CancerVaccineCenterDanaFarberCancerInstituteBostonMAUSABroadInstituteCambridgeMAUSADepartmentofHematologyYaleCancerCenterNewHavenCTParisDiderotParisFranceMassachusettsGeneralHospitalC ID: 483049
Download Pdf The PPT/PDF document "Clonalevolutioninhematologicalmalignanci..." is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
ClonalevolutioninhematologicalmalignanciesandtherapeuticDALandau1,2,3,4,SLCarter,GGetzandCJWuTheabilityofcancertoevolveandadaptisaprincipalchallengetotherapyingeneralandtotheparadigmoftargetedtherapyinparticular.Thisabilityisfueledbytheco-existenceofmultiple,geneticallyheterogeneoussubpopulationswithinthecancercell CancerVaccineCenter,Dana-FarberCancerInstitute,Boston,MA,USA;BroadInstitute,Cambridge,MA,USA;DepartmentofHematology,YaleCancerCenter,NewHaven,CT,ParisDiderot,Paris,France;MassachusettsGeneralHospitalCancerCenterandDepartmentofPathology,Boston,MA,USA;DepartmentofMedicalOncology,Dana-FarberCancerInstitute,Boston,MA,USAandDepartmentofMedicine,BrighamandWomensHospital,HarvardMedicalSchool,Boston,MA,USA.Correspondence:DrCJWu, Leukemia(2014),34 43 2014MacmillanPublishersLimitedAllrightsreserved0887-6924/14www.nature.com/leu evolutionaryoptimizationprocess,assubclonescompeteoverresourcesandadapttoexternalpressures,suchascancertherapy.Cancerprogression,therefore,isfundamentallyaprocessofmutationaldiversicationandclonalselection.Therstexperimentalevidencesupportingtheideathattumorsarecomposedofheterogeneoussubpopulationswasobtainedfrommousemodelsofsolidmalignancies.Theseexperimentsshowedthatindividualsubclonespossesseddifferentphenotypiccharacteristics,includingvaryingmetastaticpotential.Importantly,thelinkbetweenheterogeneityandresistancetotherapywasapparenteveninotherearlyexperiments.Forexample,celllinesthatexhibitedahigherdegreeofphenotypicheterogeneityalsoacquiredresistancetochemotherapy(methotrexate)atahigherratecomparedwithcelllineswithlowerphenotypicvariability.Ascancerisadiseasethatresultsfromtheaccumulationofgeneticalterations,anaturalcorollaryoftheabovestudiesisthatphenotypicevolutionmuststemfromunderlyinggenotypicevolution.Thisconcepthasbeenindeedconrmedoverthepastseveraldecadeswithincreasingtechnologicalsophistication,usingapproachesbasedoncytogeneticsandSanger(rst-generationsequencing).Mullighanetal.forexample,inanelegantsingle-nucleotidepolymorphismarrayanalysesofpediatricpre-BcellALL,demonstratedcomplexbranchedevolutionarygrowthassociatedwithdiseaserelapse.Thislandmarkstudyfurthershowedhowrelapseddiseaseisgeneticallyalteredcomparedwithdiseaseatdiagnosis.Inchroniclymphocyticleukemia(CLL),clonalevolutionwasidentiedinupto43%ofpatientsusinguorescentinsituhybridizationorcytogenetictechniques,withfrequentacquisitionofthepoorprognosticmarkersdel(11q)anddel(17p),andoccurringatahigherrateinthepoorprognosisgroupofTogether,theseexperimentalobservationshavedemon-stratedthatthegeneticmakeupofhematologicalmalignanciesisconstantlyreshapedduringdiseaseprogression.Overall,theysupporttheprescientideastheorizedbyNowell,whopostulatedthatgeneticinstabilitywouldbeexpectedtoleadtoenhancedheterogeneitywithcancerprogression,resultingindiverse,geneticallydistinct,subpopulationswithinaneoplasm.Thus,theselectionprocesswouldbeexpectedtopromotetheoutgrowthofincreasinglytsubclones,therebycontinuouslyremodelingthetnessoftheoverallIfcancerplasticityisdrivenbyclonalheterogeneity,itisimportanttoconsiderthefeaturesthatfuelthegenerationofclonalheterogeneity.Geneticinstabilityundoubtedlyhasakeyroleinthisprocess.Therateofacquisitionofnovelsomaticmutationsisprobablycloselytiedtothediversicationofthecancerpopulationandthereforeforenhancingitsevolutionarypotential,togetherwithotherfeatures,suchasthepopulation(comprehensivelyreviewedelsewhere).ApermissivegeneticcontextthateitherinhibitsDNArepair(forexample,mutations)orincreasestolerancetonovelmutationsbyremovingcriticalcheckpoints(forexample,TP53ATMmutations,enablingtolerancetowardsmassivegenomicdamageislikelytoincreasetheoveralldiversityofthetumorpopulation.Addingtothecomplexity,differentareasofthegenomemayhavedifferentratesofmutationsacquisition,29,30whichwouldneedtobetakenintoaccountwheninferringpastratesofmutationsfromgenomicinformation.Apotentiallyprovocativenotionthatarisesfromthesedataiswhethergeneticinstabilitymaybetargetedasameasuretoinhibitcancerevolution.Forexample,forhematologicalmalignancies,thedocumentedongoingmutagenicactivityofenzymesresponsibleforB-andT-cellreceptorgeneticmodicationsmaybeofparticularinterest.UNRAVELINGCLONALCOMPLEXITYWITHMASSIVELYPARALLELSEQUENCING(MPS)Althoughconrmingthebasictenetofcancerasanevolutionarydisease,theabovedescribedstudiesareinherentlylimitedintheirabilitytodecipherthetrueextentofgenomichetero-geneity,giventhelimitedamountofgeneticlesionsstudiedatanyexperimentandthelimitedsensitivityofexperimentaltechniquesthatwerethenavailabletodetectsmallersubclones.BothoftheselimitationshavebeenlargelyovercomewiththeadventofMPS.MPSoftumorshasaffordedanexponentialincreaseintheabilitytocharacterizethegeneticlandscapeofcancer.Ithasrevealedaveryhighdegreeofintertumoralheterogeneity(thatis,differentgeneticlesionaffectingdifferenttumors),withhundredsofdifferentmutationsaffectingdifferenttumorswithaprobableeffectontness.34,35Moreover,thistechnologyhasalsorevealedahighlevelofintratumoralgeneticheterogeneity(thatis,differentgeneticlesionsaffectingdifferentsubcloneswithinanindividualtumor),whichalsoaffectsputativedriverevents.Inparticular,theadventofMPShasallowedresearcherstoidentifybothsubclonalsomaticcopynumberalterations(SCNA)andsubclonalsomaticsingle-nucleotidevariations(SSNV),37,38whichcanbetrackedovertimetostudytumorevolution.39,40ThisabilitytoreconstructtheclonalstructureisderivedfromaninherentpropertyofMPS.Itinvolvesgeneratingbillionsofindependentsequencingreads,eachderivedfromasingleDNAmolecule.Thus,MPSdatarepresentaninformativerandomsampleofindividualDNAmoleculescontainedwithinatumor.AtSSNVsites,thenumberofsequencingreadssupportingthealternateandreferencebasescanbeusedtocalculateaquantitativemeasurementofthevariantallelicfraction(VAF).Insamplesderivedfromdiploidcancerssuchasacutemyeloidleukemia(AML)thatessentiallylackSCNAsandarenotcontaminatedwithnon-malignantcells,allelicfractionscanbeusedtoestimateSSNVclonalitydirectly(inwhichanyclonalSSNVshouldhaveaVAFof0.5;whileSSNVsinsubcloneswillhavelowerVAFs).However,thevastmajorityofhumancancerscontainfrequentSCNAs,withmanyofthemhavingundergonewholegenomedoublingduringtheirevolution.Inaddition,mosttumorspecimenscontainasubstantialfractionofnormalcells.Thus,inordertoaccuratelyinferthefractionoftumorcellsthatcontainanSSNVfromMPSdata,itisnecessarytoaccountforboththecopynumberattheSSNVsiteandtheoveralltumorpurityinthesample.Recently,inferencemethodshavebeendevelopedwhichattempttoaccountforthesefactorsinordertoestimatetheactualcancercellfractionharboringaspecicmutation(Figure1).Althoughmoderatesequencingdepthmayresultinconsiderableuncertaintyinthecancercellfractionestimatesofindividualmutations,thefactthatsubclonalmutationsareexpectedtoco-occurindiscretesubclonalcellpopulationshasformedthebasisofusingclusteringtechniquestobetterresolvethesubclonalstructureofbulktumorsamples.BecauseDNAfromallcellspresentinthebulksampleismixedtogetherbeforesequencing,informationregardingwhichmuta-tionsco-occurinspecicsubclonesisnotreadilyaccessiblefromanalysisofasinglesample.Ingeneral,analysisofbulkDNAfromasinglecancersamplecannotruleoutthepossibilitythatallsubclonesarenestedinsideoneanotherinalinearphylogeny.Oneexceptiontothiswasreportedinastudythatapplieddeepwholegenomesequencing(188X)ofasingleprimarybreastcancersampleandcouldinferbranchedevolutionbasedonco-occurrenceofinformativeallelesonsequencingreads.thisapproachrepresentsanelegantsolutiontothephylogeneticinferenceproblem,thedeepcoverageneededoverthewholegenomerendersitimpracticalforlargestudiesusingexistingtechnology.Wenotethatsequencingplatformscapableofreliablyproducinglongerreadswillmakethisapproachfarmorepowerful. ClonalevolutionandtherapeuticstrategiesDALandauetal 2014MacmillanPublishersLimitedLeukemia(2014)34 43 AnattractiveapproachtoinferringphylogeneticstructuresfromanalysisofbulktumorDNAistosequencemultiplespecimensfromthesameindividualscancer.Branchedevolutionaryrelation-shipscanbedetectedasclustersinwhichonesubclonemayincreaseinfrequencywhileanothersiblingsubclonemayexhibitaconcomitantdecreaseinfrequency.Thisapproachhasbeenusedtoidentifybranchedevolutioninleukemiasamplestakenbeforeandaftertreatmentaswellasinsolidtumorssampledatmultipleanatomicallocations.THESURPRISINGASPECTSOFCLONALCOMPLEXITYOFHEMATOLOGICALMALIGNANCIESOneoftherstkeylessonsgleanedfromgenome-widestudiesofhematologicalcancersisthatclonalevolutionwhichfollowsacomplexbranchedpath(wheremultiplesubpopulationco-existinthesametumorandcompeteforascendency)isatleastascommonasamorelineartrajectory(inwhichprogenyclonesreplaceparentclonesinfullselectivesweeps).Atraditionallinearmodelofsuccessiveclonalexpansionscouldhavebeenexpectedofhematologicalmalignancies,byvirtueofthemobilenatureoftheircellularnormalcounterparts(comparedwithsolidtissuesthatareoftenembeddedinxedtissuearchitecture).Intheory,thisfeaturecouldtheoreticallyhaveledtoadecreasedlevelofclonalcomplexityascancercellscanreadilymoveacrosstissuesandhenceundergomorehomogenouscellularmixing.Thisscenarioisunlikesolidtumormalignancies,inwhichthespatialcompartmentsareformed.Tothecontrary,however,wholegenome/exomeinvestigationofclonalevolutioninAML,myelodysplasticsyndrome,multiplemyelomaandCLLhaveallconsistentlydemonstratednotonlyahighdegreeofclonalheterogeneityandmarkedchangesinthegeneticmakeupofthediseaseuponrelapsebutalsobranchingratherthanlinearasthepredominantpatternofevolution(Table1).Amajorimplicationofthesendingsisthattheevolutionaryprocessisexpectedtoresultfromcomplexinteractionsamongmultiplehighlydiversepopulationsratherthanaclearsuccessionofselectivesweeps.Clonalcompetitionamongco-existingsub-populationsthatharbordriverlesionsthusshapestheeventualcompositionofthetumorsuchthatmultipleclonalvariantsarepresentatthesametime.Itisimportanttonotethatthepublishedanalysestodatehavebeenlimitedtothedetectionofmacroscopicclonalheterogeneity(clonesize1 10%oftheentirecellpopulation).Thisisbecauseonlyclonesthateitherrepresentasubstantialproportionofthecancercellmassorclonesthatbecomedominantatsomepointduringthestudiedperiodaretrackableusingcurrentmethodologies.EmergingtechnologiescapableofachievingdeepersequencingdepthofbulkDNAorsingle-cellgenomicsequencingmethodsmayenablethestudyofsmallersubpopulations.Delineatingthefullextentofcancerheterogeneitydowntothesingle-celllevelwillenableustounderstandhowtheseeminglystochasticprocessoftrialanderroratthesingle-celllevelisintegratedthroughselectiontoshapethegeneticmakeupofthetumor.Itcarriesthepotentialtorenethedichotomyofdrivervspassengermutations,byquantifyingthetnesscontributionofeachindividualmutationtoselection(manifestedinvaryingclonesizes).EPIGENETICCLONALHETEROGENEITYAlthoughgeneticalterationhasbeenthemainfocusofevolutionarydynamicsincancerthusfar,epigeneticmodicationsareprobablyresponsibleforalargepartofphenotypicthatultimatelyaffecttness.Similartogeneticalterations,epigeneticmodicationsareheritableandthereforesubjecttonaturalselection.Thecontributionofepigeneticmodicationtoselectionincancerisprobablysubstantial,asepigeneticalterationsaccumulateasthecellpopulationevolvesanddiversiesatratesestimatedtobeordersofmagnitudehighercomparedwithsomaticgeneticalterations.Indeed,alargedegreeofintratumorepigeneticheterogeneitywasrecentlydescribedinlymphomausingDNAmethylationarrays.Geneticandepigeneticchangesprobablyhavecomplexbidirectionalinteractionsandco-operatetomoldtheevolutionarylandscape.Thiscomplexandbidirectionalinterplaybetweengeneticandepigeneticfeaturesincancerhasbeenperhapsmostdeeplyexploredintheareaofcancerstemcells.Specically,earlyxenograftstudiesofALLrevealedleukemicrepopulationthatrecapitulatedthegeneticheterogeneityofthepatientsoriginalSimilarndingswerealsodemonstratedinsolidtumormalignancies.etal.concludedthatcancerstemcellsanepigeneticallyuniformpopulationaregeneticallydiverse.Ontheotherhand,evengeneticallyuniformcellsubpopulationshavebeenreportedtorevealprofoundepigeneticdifferencesleadingtodifferencesinthephenotypesofsurvivalcapacityandpluripotencypotential. Figure1.Inferringthesizeofasubpopulationaffectedbysomaticmutationsfromgenomicdata.MPSprovidesanestimateofVAF,whichiscalculatedbycountingthenumberofreadswiththevariantallelesanddividingitbythetotalnumberofreadingfromthespeciÞclocation.Thecertaintyoftheestimateisafunctionofthedepthofcoverage,usingtheBetadistribution().Subsequently,theVAFestimatesareintegratedwiththepurityandlocalcopynumberinformation()toyieldcancercellfractions().Intheexampleprovided,asomaticmutationwithaVAFof0.125,alocalcopynumberof3andapurityof67%yieldscancercellfractionestimatesof0.5. Clonalevolutionandtherapeuticstrategies DALandauetalLeukemia(2014)34 432014MacmillanPublishersLimited Together,theseobservationspromptamodelofcancerevolutioninwhichepigeneticandgeneticheterogeneityareintegrated,therebyaccountingforepigeneticheterogeneityofgeneticallyuniformpopulationsandgeneticheterogeneityofepigeneticallyuniformpopulations.Suchamodelwoulddesig-natecellswithhighself-renewalcapacity(anattributeencodedintheepigeneticstate)asthecrucialunitssubjectedtoselectionforcesingeneticevolution.Hence,theappearanceofanewsomaticmutationwithinthesespecializedcellscouldleadtotheirclonalpropagation.Atthesametime,suchamodelwouldalsoacknowledgetheexistenceofafarlesshomogenouscancerstemcellpopulationthenpreviouslyconsideredwithregardtovariousfeatures,includingdrugsensitivity.Itisimportanttoconsiderthatincancerthemovementbetweendifferentepigeneticstates(alongthespectrumofpluripotencytodifferentiation,forexample)maybealteredaswell.Inmulticellularorganisms,epigenetictransitionsaretightlycontrolledthroughnumerousregulatorymechanisms.Neoplastictransformationcanunhingethosemechanisms,revertingtoastatemorecloselyresemblingunicellularinwhichtheuidmovementacrossdiversestatescanachievehighadaptivitybytrialanderror.Epigeneticheterogeneity,thus,canbeahedgingstrategyforenhancedCancerprogression,therefore,maybeviewedasascenarioinwhichbothgeneticandepigeneticpopulationstructuresbecomeincreasinglymalleable,suchthatthelinesbetweenpopulationswithdifferentstemnesspotentialbecomemoreblurred.Withinthisframework,stemnessmayexistasafunctionalphenotype,whichcanbemanifestedbyanymemberofamalignantpopulationgiventheappropriateendogenousandexogenousfactors.Thus,ahighdegreeofinterclonalcompetitionwouldprobablyselectforcellswiththehighestself-renewingcapacityattheexpenseofmoredifferentiatedcells,ashasbeendemonstratedinCML.Therapymayalsoacceleratethisprocessbyprovidingastrongselectionforcancerstemcellsurvivalandproliferation.Acquiredgeneticalterationsprobablyhaveanimportantroleinthisscenarioaswell.Forexample,thelossof,oftenseenwithdiseaseprogression,provokesstem-cell-liketranscriptionalprograms.oncogenesmayalsoaffordleukemogenicpotentialtocommittedmyeloidprogenitors,againdemonstratingthatgeneticlesionsmayenlargetheavailablepoolofcellswithstem-likefeatures.Insummary,integratingthestemcellhierarchyandthegeneticphylogenetictreeyieldsacomplexevolutionarypicturethathasonlybeguntobeunraveled.Inconcertwithgeneticdiversicationandtnessoptimization,asimilarprocessverylikelyoccursattheepigeneticlevel.Cancerstemcellsconstituteagrowingproportionwithinthecancercellmassandhaveagreaterplasticityintermsofbidirectionalconversionfromstemcellstomoredifferentiatedcells.Together,thisleadstoenhancementofthecellularsubstrateavailableforselection,withlarge,treatment-resistantandgeneticallyheterogeneouscancerstemcellHOWDOESCLONALEVOLUTIONCONTRIBUTETORESISTANCETOTHERAPY?Relapsedmalignancyshowsanalmostuniversalphenotypicevolution,resultinginamoreaggressiveandtreatment-refractoryWeandtheothershaveshownthatfrequentgeneticevolutionunderliesthephenotypicevolution.Therefore,acentralquestionincancertreatmentiswhatistheprecisenatureoftheinteractionofclonalevolutionwithcancertherapy.Initialstudieshighlightedthepotentialroleofchemotherapytoinducenovelmutagenesisandtherebytoenhancetheprocessofgeneticdiversication(Figure2a).AlthoughstudiesofWGS(wholegemonesequencing)areinherentlylimitedbythepowertodetectminutesubcloneswithinasample,studiesinacutemyeloidmalignancieshavenonethelesssuggestedthatthenovelmutagenesismayresultfromthegenotoxiceffectsofchemotherapy,supportedalsobyachangingspectrumofsomaticsingle-nucleotidealterations.contrast,incaseofindolentbloodmalignanciessuchasCLL,evidenceforthecontributionofthechemotherapysmutagenizingeffectislimited.Previouspurineanalog-basedtherapywasnotassociatedwithanincreasedtotalnumberofmutationsinCLLandalsowasnotassociatedwithanalteredmutationalpattern.Therefore,althoughchemotherapy-inducedmutagenesishasthepotentialtocontributetofurtherclonaldiversication,othersourcesforgeneratingevolutionaryshiftsappeartobeatplay,andprobablyinvolvepre-existinggeneticvariantsorsubclones.Howthendoestherapyinduceevolutionfrompre-treatmentgeneticvariation?Twoexplanationsareconsidered,dependingonthetumorkinetics,theefcacyofcellkillwithtreatmentandotherfactorsrelatedtoboththetumortypeandthespecictreatmentstrategy.Therstisthatresistantclonesmaybeactivelyselectedbytherapy(Figure2b).Examplesforthismodelarenumerous,79 81mismatchrepairgenemutationsinrecurrentglioblastomamultiformeaftertreatmentwithtemozolomideandtheBCR-ABLT315ImutationsinCML.Indeed,thismodelofclonalevolutioninducedbytheselectivepressureoftherapymaybeparticularlyrelevantinthecontextoftargetedtherapy,as Table1.Next-generationsequencingstudiesofclonalevolutioninhematologicalmalignanciesDiseaseMethodologyNumberofcasesInsightsWGS,followedbytargeteddeepsequencing8RelapseafterchemotherapyisassociatedwithclonalevolutionandacquisitionofnewmutationsSecondaryAMLWGS,followedbytargeteddeepsequencing7SecondaryAMLclonesareoftenevolvedprogenyofMDSclonesMultiplemyelomaWES1ClonalshiftsoccuralongthehistoryofthediseaseMultiplemyelomaWES1ClonalshiftsoccuralongthehistoryofthediseaseWGS,followedbytargeteddeepsequencing3DifferentpatternsofevolutionevidentthroughcyclesoftherapyWES149(18longitudinalSubclonaldriverscananticipateclonalevolutionandimpactoutcomeEssentialSingle-cellWES1ETismonoclonalinoriginFollicularWES8EarlyandlatedriversidentiÞedAbbreviations:AML,acutemyeloidleukemia;CLL,chroniclymphocyticleukemia;ET,essentialthrombocytosis;WES,wholeexomesequencing;WGS,wgemonesequencing. ClonalevolutionandtherapeuticstrategiesDALandauetal 2014MacmillanPublishersLimitedLeukemia(2014)34 43 thetherapyisoftendirectedataparticulargeneticcontextwhichmaynotbesharedbyallsubclones.Thisrelationshipbetweentherapyandgeneticadaptationislikelytoresultinconvergentevolution,inwhichamutationthatconfersresistancewillbecomehighlyprevalentinrelapseddisease.Indeed,thisprocesshasbeenreportedinrelapsedT-cellALLaftertreatmentwithnucleoside-analogchemotherapydrugs.Analternativeprocesscontributingtotheemergenceofcontinuouslymoreaggressiveclonesmaybeentirelyindependentofdifferentialsensitivitytotherapy(Figure2c).Werecentlyobservedahighernumberoflargesubclones(10%ofcancercells)in149CLLcasesthatwereexposedtotreatmentbeforesamplingcomparedwithpatientswhoreceivedtherapyafterthesamplewasobtained.Thisndingofincreasedclonaldiversitywithtreatmentheldtrueevenafteraccountingforpotentialconfounders,suchaslongerfollow-uptime.Weinterpretthisobservationtoresult,atleastinpart,fromtheoutgrowthofmanydiversepre-existingminorbuttsubclones.Thislatterinterpretationisfurthersupportedbyourobservationofanincreasedfrequencyofsubclonal-driverevents(presumablytter)intreatedrelativetountreatedpatients.Overall,ourdatasupporttheideathatCLLtherapy,bymarkedlyreducingdiseasebulk,mayactasaclassicevolutionaryrestrictionpointandresetinterclonalWithinthisconceptualframework,whensubcloneswithhightnessalreadyexistwithinatumorpopulation,treatmentcouldfavorthedevelopmentofmoreaggressiveclones,potentiallyreducingpost-relapsesurvival.Inthiscontext,cytotoxictherapywouldeffectivelyremovetheincumbentcloneactinglikeamassextinctioneventandtherebyshifttheevolutionaryinfavorofoneormoreaggressivesubclones.Thus,highlytsubclonesprobablybenetfromtreatmentandexhibitrapidoutgrowth.Thesedataprovidemechanisticsupporttotheobservationthatthewatchandwaitstrategyfor Figure2.Threemodelsofhowcancertherapymayaccelerateclonalevolution.First,cancertherapy,particularlycontaininggenotoxicagents,caninducenovelmutagenesis().Second,therapycanaccelerateclonalevolutionbyselectingaclone(hereillustratedinred)containingamutationthatconfersresistancetothetherapeuticagentused().Theresistanceoftheselectedcloneisreßectedinthedepictionofthecellpopulationaftercytoreduction,composedalmostentirelyoftheresistantclone(inred).Athirdmodelpostulatessimilarsensitivitytotreatmentofthedifferentsubpopulations,reßectedinsimilarproportionsbeforeandaftercytoreduction().TheclearingnichealtersthedynamicevolutionarylandscapeallowingafasterriseofaÞtterclone. Clonalevolutionandtherapeuticstrategies DALandauetalLeukemia(2014)34 432014MacmillanPublishersLimited CLLleadstosuperiorclinicalresults,astheearlieradministrationofchemotherapymayaccelerateclonalevolutionandtheemergenceofttercloneswithmoreaggressivediseasephenotypes.ThisformofrelationshipbetweentherapyandevolutionmaybeparticularlyimportanttoCLL,asthiscancertypeishighlydependentongrowthandsurvivalsignalsprovidedbythelocalmicroenvironment.ThisdependencymayaugmenttheimportanceoftheroleofinterclonalcompetitionintheevolutionarydynamicsofCLL.Futurein-depthstudieswouldassistinconrmingthismodelaswellaswhetheritisgeneralizabletomalignanciesotherthanCLL,andinparticularinothermoreindolentcancers.TRANSLATINGCLONALEVOLUTIONTOTHECLINICAmajorpriorityofprecisioncancergenomicsistouseinformationongeneticlesionstodenepatientprognosis.Inanillustrativeexample,Pateletal.coulddemonstratethatlesionssuchasinternaltandemduplicationof-ITD),partialtandemduplicationin-PTD),aswellasmutationsinassociatedwithreducedoverallsurvivalinAML,whilemutationsassociatedwithimprovedoverallsurvival.Theseassociationswereindependentofestablishedriskfactors.Similareffortshavebeencarriedoutinotherhematologicalmalignancies,includingCLL,andmultiplemyeloma.Acrossthebloodmalignancies,patientswithapparentlypoorprognosticmarkerscannonethelessexhibitgoodsurvival,andviceversa.Theseveralstudiesreviewedabove(Table1)suggestthatintratumoralclonalheterogeneitymaybeanimportantcontributortothiscomplexpicture.Inaggregate,studiesofclonalevolutionhaverevealedcancerstobegeneticallyheterogeneousinspaceandtime.Hence,simplylabelinganindividualcancerasharboringageneticlesionornotisnotfullyprecise.Fromapracticalstandpoint,forasolidtumormass,orevenleukemiacellsthatarepresentindifferenttissuecompartments(thatis,bloodvsmarrowvslymphnode),multiplesamplingsmayberequiredtocorrectlyassertthegeneticlandscapeofanindividualcase(Figure3).Finally,clonalheterogeneityinandofitselfmayimpactclinicaloutcome.Ourstudieshaveshownthatthepresenceofastrongsubclonal-driverevent,butnotaclonaldriver,negativelyimpactsclinicaloutcomeinCLL.Thelinkbetweenclonalheterogeneityandspecicallythepresenceofasubclonaldrivertoadverseclinicaloutcomeaddsanadditionaldimensiontothecurrenteffortsoflinkingdiscretesomaticmutationstooutcome.Inotherwords,itisnotonlythepresenceorabsenceofamutationthatshouldbeconsideredbutalsothesizeofthesubpopulationitFromthetherapeuticstandpoint,studiesofcancergenomicshighlighttheconceptthatcancerisnotasinglediseaseentitybutratheracollectionofrelateddisorders;hence,treatmentshouldbetargetedtothemolecularsubtypeofdisease.Forexample,high-dosedaunorubicin,ascomparedwithstandard-dosedau-norubicin,improvestherateofsurvivalamongpatientswithmutationsortranslocationsinAMLbutnotamongpatientswithwild-type,andpotentialtointegratetheavailablehigh-throughputsequencingtechnologies(forexample,DNA-seq,RNA-seqandChIP-seq)toprovideapatient-specicgenomic-epigenomicmapmayprovidecrucialprognosticperspectiveandinformtherapeuticchoices.Furthermore,targetedtreatmentsthatarebasedonthepresenceofspecicmolecularlesionsmaygreatlyimprovetherapeuticresponse,asseenin,forexample,associatedmyeloproliferativedisorders.Theseactionablemutationswhereaclinicianmatchesatumormutationtoacancerdrugmayeitherbemissedgivengeneticheterogeneityintimeandspace,oralternativelymightinvolveonlyasmallsubclone,whichbegsthequestionoftheclinicalefcacywereittobesolelytargeted.Forinstance,syntheticlethalapproacheswerefoundtobehighlyeffectiveinsituationsinwhichallcancercellscontainthetargetedvariation,aswitnessedbythepotentefcacyofPARP(polyADP-ribosepolymerase)inhibitionintumorsgermlinecarriers.Theseobservationstogetherraisestheprovocativequestionofwhetheritispreferabletotargetgeneticvariationsthatarefoundinthetrunkcomparedwiththosefoundinbranchesoftheevolutionaryphylogenetictree.Intuitively,theformermaybeconsideredthesuperiorapproach.Trunkevents,bydenition,aremutationspresentinallthecellsofthemalignantprocess.Targetingofthiseventintheorycarriesthepotentialofacompleteextinctionoftheentirepopulationofmalignantcells.Conversely,itisunclearwhetherthecellremainsdependentonthespecictrunktargetafteracquiringadditionaloncogenicevents(abranchtarget),andtherefore,howwelltheywillbeimpactedbytherapydirectedagainstthesefoundertargets.Insolidmalignancies,suchasnon-smallcelllungcancer,mutationsarerarelydetectedtogether;however,whentheyco-occur,targetingthetrunk-typemutation(thatis,)isnolongereffective.canonicalmutationsarediscoveredinbenigncolonicpolypsandarethereforeprobablytobeearlier,trunk-typeevents.However,theresponsetotargetinghasbeendisappointing.Onemayhypothesizethatatleastinmoreindolentmalignancies,targetingbranchmutations(pruning)maybeaneffectivestrategy,whichcouldpromoteclonalequilibriumandhindertheselectionofmoreaggressiveThedifferentialeffectsoftargetingbranchvstrunklesionsmaybedetermined,inpart,bythecomplexepistaticrelationshipbetweendifferentgeneticlesionswithinthesameclonalpopulation.Asnewmutationsdonotoccurinisolationbutratherenterintoanestablishedgenomiclandscape,theexistinggenenetworkmayhaveaprofoundeffectonthefateofthecellanddeterminewhetherthenovelmutationwillresultincelldeathorclonalexpansion.Forexample,activationofmanyoncogenestogether,including,canleadtoastateofoncogene-inducedAsimilarrelationshiphasbeendemonstrated-inducedapoptosisthatisrelievedinthecontext Figure3.Translatingclonalheterogeneityinsightstotheclinic.Possibleprognosticandtherapeuticimplicationsofclonalheterogeneityareoutlined.ImagecourtesyBroadInstitute/LaurenSolomon;hourglassphotoiStockphoto/DominikPabis. ClonalevolutionandtherapeuticstrategiesDALandauetal 2014MacmillanPublishersLimitedLeukemia(2014)34 43 Hence,furtherstudyoftheepistaticrelationshipsinmodelsystemsaswellasinclinicaltrialswillhelpclarifyinwhatcontextoptimaleffectswillresultfromtargetingthetrunkeventandwhenitispreferabletotargetthebranches.Thebroaderevolutionaryperspectiveallowsustoviewcancerasanecologyofdifferentsubpopulationsinthecontextoftheirenvironment.Intriguingdatasuggestthat,atleastinsomecases,complexco-dependencyrelationshipsbetweensubpopulationsmayexist,inadditiontocompetition.Theunderstandingthatdiseaseiscomposedofdiversesubpopulationsisachallengetoourtraditionalschemesofclinicaltrials.Afutureinwhichbothtrunkandbrancheventsarecharacterized,andinwhichnotwocancerssharethesamegenomicfeatures,maybeenvisaged.Inthissetting,performinglarge-scaleclinicaltrialsusingpresent-daymethodologies,inparticularutilizingcombinationsoftargetedagents,mayprovehighlychallenging.Thediseasecannolongerbedenedasasingleentitycontainingauniformsetofgeneticabnormalities.Furthermore,thedegreeofgeneticheterogeneityofatumorislikelytobeanimportantdeterminantoftherapeuticAbetterunderstandingisneededoftheimpactoftherapyontheevolutionarylandscape,possiblythroughtheuseofthemoreapplicablewhole-exomesequencingtechnologiestostudylargecohortsonpatients.ResearchersmayconsiderincorporatingapproachessuchasWES(wholeexomesequencing)thatidentifyatleastlargersubpopulations(1 10%ofcancermass)andcharacterizetheirevolutioninancillarystudiesofprospectiveclinicaltrials.Suchinformationmayinformusregardingtheadaptiveprocessesresponsiblefortreatmentfailureaswellaseventuallysparkthedevelopmentofnoveltherapeuticparadigms.Forone,ithasbeenproposedthatalternativeapproachescouldpotentiallymaintaininterclonalequilibriumattheexpenseoftryingtomaximizecellkill.Thisapproachsupportspreventingtheeliminationoftherapy-sensitiveclones,asthey(theoretically)couldcontinuetosuppressthegrowthoftherapy-resistantclonesinacompetitivemannerandtherebymaintainanequilibriumstate.Asecondapproachthatrequiresfurtherconsiderationistheideaoflimitingtheunderlyingdiversicationthatservesasthesubstrateforclonalevolutionbeforethefullexpressionofthegeneticortheepigeneticheterogeneityincancerisevident.Finally,thetherapeuticchallengeposedbyacontinuouslyadaptingandreshapingmalignantprocessprovidesstrongrationaletosupportthepursuitofimmunity-basedtherapies,asthisapproachmayeffectivelypitonecomplexadaptiveprocessagainstanother.Thereisalreadylimitedevidencethatallogeneichematopoieticstemcelltransplant(anon-specicexampleofaimmunity-basedtherapy)imposesevolutionarypressuresonthetumorthataredistinctfromothertherapeuticmodalities(leading,forexample,tolossofdonor recipientmismatchedormultiplecytogeneticabnormalities).Thesealterationsdemonstratethattheleukemiccellpopulationisbeingmoldedbyapowerfulimmuneresponseandhencetotheefcacyoftheimmunity-basedtherapy.Theprocessofco-evolutionofthecancercellsandtheimmuneresponseinthesettingofeffectiveimmunotherapyisanareaofgreatinterestforfuturestudy.CONCLUSIONSANDFUTUREDIRECTIONSUnderstandingtheevolutionarycapacityofcancerisemergingasakeyelementindevelopingimprovedtherapeuticstrategiesintheeraofprecisionmedicine,asitpresentsoneofthemostformidableobstaclestothesuccessfulapplicationoftargetedAsaforementioned,theintensiveapplicationofhigh-throughputgenomicplatformshasenabledrapidprogressinourunderstandingoftheprocessofclonalevolutioninhematologicalmalignancies.Collectively,thesestudieshaveprovidedseveralcoreinsights,including:rst,thatclonalheterogeneityiscommoninmalignancybothatthegeneticandtheepigeneticlevel;second,thatclonalevolutionisfrequentlyobservedinrelationtotherapy,leadingtoemer-genceofmoreaggressiveandresistantdisease;andnally,thattheprocessofclonalevolutionislinkedtoadverseclinicalAlthoughtheserecentstudiespointtothekeyroleplayedbyclonalevolutionincancerprogression,currentperspectivesofcancerasanevolutionaryproblemarefairlylimited.Evenasknowledgeaboutgermlineandacquiredgeneticlesionsasso-ciatedwithcancershasgrownexponentially,westillknowonlylittleaboutthebackgroundrateofheterogeneitywhichisthesubstrateofevolutionandpossessonlyarudimentaryunder-standingofhowtheepigeneticprogramaffectsthissubstrate.Inthisrespect,developingmethodologiestointegratedatafromcomplementarygeneticandepigenetichigh-throughputplat-formsiskey,bothatthecellpopulationandatthesingle-celllevel.Moreover,thedynamicsofinteractionsbetweencloneswhethertheycompeteorco-dependoneachotherhasnotbeenelucidated.Additionally,theexaminationofkeymechanisticquestionrelatingtoclonalevolutionusinggenomictools(forexample,differenttypesofselectivepressure,interactionwithmicroenvironmentnichesandinteractionsbetweenmultiplegeneticlesionswithinthesamecell)hasyettobeaccomplished.Thus,theabilitytoforeseetheevolutionarytrajectoryofanyindividualcancerispresentlystillinitsinfancy.Improvingthiscapacitytopredicthowcancerwillevolvewithtreatmentcarriesasignicantpotentialtoallowustoanticipateandtailortreatmenttotheprobablefuturetrajectory(so-calledanticipation-basedOngoingtechnologicaldevelopmentsarenowgeneratingtoolsideallysuitedforthestudyofthesequestions.Recently,proof-of-principlestudiesofsingle-cellsequencinghavebeenconductedthathavecataloguedthepointmutationsinprotein-codingregions.115,116Inthenottoodistantfuture,single-cellsequencingwillallowthedetailedstudyofgeneticheterogeneitythatprovidethebackdropagainstwhichevolutionatthesubpopulationleveloccurs.Theapplicationofsingle-cellRNA-seq117tothestudyofhematologicalmalignancieswouldenablethestudyoftheheterogeneoustranscriptionalchangesandsignalingnetworksthatstemfromheterogeneoussomaticgeneticalterations.Inadditiontosingle-cellexamination,novelmethodologiesarecapableofdeconvolutingsubpopulationsfrombulkmaterialandmaybehelpfulindelineatingthebasicunderlyingprinciplesofevolutionininvivoinvitromodels,withdeepersequencingprovidingbothhighersensitivityforsmallersubclonesaswellasmorepreciseestimatesoftheirsize.TheabilitytouseWESasanalternativeapproachtoWGS,aswellastheprojecteddowntrendofsequencingcosts,willenablemulti-samplingintimeandspaceofhematologicalmalignancies,clarifyingthenatureofspatialheterogeneityinbloodmalignanciesaswellasquestionsregardingthenaturerepopulationoftheecologicalnicheuponrelapse.Itmayalsoallowthestudyofthesequestionsinlargeclinicaltrials.Theseeffortscanpotentiallyanswerquestionsoffundamentalimportancetotheclinicalapplicationoftheseinsights,namelywhatistheprognosticsignicanceofthesizeofthesubclonethatharborsageneticmarker,whenshouldwetargetbranchortrunklesionsandhowtointegratethisknowledgeincombinatorialtherapies.CONFLICTOFINTERESTTheauthorsdeclarenoconictofinterest. Clonalevolutionandtherapeuticstrategies DALandauetalLeukemia(2014)34 432014MacmillanPublishersLimited DALacknowledgessupportbytheAmericanSocietyofHematology(ResearchAwardforFellows-in-Training),andtheAmericanCancerSociety.CJWacknowledgessupportfromtheBlavatnikFamilyFoundation,theLymphomaResearchFoundation,NHLBI(1RO1HL103532-01;1RO1HL116452-01)andNCI(1R01CA155010-01A1)andisarecipientofaLeukemiaLymphomaTranslationalResearchProgramAwardandanAACRSU2CInnovativeResearchGrant.WethankallmembersoftheBroadInstitutesBiologicalSamplesandGenomeSequencingPlatforms,aswellastheCancerGenomeAnalysisgroupwhomadethisworkpossible(NHGRI-U54HG003067).1DrukerBJ,TalpazM,RestaDJ,PengB,BuchdungerE,FordJMetal.EfcacyandsafetyofaspecicinhibitoroftheBCR-ABLtyrosinekinaseinchronicmyeloidleukemia.NEnglJMed:1031 1037.2Lo-CocoF,AvvisatiG,VignettiM,ThiedeC,OrlandoSM,IacobelliSetal.acidandarsenictrioxideforacutepromyelocyticleukemia.NEnglJMed:111 121.3LynchTJ,BellDW,SordellaR,GurubhagavatulaS,OkimotoRA,BranniganBWetal.Activatingmutationsintheepidermalgrowthfactorreceptorunderlyingresponsivenessofnon-small-celllungcancertogetinib.NEnglJMed2004;3502129 2139.4FlahertyKT,PuzanovI,KimKB,RibasA,McArthurGA,SosmanJAetal.ofmutated,activatedBRAFinmetastaticmelanoma.NEnglJMed809 819.5ArkenauHT,KeffordR,LongGV.TargetingBRAFforpatientswithmelanoma.BrJCancer:392 398.6WestbrookCA,HoobermanAL,SpinoC,DodgeRK,LarsonRA,DaveyFetal.ClinicalsignicanceoftheBCR-ABLfusiongeneinadultacutelymphoblasticleukemia:aCancerandLeukemiaGroupBStudy(8762).2983 2990.7LugoTG,PendergastAM,MullerAJ,WitteON.Tyrosinekinaseactivityandtransformationpotencyofbcr-abloncogeneproducts.1079 1082.8DrukerBJ,SawyersCL,KantarjianH,RestaDJ,ReeseSF,FordJMetal.ActivityofaspecicinhibitoroftheBCR-ABLtyrosinekinaseintheblastcrisisofchronicmyeloidleukemiaandacutelymphoblasticleukemiawiththePhiladelphiachromosome.NEnglJMed:1038 1042.9OttmannOG,DrukerBJ,SawyersCL,GoldmanJM,ReiffersJ,SilverRTetal.Aphase2studyofimatinibinpatientswithrelapsedorrefractoryPhiladelphiachromosome-positiveacutelymphoidleukemias.:1965 1971.10FoaR,VitaleA,VignettiM,MeloniG,GuariniA,DeProprisMSetal.Dasatinibasrst-linetreatmentforadultpatientswithPhiladelphiachromosome-positiveacutelymphoblasticleukemia.Blood:6521 6528.11MullighanCG,PhillipsLA,SuX,MaJ,MillerCB,ShurtleffSAetal.Genomicanalysisoftheclonaloriginsofrelapsedacutelymphoblasticleukemia.Science:1377 1380.12HanahanD,WeinbergRA.Hallmarksofcancer:thenextgeneration.Cell:646 674.13WaandersE,ScheijenB,vanderMeerLT,vanReijmersdalSV,vanEmstL,KroezeYetal.TheoriginandnatureoftightlyclusteredBTG1deletionsinprecursorB-cellacutelymphoblasticleukemiasupportamodelofmulticlonalevolution.PLoSGenet2012;:e1002533.14MerloLM,PepperJW,ReidBJ,MaleyCC.Cancerasanevolutionaryandecologicalprocess.NatRevCancer:924 935.15HarrisJF,ChambersAF,HillRP,LingV.MetastaticvariantsaregeneratedspontaneouslyatahighrateinmouseKHTtumor.ProcNatlAcadSciUSA:5547 5551.16CilloC,DickJE,LingV,HillRP.Generationofdrug-resistantvariantsinmetastaticB16mousemelanomacelllines.CancerRes:2604 2608.17FearonER,VogelsteinB.Ageneticmodelforcolorectaltumorigenesis.:759 767.18ZhangL,ZnoykoI,CostaLJ,ConlinLK,DaberRD,SelfSEetal.ClonaldiversityanalysisusingSNPmicroarray:anewprognostictoolforchroniclymphocyticCancerGenet:654 665.19CalissanoC,DamleR,HayesG,MurphyE,HellersteinM,MorenoCetal.InvivointraclonalandinterclonalkineticheterogeneityinB-cellchroniclymphocytic:4832 4842.20BagnaraD,CalleaV,StelitanoC,MorabitoF,FabrisS,NeriAetal.IgVgeneintraclonaldiversicationandclonalevolutioninB-cellchroniclymphocyticBrJHaematol:50 58.21StilgenbauerS,SanderS,BullingerL,BennerA,LeupoltE,WinklerDetal.evolutioninchroniclymphocyticleukemia:acquisitionofhigh-riskgenomicaberrationsassociatedwithunmutatedVH,resistancetotherapy,andshort:1242 1245.22GunnarssonR,MansouriL,IsakssonA,GoranssonH,CahillN,JanssonMetal.Array-basedgenomicscreeningatdiagnosisandduringfollow-upinchroniclymphocyticleukemia.:1161 1169.23NowellPC.Theclonalevolutionoftumorcellpopulations.Science24FidlerIJ.TumorheterogeneityandthebiologyofcancerinvasionandCancerRes:2651 2660.25CairoMS,JordanCT,MaleyCC,ChaoC,MelnickA,ArmstrongSAetal.NCIrstInternationalWorkshoponthebiology,prevention,andtreatmentofrelapseafterallogeneichematopoieticstemcelltransplantation:reportfromthecom-mitteeonthebiologicalconsiderationsofhematologicalrelapsefollowingallogeneicstemcelltransplantationunrelatedtograft-versus-tumoreffects:stateofthescience.BiolBloodMarrowTransplant:709 728.26LoebLA.Humancancersexpressmutatorphenotypes:origin,consequencesandNatRevCancer:450 457.27BassaganyasL,BeaS,EscaramisG,TornadorC,SalaverriaI,ZapataLetal.Sporadicandreversiblechromothripsisinchroniclymphocyticleukemiarevealedbylongitudinalgenomicanalysis.Leukemia2013;E-publication:2013/04/25.28BouwmanP,JonkersJ.TheeffectsofderegulatedDNAdamagesignallingoncancerchemotherapyresponseandresistance.NatRevCancer29SchwartzM,ZlotorynskiE,KeremB.Themolecularbasisofcommonandrarefragilesites.CancerLett:13 26.30LawrenceMS,StojanovP,PolakP,KryukovGV,CibulskisK,SivachenkoAetal.Mutationalheterogeneityincancerandthesearchfornewcancer-associated:214 218.31GawadC,PepinF,CarltonVE,KlingerM,LoganAC,MiklosDBetal.evolutionoftheimmunoglobulinheavychainlocusinchildrenwithBprecursoracutelymphoblasticleukemia.Blood:4407 4417.32MahowaldGK,BaronJM,SleckmanBP.Collateraldamagefromantigenreceptorgenediversication.:1009 1012.33MeyersonM,GabrielS,GetzG.Advancesinunderstandingcancergenomesthroughsecond-generationsequencing.NatRevGenet:685 696.34PleasanceE,CheethamR,StephensP,McBrideD,HumphrayS,GreenmanCetal.Acomprehensivecatalogueofsomaticmutationsfromahumancancergenome.Nature2010;:191 196.35FutrealPA,CoinL,MarshallM,DownT,HubbardT,WoosterRetal.Acensusofhumancancergenes.NatRevCancer:177 183.36GerlingerM,RowanAJ,HorswellS,LarkinJ,EndesfelderD,GronroosEetal.IntratumorheterogeneityandbranchedevolutionrevealedbymultiregionNEnglJMed:883 892.37CarterSL,CibulskisK,HelmanE,McKennaA,ShenH,ZackTetal.quanticationofsomaticDNAalterationsinhumancancer.NatBiotechnol:413 421.38Nik-ZainalS,VanLooP,WedgeDC,AlexandrovLB,GreenmanCD,LauKWetal.Thelifehistoryof21breastcancers.:994 1007.39DingL,LeyTJ,LarsonDE,MillerCA,KoboldtDC,WelchJSetal.Clonalevolutioninrelapsedacutemyeloidleukaemiarevealedbywhole-genomesequencing.:506 510.40LandauD,CarterS,StojanovP,McKennaA,StevensonK,LawrenceMetal.Evolutionandimpactofsubclonalmutationsinchroniclymphocyticleukemia.:714 726.41BeroukhimR,MermelCH,PorterD,WeiG,RaychaudhuriS,DonovanJetal.Thelandscapeofsomaticcopy-numberalterationacrosshumancancers.:899 905.42ShahSP,RothA,GoyaR,OloumiA,HaG,ZhaoYetal.Theclonalandmutationalevolutionspectrumofprimarytriple-negativebreastcancers.43MaleyCC,GalipeauPC,LiX,SanchezCA,PaulsonTG,ReidBJ.Selectivelyadvantageousmutationsandhitchhikersinneoplasms:p16lesionsareselectedinBarrettsesophagus.CancerRes:3414 3427.44BozicI,AntalT,OhtsukiH,CarterH,KimD,ChenSetal.Accumulationofdriverandpassengermutationsduringtumorprogression.ProcNatlAcadSciUSA:18545 18550.45WelchJS,LeyTJ,LinkDC,MillerCA,LarsonDE,KoboldtDCetal.Theoriginandevolutionofmutationsinacutemyeloidleukemia.46WalterMJ,ShenD,DingL,ShaoJ,KoboldtDC,ChenKetal.Clonalarchitectureofsecondaryacutemyeloidleukemia.NEnglJMed:1090 1098.47EganJB,ShiCX,TembeW,ChristoforidesA,KurdogluA,SinariSetal.Whole-genomesequencingofmultiplemyelomafromdiagnosistoplasmacellleukemiarevealsgenomicinitiatingevents,evolution,andclonaltides.:1060 1066. ClonalevolutionandtherapeuticstrategiesDALandauetal 2014MacmillanPublishersLimitedLeukemia(2014)34 43 48KeatsJJ,ChesiM,EganJB,GarbittVM,PalmerSE,BraggioEetal.competitionwithalternatingdominanceinmultiplemyeloma.49SnuderlM,FazlollahiL,LeLP,NittaM,ZhelyazkovaBH,DavidsonCJetal.amplicationofmultiplereceptortyrosinekinasegenesinglioblastoma.:810 817.50EneCI,FineHA.Manytumorsinone:adauntingtherapeuticprospect.:695 697.51QuentmeierH,AminiRM,BerglundM,DirksWG,EhrentrautS,GeffersRetal.U-2932:twoclonesinonecellline,atoolforthestudyofclonalevolution.:1155 1164.52CampbellPJ,PleasanceED,StephensPJ,DicksE,RanceR,GoodheadIetal.Subclonalphylogeneticstructuresincancerrevealedbyultra-deepsequencing.ProcNatlAcadSciUSA:13081 13086.53SmithAE,MohamedaliAM,KulasekararajA,LimZ,GakenJ,LeaNCetal.Next-generationsequencingoftheTET2genein355MDSandCMMLpatientsrevealslow-abundancemutantcloneswithearlyorigins,butindicatesnodeniteprognosticvalue.:3923 3932.54NavinN,KendallJ,TrogeJ,AndrewsP,RodgersL,McIndooJetal.Tumourevolutioninferredbysingle-cellsequencing.:90 94.55SmithZD,MeissnerA.DNAmethylation:rolesinmammaliandevelopment.RevGenet:204 220.56SiegmundKD,MarjoramP,WooYJ,TavareS,ShibataD.InferringclonalexpansionandcancerstemcelldynamicsfromDNAmethylationpatternsincolorectalcancers.ProcNatlAcadSciUSA:4828 4833.57DeS,ShaknovichR,RiesterM,ElementoO,GengH,KormakssonMetal.AberrationinDNAmethylationinB-celllymphomashasacomplexoriginandincreaseswithdiseaseseverity.PLoSGenet:e1003137.58AndersonK,LutzC,vanDelftFW,BatemanCM,GuoY,ColmanSMetal.variegationofclonalarchitectureandpropagatingcellsinleukaemia.:356 361.59ClappierE,GerbyB,SigauxF,DelordM,TouzriF,HernandezLetal.selectioninxenograftedhumanTcellacutelymphoblasticleukemiarecapitu-latesgainofmalignancyatrelapse.JExpMed:653 661.60PiccirilloSG,CombiR,CajolaL,PatriziA,RedaelliS,BentivegnaAetal.poolsofcancerstem-likecellscoexistwithinhumanglioblastomasanddisplaydifferenttumorigenicityandindependentgenomicevolution.Oncogene:1807 1811.61KresoA,OBrienCA,vanGalenP,GanOI,NottaF,BrownAMetal.Variableclonalrepopulationdynamicsinuencechemotherapyresponseincolorectalcancer.:543 548.62SpencerSL,GaudetS,AlbeckJG,BurkeJM,SorgerPK.Non-geneticoriginsofcell-to-cellvariabilityinTRAIL-inducedapoptosis.:428 432.63GreavesM.Cancerstemcells:backtoDarwin?SeminCancerBiol:65 70.64AriasAM,HaywardP.Filteringtranscriptionalnoiseduringdevelopment:conceptsandmechanisms.NatRevGenet:34 44.65BalazsiG,vanOudenaardenA,CollinsJJ.Cellulardecisionmakingandbiologicalnoise:frommicrobestomammals.:910 925.66GuptaPB,ChafferCL,WeinbergRA.Cancerstemcells:mirageorreality?NatMed:1010 1012.67JamiesonCH,AillesLE,DyllaSJ,MuijtjensM,JonesC,ZehnderJLetal.Granulocyte-macrophageprogenitorsascandidateleukemicstemcellsinblast-crisisCML.NEnglJMed:657 667.68GuptaPB,OnderTT,JiangG,TaoK,KuperwasserC,WeinbergRAetal.Identicationofselectiveinhibitorsofcancerstemcellsbyhigh-throughput:645 659.69MizunoH,SpikeBT,WahlGM,LevineAJ.Inactivationofp53inbreastcancerscorrelateswithstemcelltranscriptionalsignatures.ProcNatlAcadSciUSA:22745 22750.70CicaleseA,BonizziG,PasiCE,FarettaM,RonzoniS,GiuliniBetal.Thetumorsuppressorp53regulatespolarityofself-renewingdivisionsinmammarystem:1083 1095.71HuntlyBJ,ShigematsuH,DeguchiK,LeeBH,MizunoS,DuclosNetal.butnotBCR-ABL,conferspropertiesofleukemicstemcellstocommittedmurinehematopoieticprogenitors.CancerCell:587 596.72QuintanaE,ShackletonM,SabelMS,FullenDR,JohnsonTM,MorrisonSJ.Efcienttumourformationbysinglehumanmelanomacells.73PeceS,TosoniD,ConfalonieriS,MazzarolG,VecchiM,RonzoniSetal.andmolecularheterogeneityofbreastcancerscorrelateswiththeircancerstemcellcontent.:62 73.74FieldingAK,RichardsSM,ChopraR,LazarusHM,LitzowMR,BuckGetal.Outcomeof609adultsafterrelapseofacutelymphoblasticleukemia(ALL);anMRCUKALL12/ECOG2993study.:944 950.75ParsonsD,JonesS,ZhangX,LinJ,LearyR,AngenendtPetal.Anintegratedgenomicanalysisofhumanglioblastomamultiforme.76SchuhA,BecqJ,HumphrayS,AlexaA,BurnsA,CliffordRetal.chroniclymphocyticleukemiaprogressionbywholegenomesequencingrevealsheterogeneousclonalevolutionpatterns.:4191 4196.77WangL,LawrenceMS,WanY,StojanovP,SougnezC,StevensonKetal.SF3B1andothernovelcancergenesinchroniclymphocyticleukemia.NEnglJMed:2497 2506.78GreavesM,MaleyCC.Clonalevolutionincancer.Nature:306 313.79MaheswaranS,SequistLV,NagrathS,UlkusL,BranniganB,ColluraCVetal.DetectionofmutationsinEGFRincirculatinglung-cancercells.NEnglJMed:366 377.80EdwardsSL,BroughR,LordCJ,NatrajanR,VatchevaR,LevineDAetal.ResistancetotherapycausedbyintragenicdeletioninBRCA2.81InukaiM,ToyookaS,ItoS,AsanoH,IchiharaS,SohJetal.PresenceofepidermalgrowthfactorreceptorgeneT790Mmutationasaminorcloneinnon-smallcelllungcancer.CancerRes:7854 7858.82CleversH.Thecancerstemcell:premises,promisesandchallenges.NatMed:313 319.83YipS,MiaoJ,CahillDP,IafrateAJ,AldapeK,NuttCLetal.MSH6mutationsariseinglioblastomasduringtemozolomidetherapyandmediatetemozolomideClinCancerRes:4622 4629.84ShahNP,NicollJM,NagarB,GorreME,PaquetteRL,KuriyanJetal.MultipleBCR-ABLkinasedomainmutationsconferpolyclonalresistancetothetyrosinekinaseinhibitorimatinib(STI571)inchronicphaseandblastcrisischronicmyeloidleukemia.CancerCell:117 125.85Roche-LestienneC,Soenen-CornuV,Grardel-DuosN,LaiJL,PhilippeN,FaconTetal.SeveraltypesofmutationsoftheAblgenecanbefoundinchronicmyeloidleukemiapatientsresistanttoSTI571,andtheycanpre-existtotheonsetof:1014 1018.86TzonevaG,Perez-GarciaA,CarpenterZ,KhiabanianH,ToselloV,AllegrettaMetal.ActivatingmutationsintheNT5C2nucleotidasegenedrivechemotherapyresistanceinrelapsedALL.NatMed2013;:368 371.87WuCJ.CLLclonalheterogeneity:anecologyofcompetingsubpopulations.:4117 4118.88MorganGJ,WalkerBA,DaviesFE.Thegeneticarchitectureofmultiplemyeloma.NatRevCancer:335 348.89JablonskiD.Lessonsfromthepast:evolutionaryimpactsofmassextinctions.ProcNatlAcadSciUSA:5393 5398.90NowakMA,SigmundK.Evolutionarydynamicsofbiologicalgames.:793 799.91VincentTL,GatenbyRA.Anevolutionarymodelforinitiation,promotion,andprogressionincarcinogenesis.IntJOncol:729 737.92MaleyCC,GalipeauPC,FinleyJC,WongsurawatVJ,LiX,SanchezCAetal.Geneticclonaldiversitypredictsprogressiontoesophagealadenocarcinoma.NatGenet:468 473.93CLLTrialistsCollaborativeGroup.Chemotherapeuticoptionsinchroniclymphocyticleukemia:ameta-analysisoftherandomizedtrials.JNatlCancerInst:861 868.94BurgerJA,GhiaP,RosenwaldA,Caligaris-CappioF.ThemicroenvironmentinmatureB-cellmalignancies:atargetfornewtreatmentstrategies.:3367 3375.95PatelJP,GonenM,FigueroaME,FernandezH,SunZ,RacevskisJetal.relevanceofintegratedgeneticprolinginacutemyeloidleukemia.NEnglJMed:1079 1089.96DohnerH,StilgenbauerS,BennerA,LeupoltE,KroberA,BullingerLetal.Genomicaberrationsandsurvivalinchroniclymphocyticleukemia.NEnglJMed:1910 1916.97RossiD,RasiS,SpinaV,BruscagginA,MontiS,CiardulloCetal.Integratedmutationalandcytogeneticanalysisidentiesnewprognosticsubgroupsinchroniclymphocyticleukemia.:1403 1412.98Avet-LoiseauH,LiC,MagrangeasF,GouraudW,CharbonnelC,HarousseauJLetal.Prognosticsignicanceofcopy-numberalterationsinmultiplemyeloma.JClinOncol2009;:4585 4590.99SwantonC.Intratumorheterogeneity:evolutionthroughspaceandtime.:4875 4882.100CancerGenomeAtlasResearchNetwork.Genomicandepigenomiclandscapesofadultdenovoacutemyeloidleukemia.NEnglJMed:2059 2074.101PardananiA,TefferiA.Imatinibtherapyforhypereosinophilicsyndromeandeosinophilia-associatedmyeloproliferativedisorders.LeukRes(Suppl1): Clonalevolutionandtherapeuticstrategies DALandauetalLeukemia(2014)34 432014MacmillanPublishersLimited 102FongPC,BossDS,YapTA,TuttA,WuP,Mergui-RoelvinkMetal.Inhibitionofpoly(ADP-ribose)polymeraseintumorsfromBRCAmutationcarriers.NEnglJMed:123 134.103ZhuCQ,daCunhaSantosG,DingK,SakuradaA,CutzJC,LiuNetal.RoleofKRASandEGFRasbiomarkersofresponsetoerlotinibinNationalCancerInstituteofCanadaClinicalTrialsGroupStudyBR.21.JClinOncol:4268 4275.104YangH,HigginsB,KolinskyK,PackmanK,BradleyWD,LeeRJetal.activityofBRAFinhibitorvemurafenibinpreclinicalmodelsofBRAF-mutantcolorectalcancer.CancerRes:779 789.105SerranoM,LinAW,McCurrachME,BeachD,LoweSW.Oncogenicrasprovokesprematurecellsenescenceassociatedwithaccumulationofp53andp16INK4a.:593 602.106BissonnetteRP,EcheverriF,MahboubiA,GreenDR.Apoptoticcelldeathinducedbyc-mycisinhibitedbybcl-2.Nature:552 554.107IndaMM,BonaviaR,MukasaA,NaritaY,SahDW,VandenbergSetal.TumorheterogeneityisanactiveprocessmaintainedbyamutantEGFR-inducedcytokinecircuitinglioblastoma.GenesDev:1731 1745.108GilliesRJ,VerduzcoD,GatenbyRA.Evolutionarydynamicsofcarcinogenesisandwhytargetedtherapydoesnotwork.NatRevCancer:487 493.109MerloLM,MaleyCC.Theroleofgeneticdiversityincancer.JClinInvest:401 403.110GatenbyRA,SilvaAS,GilliesRJ,FriedenBR.Adaptivetherapy.CancerRes:4894 4903.111WaterhouseM,PfeiferD,PanticM,EmmerichF,BertzH,FinkeJ.Genome-wideprolinginAMLpatientsrelapsingafterallogeneichematopoieticcelltransplantation.BiolBloodMarrowTransplant:1450 1459,e1.112StolzelF,HackmannK,KuithanF,MohrB,FusselM,OelschlagelUetal.evolutionincludingpartiallossofhumanleukocyteantigengenesfavoringextramedullaryacutemyeloidleukemiarelapseaftermatchedrelatedallogeneichematopoieticstemcelltransplantation.Transplantation:744 749.113BacherU,HaferlachT,AlpermannT,ZengerM,KrogerN,BeelenDWetal.ComparisonofcytogeneticclonalevolutionpatternsfollowingallogeneichematopoietictransplantationversusconventionaltreatmentinpatientsatrelapseofAML.BiolBloodMarrowTransplant:1649 1657.114PuenteXS,Lopez-OtinC.TheevolutionarybiographyofchroniclymphocyticNatGenet:229 231.115HouY,SongL,ZhuP,ZhangB,TaoY,XuXetal.Single-cellexomesequencingandmonoclonalevolutionofaJAK2-negativemyeloproliferativeneoplasm.:873 885.116XuX,HouY,YinX,BaoL,TangA,SongLetal.Single-cellexomesequencingrevealssingle-nucleotidemutationcharacteristicsofakidneytumor.:886 895.117ShalekAK,SatijaR,AdiconisX,GertnerRS,GaublommeJT,RaychowdhuryRetal.Single-celltranscriptomicsrevealsbimodalityinexpressionandsplicinginimmunecells.Nature:236 240.118Weston-BellN,GibsonJ,JohnM,EnnisS,PfeiferS,CezardTetal.Exomesequencingintrackingclonalevolutioninmultiplemyelomafollowingtherapy.:1188 1191.119GreenMR,GentlesAJ,NairRV,IrishJM,KihiraS,LiuCLetal.Hierarchyinsomaticmutationsarisingduringgenomicevolutionandprogressionoffollicular:1604 1611. ClonalevolutionandtherapeuticstrategiesDALandauetal 2014MacmillanPublishersLimitedLeukemia(2014)34 43