithoropharynxcancer(OPC)inRadia-tionTherapyOncologyGroup(RTOG)protocol9003[abstract].JClinOncol.2010;28:(15suppl).Abstract5510.

.LorchJ,PosnerM,GoloubevaO,etal.Five-yearsurvival(OS)and

patternsoffailureforhumanpapillomavirus(HPV)positiveandnegativeoropharynxcancer(OPC)intheTAX324clinicaltrial:

31

ReviewArticle

resultsofsequentialtherapy[abstract].AnnOncol.2010;21(suppl8).Abstract1004O.

55.

CenterforDiseaseControlandPrevention.Humanpapillomavirus:HPVinformationforclinicians.April2007.Availableat:http://www.cdc.gov/std/hpv/common-clinicians/ClinicianBro-fp.pdfAccessedAugust18,2009.

56.

KirnbauerR,HubbertNL,WheelerCM,etal.Avirus-likeparticleenzyme-linkedimmunosorbentassaydetectsserumantibodiesinamajorityofwomeninfectedwithhumanpapillomavirustype16.JNatlCancerInst.1994;86:494-499.

57.

AngKK,ZhangQE,RosenthalDI,etal.PhaseIIItrial(RTOG0522)ofconcurrentacceleratedradiationpluscisplatinwithorwithoutcetuximab(Cetz)forlocally-advancedHNSCC[abstract].JClinOncol.2011;29(suppl).Abstract5500.

58.WilliamsHK.Molecularpathogenesisoforalsquamouscarcinoma.MolPathol.2000;53:165-172.

59.

Khambata-FordS,GarrettCR,MeropolNJ,etal.ExpressionofepiregulinandamphiregulinandK-RASmutationstatuspredictdiseasecontrolinmetastaticcolorectalcancerpatientstreatedwithcetuximab.JClinOncol.2007;25:3230-3237.

60.KarapetisCS,Khambata-FordS,JonkerDJ,etal.K-RASmutationsandbenefitfromcetuximabinadvancedcolorectalcancer.NEnglJMed.2008;359:1757-1765.

61.

BokemeyerC,BondarenkoI,HartmannJT,etal.KRASstatusandefficacyoffirst-linetreatmentofpatientswithmetastaticcolorectalcancer(mCRC)withFOLFOXwithorwithoutcetuximab:theOPUSexperience[abstract].JClinOncol.2008;26(suppl).Abstract4000.

62.

AllegraCJ,JessupJM,SomerfieldMR,etal.AmericanSocietyofClinicalOncologyprovisionalclinicalopinion:testingforKRASgenemutationsinpatientwithmetastaticcolorectalcarcinomatopredictresponsetoanti-epidermalgrowthfactorreceptormonoclo-nalantibodytherapy.JClinOncol.2009;27:2091-2096.

63.EuropeanMedicinesAgency.Summaryofproductcharacteristics.Availableat:http://www.emea.europa.eu/humandocs/PDFs/EPAR/erbitux/H-558-PI-en.pdf.AccessedonOctober1,2010.

.

RathckeIO,GottschlichS,GoroghT,etal.IncidenceofpointmutationsinKi-rascodon12and13insquamousepithelialcarci-nomasofthehead-neckregion.Laryngorhinootologie.1996;75:465-470.

65.YarbroughWG,ShoresC,WitsellDL,etal.Rasmutationsandexpressioninheadandnecksquamouscellcarcinomas.Laryngo-scope.1994;104:1337-1347.

66.

BissadaE,Abou-ChacraZ,WengX,etal.PrevalenceofK-RAScodon12mutationsinlocallyadvancedheadandnecksquamouscellcarcinomaandinfluencewithregardstoresponsetochemora-diationtherapy[abstract].JClinOncol.2008;26(suppl).Abstract17005.

67.HoaM,DavisSL,AmesSJ,etal.Amplificationofwild-typeK-RASpromotesgrowthofheadandnecksquamouscellcarci-noma.CancerRes.2002;62:71-7156.

68.MauriziM,AlmadoriG,FerradinaG,etal.Prognosticsignificanceofepidermalgrowthfactorreceptorinlaryngealsquamouscellcar-cinoma.BrJCancer.1996;74:1253-1257.

69.

RubinGrandisJ,TweardyDJ.ElevatedlevelsoftransforminggrowthfactoraandepidermalgrowthfactorreceptormessengerRNAareearlymarkersofcarcinogenesisinheadandneckcancer.CancerRes.1993;53:3579-3584.

3270.DassonvilleO,FormentoJL,FrancoualM,etal.Expressionofepidermal

growthfactorreceptorandsurvivalinupperaerodigestivetractcancer.JClinOncol.1993;11:1873-1878.

71.AngKK,BerkeyBA,TuX,etal.Impactofepidermalgrowthfac-torreceptorexpressiononsurvivalandpatternofrelapseinpatientswithadvancedheadandneckcarcinoma.CancerRes.2002;62:7350-7356.

72.BurtnessB,GoldwasserMA,FloodW,etal.PhaseIIIrandomized

trialofcisplatinplusplacebocomparedwithcisplatinpluscetuxi-mabinmetastatic/recurrentheadandneckcancer:anEasternCo-operativeOncologyGroupstudy.JClinOncol.2005;24:86-86.

73.AtkinsD,ReiffenK-A,TegtmeierCL,etal.Immunohistochemical

detectionofEGFRinparaffin-embeddedtumortissues:variationinstainingintensityduetochoiceoffixativeandstoragetimeoftissuesections.JHistochemCytochem.2004;52:3-901.

74.VermorkenJB,MesiaR,RemenarE,etal.Predictorsofefficacyin

theEXTREMEstudy:cetuximabplusplatinum-basedtherapyfirst-lineinpatientswithrecurrentand/ormetastatic(R/M)squamouscellcarcinomaoftheheadandneck(SCCHN)[abstract].AnnOncol.2008;19(suppl).Abstract6870.

75.PsyrriA,YuZ,WeinbergerPM,etal.Quantitativedetermination

ofnuclearandcytoplasmicepidermalgrowthfactorreceptorexpres-sioninoropharyngealsquamouscellcancerbyusingautomatedquantitativeanalysis.ClinCancerRes.2005;11:5856-5862.

76.SokJC,CoppelliFM,ThomasSM,etal.Mutantepidermalgrowth

factorreceptor(EGFRvIII)contributestoheadandneckcancergrowthandresistancetoEGFRtargeting.ClinCancerRes.2006;12:50-5073.

77.Loeffler-RaggJ,Witsch-BaumgartnerM,TzankovA,etal.Low

incidenceofmutationsinEGFRkinasedomaininCaucasianpatientswithheadandnecksquamouscellcarcinoma.EurJCancer.2006;42:109-111.

78.EberhardDA,JohnsonBE,AmlerLC,etal.Mutationsintheepi-dermalgrowthfactorreceptorandinKRASarepredictiveandprognosticindicatorsinpatientswithnon-small-celllungcancertreatedwithchemotherapyaloneandincombinationwitherlotinib.JClinOncol.2005;23:5900-5909.

79.LeeJW,SoungYH,KimSY,etal.SomaticmutationsofEGFR

geneinsquamouscellcarcinomaoftheheadandneck.ClinCancerRes.2005;11:2879-2882.

80.CarboneDP,SeymourL,DingK,etal.Serumproteomicpredictionof

outcomesinadvancedNSCLCpatientstreatedwitherlotinib/placebointheNCICClinicalTrialsGroupBR.21trial[abstract].JThoracOncol.2010;5(suppl1).Abstract203O.

81.ChungCH,SeeleyEH,RoderH,etal.Detectionoftumorepider-malgrowthfactorreceptorpathwaydependencebyserummassspectrometryincancerpatients.CancerEpidemiolBiomarkersPrev.2010;19:358-365.

82.ByersLA,HolsingerFC,KiesMS,etal.Serumsignatureofhy-poxia-regulatedfactorsisassociatedwithprogressionafterinductiontherapyinheadandnecksquamouscellcancer.MolCancerTher.2010;9:1755-1763.

83.FerrisRL,LeeSC,FeinsteinC,etal.Serumbiomarkersaspredic-torsofclinicaloutcomeaftercetuximab-basedtherapyinpatientswithlocally-advancedsquamous-cellcarcinomaoftheheadandneck[abstract].JClinOncol.2009;27(15suppl).Abstract6035.

Cancer

August15,2012