28car

DOI:10.1126/scitranslmed.3008226

,224ra25(2014);6SciTranslMed

AcuteLymphoblasticLeukemia

EfficacyandToxicityManagementof19-28zCARTCellTherapyinBCell

Editor'sSummary

CARTcelltherapy.

atasupporttheneedforfurthermulticentertrialsfor

especiallyimportantbecautreatmentfor

shouldallowforidentificationofthesubtofpatientswhowilllikelyrequiretherapeuticinterventionwith

undthatrumC-reactiveprotein(CRP)associatedwiththeverityofCRS,which

theauthorscarefullycharacterizedcytokinereleasyndrome(CRS),whichisariesoftoxicitiesassociatedwith

er,−−patientstotransitiontoallogeneichematopoieticstemcelltransplantation

TheCD19-targetingCARTcelltherapyresultedinan88%completeresponrate,whichallowedmostofthe

16relapdorrefractoryadultpatients.

imericantigenreceptors(CARs).Davila

gingtherapyforadultB-ALListhroughTcellsthattargettumorcells

RelapdorrefractoryBacutelymphoblasticleukemia(B-ALL)inadultshasapoorprognosis,withanexpected

CARvingOutaNicheforCARTCellImmunotherapy

/content/6/224/

canbefoundat:

andotherrvices,includinghigh-resolutionfigures,Acompleteelectronicversionofthisarticle

/content/suppl/2014/02/14/

canbefoundintheonlineversionofthisarticleat:SupplementaryMaterial

/content/scitransmed/5/215/

/content/scitransmed/3/95/

/content/scitransmed/4/132/

/content/scitransmed/5/177/

canbefoundonlineat:RelatedResourcesforthisarticle

/about/

inwholeorinpartcanbefoundat:article

permissiontoreproducethisofthisarticleoraboutobtainingreprintsInformationaboutobtaining

le

NW,Washington,ght2014bytheAmericanAssociationfortheAdvancementofScience;all

lastweekinDecember,bytheAmericanAssociationfortheAdvancementofScience,1200NewYorkAvenue

(printISSN1946-6234;onlineISSN1946-6242)ispublishedweekly,excepttheScienceTranslationalMedicine

o

n

F

e

b

r

u

a

r

y

2

1

,

2

0

1

4

s

t

m.

s

ci

e

n

c

e

m

a

g.

o

r

g

D

o

w

nl

o

a

d

e

d

f

r

o

m

CANCER

EfficacyandToxicityManagementof19-28zCARTCell

TherapyinBCellAcuteLymphoblasticLeukemia

,1IsabelleRiviere,1,2,3,4XiuyanWang,4ShirleyBartido,4JaePark,1

KevinCurran,,1JolantaStefanski,4OrianaBorquez-Ojeda,4

MalgorzataOlszewska,4JinrongQu,4TeresaWasielewska,4QingHe,4MitsuFink,4

HimalyShinglot,4MaherYoussif,4MarkSatter,4YongzengWang,4JamesHoy,4

HildaQuintanilla,1ElizabethHalton,1YvetteBernal,sira,,6

MithatGonen,,8PeterMaslak,1DanDouer,ni,9SergioGiralt,1,2

MichelSadelain,1,2,3*RenierBrentjens1,2,3*

Wereporton16patientswithrelapdorrefractoryBcellacutelymphoblasticleukemia(B-ALL)thatwetreatedwith

autologousTcellxpressingthe19-28zchimericantigenreceptor(CAR)rall

completeresponratewas88%,whichallowedustotransitionmostofthepatientstoastandard-of-carealloge-

neichematopoieticstemcelltransplant(allo-SCT).Thistherapywasaffectiveinhigh-riskpatientswithPhiladelphia

chromosome–positive(Ph+)hsystematic

analysisofclinicaldataandrumcytokinelevelsoverthefirst21daysafterTcellinfusion,wehavedefineddiagnostic

criteriaforaverecytokinereleasyndrome(sCRS),withthegoalofbetteridentifyingthesubtofpatientswho

willlikelyrequiretherapeuticinterventionwithcorticosteroidsorinterleukin-6receptorblockadetocurbthesCRS.

Additionally,wefoundthatrumC-reactiveprotein,areadilyavailablelaboratorystudy,canrveasareliable

er,ourdataprovidestrongsupportforconductingamulticenterpha

2studytofurtherevaluate19-28zCARTcellsinB-ALLandaroadmapforpatientmanagementatcentersnowcon-

templatingtheuofCARTcelltherapy.

INTRODUCTION

Tcelltherapywithtumor-targetedchimericantigenreceptor(CAR)–

modifiedTcellshasrecentlytransitionedfromthelaboratorytothe

clinicandyieldedoutcomesthatsupportthetremendouspotentialof

thisapproachtocancertherapy(1–3).CARsareartificialreceptorsthat

redirectantigenspecificity,activateTcells,andfurtherenhanceTcell

functionthroughtheircostimulatorycomponent(4,5).Threegroups,

includingourown,havereportedobjectivetumorresponswhenin-

fusingautologousTcellsgeneticallymodifiedwithCD19-targeted

CARsintopatientswithchroniclymphocyticleukemia(CLL)andother

indolentnon-Hodgkin’slymphomas(3,6,7).Wenextdemonstrated

potentantitumorbenefitafterinfusingCD19-targeted19-28zCART

cellsintofiveadultswithrelapdorrefractoryBcellacutelymphoblas-

ticleukemia(B-ALL)(1).Inadults,relapdB-ALLhasamarkedly

poorprognosiswithanexpectedmediansurvivaloflessthan6months

(8,9).Inthisttingofhighlychemotherapy-resistant,rapidlyprogres-

sivedia,therapywithCD19-targetedCARTcellsresultedin

completemolecularremissions(CRm),asassdbyimmunoglobulin

heavychain(IgH)deepquencing,infiveoffivetreatedpatients.

AchievingCRminthischemotherapy-refractorypopulationallowed

forsubquentallogeneicstemcelltransplants(allo-SCT)inclinically

eligiblesubjects,thestandardofcareinadultsforthisdiaafterre-

lap(8).Thepromisingclinicaloutcomeswereconfirmedbyinves-

tigatorsfromtheChildren’sHospitalofPennsylvaniainacareportof

twopediatricpatientswithrelapdB-ALLtreatedwithasimilarCD19

CARTcelltherapy(2).Wehavenowtreatedanadditional11patients

nicaloutcomesinthe

CD19-targetedCARTcell–treatedpatientsconfirmtheclinicalefficacy

ofthisapproachenwithourinitialresults;19-28zCARTcellsin-

ducedcompleteremissions(CRs)inthevastmajorityofpatients,

enablingmanytotransitiontoanallo-SCT.

InfusionofCD19CARTcellscanbeassociatedwithtoxicities

includinghigh-gradefevers,hypotension,hypoxia,andneurologic

disturbancesthatmayrequireaggressivemedicalsupport(1–3).This

syndromeoftoxicitieshasbeendescribedasacytokinereleasyn-

drome(CRS)likelyrelatedtoaprogressivesystemicinflammatorypro-

cessinitiatedandmaintainedbytheinfudCARTcellsactivatedin

r,the

clinicalandlaboratoryevaluationofthissyndromehasbeenlimited

todataderivedfromonlyafewpatientsincareports(1–3).Thepau-

cityofpublishedresultsfromwhichtodefineorunderstandtheCRS

markedlylimitstheclinicalinvestigator’sabilitytoeitherpredictthe

likelihoodoranticipatetheverityofthisassociatedspectrumof

CARTcell–mediatedtoxicities.

Byanalyzingall16adultswithrelapdorrefractoryB-ALLtreated

atourcenter,wehaveestablishedlaboratoryandclinicalcriteriaforthe

diagnosisoftheCARTcell–relatedvereCRS(sCRS).Usingthe

criteria,weestablishedguidelinesforinfusionofCARTcellsandthe

subquentclinicalmanagement,partofwhichincludestherial

monitoringofC-reactiveprotein(CRP).Wehavefoundthatdaily

1DepartmentofMedicine,MemorialSloan-KetteringCancerCenter,NewYork,NY10065,

USA.2CenterforCellEngineering,MemorialSloan-KetteringCancerCenter,NewYork,NY

10065,USA.3MolecularPharmacologyandChemistryProgram,MemorialSloan-Kettering

CancerCenter,NewYork,NY10065,USA.4CellTherapyandCellEngineeringFacility,

MemorialSloan-KetteringCancerCenter,NewYork,NY10065,USA.5Departmentof

Pediatrics,MemorialSloan-KetteringCancerCenter,NewYork,NY10065,USA.6Department

ofPathology,MemorialSloan-KetteringCancerCenter,NewYork,NY10065,USA.7De-

partmentofEpidemiologyandBiostatistics,MemorialSloan-KetteringCancerCenter,New

York,NY10065,USA.8LeukemiaService,NewYork-Presbyterian/WeillCornell,NewYork,NY

10065,USA.9LeukemiaService,NewYork-Presbyterian/Columbia,NewYork,NY10032,USA.

*Correspondingauthor.E-mail:brentjer@(R.B.);m-sadelain@(M.S.)

RESEARCHARTICLE

19February2014Vol6Issue224224ra251

o

n

F

e

b

r

u

a

r

y

2

1

,

2

0

1

4

s

t

m.

s

ci

e

n

c

e

m

a

g.

o

r

g

D

o

w

nl

o

a

d

e

d

f

r

o

m

monitoringofCRPincombinationwithsimpleclinicalparameters

allowsustoidentifypatientsinneedofintensivemedicalmonitoring

odifiedguidelines

willbeufulastheCARtechnology,developedandcurrentlyudin

onlyafewspecializedcenters,isadaptedtoalargernumberofmed-

asisofthe

remarkablyrobustclinicalresultsandourtoxicitymanagemental-

gorithm,wewillsoonopenamulticenterpha2clinicaltrialtofurther

evaluatetheefficacyof19-28zCARTcellsandprospectivelyvalidate

ourpropodCRSmonitoringandinterventionguidelinesinpatients

withB-ALL.

RESULTS

Clinicaltrial

Wehavetreated16patientsonour19-28zCARTcellpha1trial(1).

ThistrialisopentoadultswithB-ALLeitherinCR1orwithrelapdor

refractorydia;however,patientsaretreatedwith19-28zTcellsonly

tientstreatedtodatehavebeen

enrolledundertherelapdarm(fig.S1).Enrolledpatientsundergoleu-

kapheresis,andthowithrelapdorrefractoryB-ALLreceive“phy-

sician’schoice”followed,regardlessof

diarespon,bycyclophosphamideconditioningchemotherapy

ianageofourtreatedpa-

tientsis50years(Table1).Additionalpoor-riskfactorsinourtreatment

cohortincludePhiladelphiachromosome–positive(Ph+)dia(n=4)

aswellasrelapafterallo-SCT(n=4),rulingoutlective“goodrisk”

patientenrollmentonthistrialasapotentialfactortoconfoundclinical

tentwithchemotherapy-resistantdiainthepa-

tientsisthehighrateofresidualdiaaftersalvagetherapyandbefore

thetimeofTcellinfusion(88%,Table1).

CARTcellproductsweresuccessfullygeneratedatthedoof3×

106CARTcells/kgin15of16patientsdespitelowTcellnumbers(as

lowas3.7%)intheleukapheresisproductoftheheavilypretreated

patients(tableS1).MSK-ALL09onlyreceived16%oftheprescribed

Tcelldo,-

temptsshowedlowgenetransferefficiencyandpoorTcellexpansion,

possiblyduetothequalityofthestartingTcellproductbecaudo

asno

otherpatientenrolledonthistrialthatdidnothaveanadequatedo

production,andthisdowasnotarequirementfor19-28zCARTcell

treatment.g-Retroviral19-28zCARgenetransferwasoverallrobust,

with5to60%(mean,24%)19-28zCARexpressioninend-of-production

Tcells.

Clinicaloutcomes

Infusionof19-28zCARTcellsaftersalvagechemotherapymarkedly

enhancedtheoverallcompleteresponrate,compodofbothpatients

withaCRandaCRwithincompletecountrecovery(CRi),to88%.This

isahigherCRratethanthatexpectedwithsalvagechemotherapyalone

[Table2and(8–10)].After19-28zCARTcellinfusion,theoverallCR

ratewas78%intheninepatientswithgrossmorphologicresidualleu-

ranalysofCRstatusinclud-

edstudiestodetectminimalresidualdia(MRD)byflowcytometry,

quantitativepolymerachainreaction(qPCR)forthebcr-abl

transcriptinpatientswithPh+B-ALL,and,wheneverfeasible,deep-

quencingforIgHrearrangements(11)associatedwithmalignantclones

(Table2).Overall,75%oftreatedpatientsachievedanMRD-negative

(MRD−)orCRmdiastatusbadononeormoreoftheaboveMRD

assays(Table2).TheCRandCRmrates,obtainedinaverypoor

,cen-

tralnervoussystem.

ents(N=16)%

Sex

Male1275

Female425

Age(years)

Median50

Range

18–29425

30–59744

≥60531

BalineBMcytogenetics

Unfavorable744

Ph+425

Intermediate956

Previousallo-SCT

Yes425

No1275

Extramedullarydia

CNS212

Other16

None1381

DurationofCR1(months)

Median8

Range

<6531

6–24744

>24425

Numberofsalvageregimens

1956

2425

≥3319

Refractorytoimmediateprevioustherapy

Yes1488

No212

B-ALLtumorburdenintheBMbeforeCARTcellinfusion(n=15)*

MRD

−

213

MRD+533

<50%blasts213

≥50%640

*Onepatienthadonlygrosxtramedullarydia(nodetectablediaintheBM).

RESEARCHARTICLE

19February2014Vol6Issue224224ra252

o

n

F

e

b

r

u

a

r

y

2

1

,

2

0

1

4

s

t

m.

s

ci

e

n

c

e

m

a

g.

o

r

g

D

o

w

nl

o

a

d

e

d

f

r

o

m

prognosticpatientpopulation,farexceedexpectationsbadonhistor-

icaldataofrelapdadultB-ALL(8–10)andareconsistentwithapro-

foundantitumoreffectmediatedby19-28zCARTcells(Table2).

Furthermore,inpatientswhereinthemalignanttumorclonecouldbe

monitoredinbonemarrow(BM)bydeepquencing,wefoundrapid

eliminationofthemalignantB-ALLtumorcloneafter19-28zCART

cellinfusion(tableS2).Concomitantmonitoringfor19-28zCARTcell

persistenceintheBMrevealedthatnearlyallpatientshadapeakof

CARTcellswithin1to2weeksaftertheinfusionandthatthe

numbersdecreadtoloworundetectableby2to3months(table

S2).AnalysisforCARTcellpersistencewaslimitedinpatientssub-

herobrvedthatinthefourpa-

tientstreatedafterapost–allo-SCTrelap,therewasnoclinical

evidenceofgraft-versus-hostdiadespitethefactthattheinfud

19-28zCARTcellswereofdonororigin.

TheCARTcell–associatedCRS

CAR-engineeredTcellscaninduceinsomepatientsaclinicalsyndrome

offevers,hypotension,hypoxia,andneurologicchangesassociatedwith

markedelevationsofrumcytokines(1–3).Thisspectrumofclinical

andlaboratoryfindingshasbeentermedaCRS,which,giventhe

anecdotalnatureofthisphenomenon,hasremainedlargelyundefined.

Wethereforeanalyzedourcohorttoarchforclinicalorlaboratory

resultsthatmightrveasdiagnosticindicatorsforaclinicallymeaning-

fulorvereCRS,predictablyrequiringadditionaltherapeuticinterven-

end,weidentifiedatofcriteriaforthediagnosisofan

sCRSbadontheprenceoffevers,elevationofcharacteristiccyto-

kines,andclinicaltoxicities(Table3,tableS3,andFig.1).Patientswith

evidenceofCRStypicallyhavefeversthatstartabout24hoursafterin-

fusionwith19-28zCARTcellsandcanpersistforveraldays(Fig.1A).

Feversare,however,notalwaystheharbingerformoreclinicallyrelevant

eforeevaluatedcytokineincreastodiscernbetween

patientswithsCRSassociatedwithclinicaldeteriorationandpatients

whofeversanddiscomfortwouldresolvespontaneouslyorwith

ortanceofthisdistinctionistoavoidpre-

matureinterventionthatmaydiminishTcellpersistenceorefficacy.

WehavepreviouslycorrelatedcytokinelevelstopretreatmentB-ALL

tumorburden(1),albeitinasmallsamplesize(n=5),whichprecluded

argercohortofpatients,wenotonlycon-

firmedthiscorrelationbutalsoidentified7cytokinesof39measured,

whoelevationcorrelated(r=0.43to0.88)topretreatmenttumorbur-

den(Fig.1B)andalsotoansCRS(tableS4).Withinthispanelofven

cytokines,weobrvedthatpatientswithCRSrequiringintensivemedical

interventionhada75-foldincreaoverpretreatmentbalinelevelsin

twoofthelectedvencytokines(Table3).Furthermore,thopatients

withsCRSuniversallyexhibitedatleastoneofthefollowingclinicalman-

ifestations:hypoxia,hypotension,and/,on

thebasisofthecombinedclinicalandcytokinedata,wecouldaccurately

defineansCRSinthopatientswiththetriadofpersistentfevers(38°C)

formorethan3days,lectedcytokineelevations,andadditionalclin-

ationofthecriteriaenablesstratifica-

tionofpatientsintothesCRSgroup,whichrequiresclorobrvation

andislikelytorequiremedicalandpharmacologicintervention,and

anothergroup(nCRS)madeupofpatientswhotoleratetherapyandonly

tternCRScohort

includespatientswithamildCRS,characterizedbylow-gradefever

andmildcytokineincreas,orabntCRS,definedasnofeversand/or

nosignificantcytokineelevations(Fig.1).Thisisaclinicallymeaningful

stratificationbecausCRSpatientsareinthehospitalforanaverageof

56.7days(SD,28.6;range,20to104),whereasnCRSpatientsareinthe

hospitalforanaverageof15.1days(SD,18.8;range,4to61).

ManagementoftheCRS

CRS-associatedtoxicities,whenvere,requireintensivemedicalman-

agementincludingsupportwithvasoactivepressors,mechanicalven-

tilation,antiepileptics,r,althoughthe

yofclinicaloutcomes.

ents(N=16)%

Overallcompleterespontosalvagechemotherapy*744

Overallcompleteresponto19-28zCARTcells14

†

88

Inpatientswithmorphologicresidualleukemia(n=9)778

Completeremission(CR)1063

Completeremissionwithincompletecountrecovery(CRi)425

Molecularcompleteremission(CRm)

‡

12

†

75

MediantimetoCR/CRi(days)24.5

Post-CARTallo-SCT(n=10eligiblepatients)§770

*Overallcompleterespon=CR+CRi(determinedwithoutregardtoCRmstatus).†IncludestwopatientswhowereinCRmbeforeCARTcellinfusion.‡CRmorMRD

−

asdetermined

byflowcytometryand/ordeepquencingfortheindexIgHclonotypeand/orqPCRforthebcr-abltranscript.§Threepatientshadmedicalcontraindicationtoallo-SCT,twopatientsinCR

havedeclinedpotentialallo-SCT,andoneisbeingevaluatedforanallo-SCT.

sticcriteriaforsCRScondarytoCARTcells.

CriteriaforsCRS

Feversforatleastthreeconcutivedays

Twocytokinemaxfoldchangesofatleast75oronecytokinemaxfold

changeofatleast250

Atleastoneclinicalsignoftoxicitysuchashypotension(requiringatleast

oneintravenousvasoactivepressor)or,

Hypoxia(PO

2

<90%)or,

Neurologicdisorders(includingmentalstatuschanges,obtundation,

andizures)

RESEARCHARTICLE

19February2014Vol6Issue224224ra253

o

n

F

e

b

r

u

a

r

y

2

1

,

2

0

1

4

s

t

m.

s

ci

e

n

c

e

m

a

g.

o

r

g

D

o

w

nl

o

a

d

e

d

f

r

o

m

toxicitiesareconcerning,theyareaby-productof19-28zCARTcell

functionand,todate,tedourinitial

threesCRSpatientswithlymphotoxichigh-dosteroids,>100mgdaily

ofprednisoneequivalent,whichrapidlyreverdsymptomsbutconcur-

rentlyablated19-28zCARTcells(Fig.2).Theinterleukin-6receptor

(IL-6R)–blockingmonoclonalantibody(mAb)tocilizumabmayalso

amelioratesCRS,asinitiallyreportedbyGruppetal.(2)whodemon-

efore

treatedournextthreesCRSpatients(prenting27-to400-foldincreas

inrumIL-6)withtocilizumabalone(Fig.2),whichreducedpatients’

feversandsCRSsymptomswithin1to3dayssimilartosteroidtherapy,

butdidnotresultindampenedexpansionofthe19-28zCARTcells

rresultswerenotedinthe

BMbydeepquencing(Fig.3).Inaggregate,wedetectedafivefold

decreain19-28zCARTcellsintheBMofsCRSpatientstreatedwith

high-dosteroidsrelativetosCRSpatientstreatedeitherconrvatively

orwithtocilizumabalone(Fig.3).

SuppressionofCARTcellexpansionpresumablyhasanegative

,deepquencingfortheIgH

rearrangementassociatedwiththemalignantB-ALLclonerevealedthat

thethreesCRSpatientstreatedwithhigh-dosteroidsallexperienceda

recurrenceofdiadespiteinitiallyachievingaCRmafter19-28zCAR

Tcellinfusion(tableS2).Unfortunately,twoofthepatientsdidnot

undergotherecommendedallo-SCTwhileMRD–,becauofeither

medicalcontraindications(MSK-ALL04)orhavingdeclinedfurther

therapy(MSK-ALL07),-ALL05hadavery

lowlevelofdetectablerecurrentdiaintheBMbydeepquencing

andachievedaCRmafterallo-SCT(tableS2).

CARTcell–mediatedsCRS-associatedneurologictoxicities

PatientswithsCRSmayalsodevelopreversibleneurologiccomplica-

tsmaydevelop

agradualprogressionofconfusion,word-findingdifficulty,andaphasia

ecas,theneurologiccom-

plicationsrequiredintubationandmechanicalventilationforairway

protection(tableS4).Patientswithneurologiccomplicationswereeval-

uatedwithcomputedtomographyandmagneticresonanceimagingof

thebrain,whichwasnonrevealing,aswellalectroencephalograms

(EEGs)sconfirmedizure-likeactiv-

ity,isofcerebro-

spinalfluid(CSF)obtainedbylumbarpunctureinthreepatientsat

thetimeofovertneurologiccomplicationsrevealedalymphocytosis,

which,byfurtherqPCRanalys,wasfoundtobecompodof,atleast

inpart,19-28zCARTcells(tableS5).Oneofthepatients(MSK-

ALL14)hadaprevioushistoryofCNSdia,whichhadresolvedat

ertwopatientsdidnothaveany

ghCSFsam-

pleswereobtainedonlyinasubtofpatientsandonlyinthettingof

clinicalneurologiccomplications,19-28zCARTcellswerenotdetected

m-

ple,CSFfromMSK-ALL16,obtainedatatimeoffeversanddelirium,

showednoevidenceofCARTcellsbydirectmicroscopicexamination

ormorensitiveqPCR(tableS5).

CRPasanindicatoroftheCRS

Wehaveobrvedthatpatientswithmorphologicdiaatthetimeof

19-28zCARTcellinfusionhaveagreaterriskfordevelopingsCRS

(Figs.1and2andtableS4).Unfortunately,rapidanddailyreal-time

teristicsoftheCRS.(A)Averagemaxtemperatureson

days1to11afterCARTcellinfusioninpatientswithsCRScomparedto

hedlineisat38°Cto

-wayanalysisofvariance(ANOVA)

analysisbetweenthesCRSandnCRSgroupsrevealedaP=0.019(n=

22).(B)Maxfoldchangesofveninflammatorycytokineslectedfor

edarethemax

foldchangesrelativetopretreatmentvaluesoverdays1to21afterCAR

hlightedboxreprentschanges75-foldand

ationwasassdforpretreatmenttumorburdenandcy-

armanrank

correlationcoefficientwascalculatedwithpretreatmenttumorburden,

measuredbydeepquencing,andcytokineconcentration(pg/ml),and

-g,interferon-g;GM-CSF,granulocyte-

macrophagecolony-stimulatingfactor.

RESEARCHARTICLE

19February2014Vol6Issue224224ra254

o

n

F

e

b

r

u

a

r

y

2

1

,

2

0

1

4

s

t

m.

s

ci

e

n

c

e

m

a

g.

o

r

g

D

o

w

nl

o

a

d

e

d

f

r

o

m

analysisofrumcytokines,whichcouldguideclinicaldecision-making

beforetheontofvereCARTcell–associatedtoxicities,isnotfeasible

becauoftechnologicallimitationswithcytokinemeasurements.

Therefore,wesoughtalaboratoryindicatorforCRSveritythatcould

dontheacute-

phareactant,CRP,badonthewell-documentedassociationbe-

tweenrumIL-6andCRPlevels(12)andtheclinicalamelioration

ofthesCRSaffordedbyIL-6Rblockade[Fig.2and(2)].

Weretrospectivelyanalyzedrumsamplesfromallpatientstreated

onthistrialanddeterminedthatonlythopatientswhometthecrite-

riaforsCRShadaCRPlevelof≥20mg/rvedaclear

differencebetweentheCRPlevelsofpatientswithansCRSversuspa-

tientsclassifiedashavingeithermildornoCRS(Fig.4andfig.S2).Pa-

tientstreatedwithhigh-dosteroidswereexcludedfromthisanalysis

giventheinvercorrelationbetweenhigh-dosteroidtreatmentand

rmore,receiveroperatingcharacteristic(ROC)

curveanalysissuggestsCRPasanexcel-

lentindicatorforsCRS(fig.S3).Maxi-

mumvalueoftheCRPbeforesCRShas

-

rvedthatpatientswhoCRPexceeds

thethresholdareparticularlyathighrisk

forCRS(nsitivity,86%;specificity,100%).

CARTcellsasabridgetoallo-SCT

Asperthecurrentstandardofcarefor

adultswithrelapdorrefractoryB-ALL,

theinitialprimaryaimoftherapyistore-

induceaCR(8–10).This,inturn,renders

thepatienteligibleforanallo-SCT,which

is,atprent,theonlytherapeuticmodal-

16pa-

tientstreatedonthisprotocol,3were

ineligibleforallo-SCTbecauofafailure

toachieveaCRdespite19-28zCARTcell

infusion,3patientsinCRwereineligible

becauofmedicalcomorbiditiesthatpre-

existed19-28zCARTcelltherapy,and2

patientsinCRwereeligibleforallo-SCT

butdeclinedfurthertherapy(Tables2

and4).Onepatientiscurrentlybeing

date,7ofthe16(44%)treatedpatients

havesuccessfullyundergoneanallo-SCT

post-CARTcelltherapywithnorelaps.

DISCUSSION

Ourresultsstronglysupportthetherapeu-

ticpotentialforourfirst-in-classCD19-

gh

theresultswereobtainedinasingle-center

pha1study,theysupportfurtherevalu-

ationof19-28zCARTcelltherapyforthis

verypoorprognosispopulationinamul-

ticenterpha2clinicaltrial.

PatientswithrelapdB-ALLhavefew

treatmentoptionsandahistoricalremis-

sionratewith“standard-of-care”salvage

chemotherapyofabout30%(8–10).Con-

sistentwiththisoverallpoor-riskprog-

nosis,nearlyallofourpatients(88%)were

refractorytothephysician’schoicesalvage

therapygivenbeforeCARTcellinfusion

(Table1).Incontrast,patientstreatedwith

P

r

e

-

R

x

2

d

4

d

6

d

8

d

1

0

d

1

2

d

1

4

d

1

6

d

1

8

d

2

0

d

0

50

100

150

34

36

38

40

42

Timepoint

MSK-ALL05

d6d19

P

r

e

-

R

x

2

d

4

d

6

d

8

d

1

0

d

1

2

d

1

4

d

1

6

d

1

8

d

2

0

d

2

2

d

0

50

100

150

34

36

38

40

42

Timepoint

#C

A

R

T

c

e

l

l

s

/

µl

C

A

R

T

c

e

l

l

s

/

µl

MSK-ALL14

d6d14

P

r

e

-

R

x

2

d

4

d

6

d

8

d

1

0

d

1

2

d

1

4

d

1

6

d

1

8

d

2

0

d

2

2

d

0

50

100

150

34

36

38

40

42

Timepoint

MSK-ALL07

d10d9d16

P

r

e

-

R

x

2

d

4

d

6

d

8

d

1

0

d

1

1

d

1

2

d

1

4

d

1

6

d

1

8

d

2

0

d

2

2

d

0

50

100

150

1000

3000

34

36

38

40

42

Timepoint

MSK-ALL13

d6

d9

P

r

e

-

R

x

2

d

4

d

6

d

8

d

1

0

d

1

2

d

1

4

d

1

6

d

1

8

d

2

0

d

2

2

d

0

50

100

150

34

36

38

40

42

Timepoint

#C

A

R

T

c

e

l

l

s

/

µl

MSK-ALL04

d8d6d27

P

r

e

-

R

x

2

d

4

d

6

d

8

d

1

0

d

1

2

d

1

4

d

1

6

d

1

8

d

2

0

d

2

2

d

0

50

100

150

34

36

38

40

42

Timepoint

MSK-ALL17

d10

Steroidsadministered

Tocilizumabadministered

T

max

CARTcells/µl

Temperature(

ºC)

Temperature(

ºC)

Temperature(

ºC)

ectofsteroidsand/ortocilizumabontheexpansionofCARTcellsinpatientswith

berofCARTcellspermicroliterofwholeblood,detectedbyqPCR,wasmeasuredinsamples

peraturesondays1to11are

tion,thedayswhensteroidsortocilizumabwasadministeredtomanagesCRSare

dashedlinereprentsthedurationofsteroidtreatment,andthegraydashedlineisat

the38°Cfeverthreshold.

RESEARCHARTICLE

19February2014Vol6Issue224224ra255

o

n

F

e

b

r

u

a

r

y

2

1

,

2

0

1

4

s

t

m.

s

ci

e

n

c

e

m

a

g.

o

r

g

D

o

w

nl

o

a

d

e

d

f

r

o

m

19-28zCD19CARTcellshadaveryhighoverallcompleterespon

rate(88%),with86%ofthepatientsfromthisCRgroupfurtherclas-

sified(12of14patients)asMRD−(CRm)(Tables2and4).Subjectswith

MRDorovertmorphologicresidualleukemiaaftersalvagechemo-

rvedsimilarlyhighCRrates

inbothgroupsafter19-28zCARTcellinfusion(Table2).

Wewereabletosuccessfullytransitionvenpatients(44%ofall

patients)tostandard-of-caretherapywithanallo-SCT(Tables2and

4).Thisispeciallymeaningfulwhencomparedtothereportedhistor-

icallylowfrequency(5%)ofrelapdorrefractoryadultB-ALLpatients

whoultimatelytransitiontoallo-SCTaftersalvagechemotherapy(13).

Thus,19-28zCARTcelltherapymayreprentaneffective“bridge”to

emostofourpatientsunderwentallo-SCTinthe

ttingofaCRm,wehypothesizethattransplantsperformedunder

theoptimalconditionswillmarkedly,ifnotcompletely,reducethe

historical30%diarelaprateofB-ALLpatientsafterallo-SCT

(14,15).Tothind,therehavebeennorelapsinthevenpatients

treatedwithanallo-SCTafter19-28zCARTcelltherapy(post–allo-

SCTfollow-uprangesfrom2to24months),althoughtwoofthepa-

ients

whodidnottransitiontoanallo-SCTafterTcelltherapydidsofora

varietyofreasons,includingsuboptimalrespontoCARTcelltherapy

(n=3),decliningallo-SCTdespiteachievingCRmafter19-28zCART

celltherapy(n=2),andpreexistingmedicalcontraindicationstoan

allo-SCTinpatientswithaCRorCRmafter19-28zCARTcelltherapy

(n=3).Onerecentlytreatedpatientispendinganevaluationforanallo-

,nopatientwasprecludedfromallo-SCTbe-

cauoftoxicitiesassociatedwith19-28zCARTcelltherapy.

Thedesignofthispha1clinicaltrialstipulatedthatafterenroll-

mentandleukapheresis,patientsweregivenreinductionsalvage

chemotherapyandlaterinfudwithautologous19-28zCARTcells

(fig.S1).Thelowefficacyandprolongeddurationofmyelosuppression,

aswellasnumerousothertoxicsideeffects,associatedwithsalvage

chemotherapyresultinmanypatientshavingmorbidityand/ormor-

tality,whichprecludesfurthertreatmentandmayevendisqualifysome

patientsfromanallo-SCT(8,9,15).Incontrasttoourinitialexpecta-

tionsthatsalvagechemotherapywouldenhanceCARTcellantitumor

efficacy,weobrvedsimilarclinicaloutcomesinpatientswhoachieved

aCRaftersalvagetherapyaswellasthopatientswhodidnot(Tables2

and4).Consideringthetoxicitiesassociatedwithsalvagereinduction

chemotherapy,andtheoverallhighcompleteresponratesto19-

28zCARTcelltherapy,onemayquestiontheutilityorwisdomofgiv-

ingpatientstoxichigh-dochemotherapiesbefore19-28zCARTcell

infusions.

Consistentwithourpreviousreports(1,6),thepersistenceofthe19-28z

CARTcellsinALLpatientsisabout3months(tableS2).Thisisincontrast

toatleastoneB-ALLpediatricpatientandveralCLLpatientsreportedby

theUniversityofPennsylvania(2,7,16),whoexhibitedCARTcellpersist-

enceandpersistingBcellaplasiaforveralmonthstoevenmorethana

thesizethatthe19-28zCARTcellexpansionandsubquent

contractionareCD19antigen–dependent,resultinginTcellclearance

uponeliminationofnormalandmalignantandBcells(1,6),asenin

ingly,thepersistence

ofCD19-targetedCARsincorporatinga4-1BBmoietyratherthanCD28

asudbytheUniversityofPennsylvaniamaybedue,atleastinpart,to

antigen-independentsignalingthroughthe4-1BBCAR,aspreviously

demonstratedinpreclinicalstudies(17).Additionaloralternativemecha-

urrentlydevelopingahumananti-mouanti-

bodyassaytodeterminewhetherimmune-mediatedrejectionmightbea

symptomsappearearlierwithCD28-containingCD19-targetedCART

cells,reportedbyboththeNationalCancerInstitute(NCI)andMemorial

n

C

R

S

s

C

R

S

(

s

t

e

r

oi

d

s

)

s

C

R

S

(

n

o

s

t

e

r

oi

d

s

)

0

2000

4000

6000

8000

10,000

1

9

-

2

8

z

C

A

R

T

c

e

l

l

s

p

e

r

1

0

B

M

c

e

l

l

s

*

isolatedfrompatientsandsubmittedtoAdaptiveBiotechnologiesfordeep

maxnumberofCARTcellsintheBMwithin6weeksofCARTcellinfusionis

nandSDaredepictedforthepatientgroupsstratifiedon

thebasisofCRSanditsmanagement.*P=0.048,one-wayttestbetween

thetwogroups(n=6).

P

r

e

-

R

x

0

1

2

3

4

5

67

8

9

1

0

1

1

1

2

1

3

1

4

1

5

1

6

1

7

1

8

0

10

20

30

40

Daypost-infusion

C

R

P

(

m

g

/

dl

)

nCRS

sCRS

****

measuredbeforetreatmentandfromdays1to18afterCARTcellinfusion.

ThegreenlinesreprentCRPlevelsfrompatientswhometthediagnostic

criteriaforsCRS(n=4)andweretreatedwitheithertocilizumabornothing.

ars

ydashedlineisat20mg/dl,whichindicatesthethresh-

oldwherepatientsareathighriskforclinicalcomplicationscondaryto

sCRS.*P<0.05,ic

Pvaluesforthetimepointsareasfollows:day2,P=0.035(n=13);day4,

P=0.025(n=12);day5,P=0.019(n=11);andday9,P=0.01(n=8).

RESEARCHARTICLE

19February2014Vol6Issue224224ra256

o

n

F

e

b

r

u

a

r

y

2

1

,

2

0

1

4

s

t

m.

s

ci

e

n

c

e

m

a

g.

o

r

g

D

o

w

nl

o

a

d

e

d

f

r

o

m

Sloan-KetteringCancerCenter(MSKCC),comparedto4-1BB–containing

evethatthisisduetoamorerapid

Tcellexpansionintheformergroupthaninthelatter,onthebasisofthis

currentreportandpreviouslypublishedobrvations(1–3,6,7,16,18).

Althoughtheoverallcompleteresponratewasdramatic,there

werealsoexamplesoffailuretoreinduceaCRinpatientswithmorpho-

logicresidualdiaorfailuretoinduceaCRminpatientswithMRD.

TwosuchtreatmentfailuresoccurredinpatientswithMRDatthetime

ofCARTcellinfusion,whereastheothertwopatientshadovertmor-

-

tient(MSK-ALL09)receivedalow19-28zTcelldo(16%,tableS1)

withoutsubquentevidenceofpost-infusioninvivoCARTcellexpan-

sion(tableS2).BothMRDtreatmentfailureshadlowtonodetectable

19-28zCARTcellsintheBMafterinfusion,incontrasttothopa-

tientswithtreatmentrespons,pointingtolimitedTcellexpansionin

thepatientsasonemechanismcontributingtotreatmentfailure(Fig.

3andtableS2).ThelackofresponinMSK-ALL08,whohaddia

involvementonlywithinalargeabdominallymphnodemass,maybe

duetolimitedTcelltraffickingorimmunosuppressionofCARTcells

withinthixtramedullarytumormicroenvironment.

Thetoxicitiesassociatedwiththeinfusionof19-28zCARTcellsare

thetoxicitiesassociatedwithconventional

salvagechemotherapy,thoassociatedwithinfudCARTcellsarere-

latedtolarge-scale,synchronousTcellactivationupontargetingof

CD19+aticrumcytokineanalysallowedus

tolectasmallpanelofcytokinesthatarestronglyassociatedwithan

sCRS(Fig.1B).Identificationofthevencytokines,commonly

elevatedwiththesCRS,allowedustodeveloplaboratoryandclinical

criteriafortheformaldiagnosisofansCRS(Table3).Onthebasisof

ouranalysis,patientswithfeversaloneand/orelevatedrumcytokines

intheabnceofadditionalclinicallyapparenttoxicitiesareunlikelyto

requireanythingmorethanobrvationormodestmedicalinterven-

,deepquencingforIgH

rearrangement;DUCBT,doubleumbilicalcordbloodtransplant;FC,flow

cytometry;HUCT,haplo-umbilicalcordtransplant;MTX,methotrexate;

Peg,pegasparigina;Pred,prednisone;qPCR,quantitativePCRfor

bcr-abltranscript;RD,relateddonor;TCD,Tcell–depleted;UD,unrelated

donor;Vinc,vincristine;NA,notavailable.

PatientIDAge

Cytogenetics

atdiagnosis

Salvagetherapy

Diaresponto

salvagetherapy

Diaresponto

CARTcells

Allo-SCT

MSK-ALL0166NormalkaryotypeVinc/Pred/PegMRD+byDSMRD

−

byDS10/10TCDRDat64dayspost

MSK-ALL0356NormalkaryotypeVinc/Pred/PegMRD

−

byFCMRD

−

byFC10/10TCDMRDat43dayspost

MSK-ALL0459t(9;11),9p21deletionVinc/PredRefractorydia,63%

blastsinBM

MRD

−

byDSIneligiblebecauofmedical

contraindications

MSK-ALL05589p21deletionHigh-docytarabine/

mitoxantrone

Refractorydia,70%

blastsinBM

MRD

−

byDSTCDDUCBTat69dayspost

MSK-ALL0623NormalkaryotypeModifiedNYIIMRD+MRD

−

byDS8/10TCDUDat121dayspost

ConsolidationI(27)

MSK-ALL07309qisochrome,12p13deletionVinc/Pred/PegRefractorydia,5–10%

blastsinBM

MRD

−

byDSDeclined

MSK-ALL0874Complexincluding

11q23deletion

MTXBM-negative,

+extramedullarydia

NoresponNorespon

MSK-ALL0923NAModifiedNYIIMRD+byFCMRD

−

byFC,MRD+

byDS

Ineligiblebecauofmedical

contraindications

ConsolidationI

MSK-ALL1027NormalModifiedNYIIMRD

−

byFCMRD

−

byFCIneligiblebecauofmedical

contraindications

ConsolidationI

MSK-ALL1132Ph+Vinc/PegMRD+byqPCRMRD

−

byqPCR10/10UDat~90dayspost

MSK-ALL12*42Ph+ClofarabineRefractorydia,97%

blastsinBM

NoresponNorespon

MSK-ALL13*36Ph+InotuzumabRefractorydia,60%

blastsinBM

MRD

−

byDSandqPCRDeclined

MSKALL1460NAVinc/Pred/PegRefractorydia,52%

blastsinBM

MRD

−

byFCHUCTat~60dayspost

MSKALL15*27t(2;12),monosomy7L20(28)Refractorydia,23%

blastsinBM

MRD

−

byDS10/10UDat49dayspost

MSK-ALL16*63Ph+,11qdeletionPOMP(29)MRD+byqPCRMRD+byqPCRNorespon

MSK-ALL1759ComplexVinc/PredRefractorydia,85%

blastsinBM

MRD

−

byFCAwaitingallo-SCTevaluation

*Enrolledandtreatedafteranallo-SCT.

RESEARCHARTICLE

19February2014Vol6Issue224224ra257

o

n

F

e

b

r

u

a

r

y

2

1

,

2

0

1

4

s

t

m.

s

ci

e

n

c

e

m

a

g.

o

r

g

D

o

w

nl

o

a

d

e

d

f

r

o

m

rast,patientswhomeetthesCRScriteriaarelikelytorequire

diagnosticcriteriaweestablishedwillbeufultonormalizeevaluation

ofthetoxicitiesacrossmultipletrialsatdifferentmedicalcenters,for

developingpreclinicalmodelstounderstandthemechanismbehindthe

CRS(19),andtofurtheroptimizetheclinicalmanagementofthissyndrome.

OurinitialattemptstomanagesCRShaveincludedtreatingpatients

withhigh-dosteroidsand/ortocilizumab,anIL-6R–blockingmAb.

WehavefoundthatthemanneroftreatingsCRSmayaffectclinicalout-

strationofhighlymphotoxicdosofsteroidsasatreat-

mentofsCRSinpatientsMSK-ALL04,MSK-ALL05,andMSK-ALL07

resultedinarapidreversaloftheirfevers,cytokines,andotherclinical

symptomsbutabrogated19-28zCARTcellexpansionandpersistence

(Fig.2).Incontrast,administrationoftocilizumabasafirst-linetherapy

forsCRSinpatientsMSK-ALL13,MSK-ALL14,andMSK-ALL17simi-

larlyreducedfeversandamelioratedclinicalsymptomswithoutappar-

enteffecton19-28zCARTcellexpansionandpersistence(Fig.2).Two

patients,MSK-ALL13andMSK-ALL14,hadexpansionof19-28zCAR

Tcellsaftertheirfirsttreatmentwithtocilizumab,withMSK-ALL13

demonstratinganalmost7000-foldinvivoexpansionaftertreatment.

Thelymphotoxiceffectofsteroidsappearstoaffectnotonlytheinvivo

expansionoftheinfudCARTcellsbutalsotheclinicaloutcomeof

eepatientstreatedwithhigh-dosteroidsre-

lapddespitepreviouslyachievingaCRm(MRDtomorphologicre-

lap),whereasuntreatedpatientsorthotreatedwithtocilizumabalone

t

knowiflowerdosofsteroidsmightbeaffectiveatdecreasingsCRS

esultsstronglysuggestthattoci-

lizumabshouldbeudinthefirst-linetreatmentofsCRS,withhigh-do

steroidsbeingrervedforthopatientswithverelife-threateningCRS

unresponsivetotocilizumab.

We,aswellasothers(1–3),haveobrvedanumberofclinically

alarmingneurologicchangesassociatedwiththesCRS(tableS4).

Becauofsimilarpublishedneurologicchangesafterblinatumo-

mabinfusionorCD28mAbligation(20,21),which,inbothcas,

resultedinrobustTcellactivation,wespeculatethattheneuro-

logictoxicitiesarifromageneralizedTcell–mediatedinflamma-

torystateratherthandirecttoxicitymediatedby19-28zCART

,nodetectable19-28zCARTcellswere

foundintheCSFofMSK-ALL16,despitetheclinicallyevidentand

persistentdeliriumatthetimeofCSFcollection(tableS5).

Understandingthemechanismsunderlyingtheneurologiccompli-

cationsenwithCARTcelltherapyinthettingofansCRS,as

wellasmoreefficientmanagementofthetoxicities,willrequire

moreintensiveclinicalandpreclinicalinvestigation(19).Fortu-

nately,thecomplicationshavebeenmedicallymanageableand

fullyreversibleinourpatientcohort.

WehaveidentifiedCRPasapotentiallaboratoryindicatorforthe

sCRS,consideringthatcytokinemonitoringisunlikelytobeperformed

pectivereviewofpa-

tientrumCRPlevelsovertime(fig.S2)revealedthatpatientswith

sCRSwhoreceivedsteroidsortocilizumabweretreatedatorneartheir

onally,wefoundthatpatientswithsCRStreated

withsteroidsand/ortocilizumabexhibitedarapiddropinrumCRP,

consistentwithclinicalresolutionofthesCRS(fig.S2).Wethereforepro-

pothatanypatientwhohasfeversandaCRP≥20mg/dlshouldbe

managedasiftheyhavesCRSandbeconsideredathighriskforclinical

complications(Table3),aguidelinethatweplantovalidateprospectively.

PosthoccytokinemonitoringwillstillbeufultoconfirmsCRSandfor

asisofour

experience,wehavedevelopedclinicalguidelinesforthemanagement

ofpatientsbeingtreatedwithCARTcells(fig.S4).

GroupingpatientsaccordingtotheirCRSstatus,sCRS(n=7)versus

mildornoCRS(n=9),alignssignificantlywiththepretreatmentblast

burdenbefore19-28zCARTcellinfusion(P<0.05,Table4).Thus,all

venpatientswhodevelopedsCRShadmorphologicresidualleukemia

andachievedaCRm,whereastheninepatientswithmildornoCRS

includedvenpatientswithMRDandtwowithmorphologicresidual

dia,butnotreatmentrespon(Table4).Inourpreviousreport

(1),weobrvedacorrelationbetweentumorburdenandcytokineele-

vationbutnotbetweentumorburdenandoutcome,indicatingthat

treatment-associatedtoxicitywasnotrequisiteforanefficient19-28z

CARTcell–rlargercohort,wecon-

tinuetoreportnodifferencesintheclinicaloutcomesofpatientswith

MRDversusthopatientswithovertmorphologicresidualleukemia.

However,patientswithsCRS,andthereforewithmorphologicresidual

leukemiabefore19-28zCARTcellinfusion,havegreaterexpansionof

yberelatedtoabundant

CD19expressiononresidualleukemiainthepatients,incontrastto

loworabntlevelsofCD19inpatientswithMRD(22),oradampening

effectofnormalBcells,whichmaypredominateinpatientswithMRD.

Together,ourabilitytoanticipateandmanagetoxicitiesinpatientstreated

with19-28zCARTcellswillgreatlyenhancetheimplementationofmul-

ticenterpha2studies,whichthefindingsreportedhereinstrongly

support.

MATERIALSANDMETHODS

Clinicalprotocoldesign

Thisisapha1protocol(#NCT01044069)thathas

beendescribedindetail,andtheprotocolisavailableassupplemental

material[fig.S1and(1)].Briefly,itisopentoadultswithB-ALLintheir

firstCRorwithrelapd/r,onlypatientswith

relapdorrefractoryB-ALLareeligibleforinfusionwith19-28zCAR

ientsaregivenaconditioningchemotherapyagent,cy-

clophosphamide(1.5to3.0g/m2),followedbyafractionateddo

(1/

3

doonday1and2/

3

doonthefollowingday)of19-28zCAR

eunderevaluationis3×106CARTcells/tsare

treatedintheinpatientttingtomanagepotentialtoxicitiesafter19-28z

tsachievingaCRafterCARTcelltherapy

werereferredtotheMSKCCBMtransplantationrviceforevaluation

titutionalReviewBoardat

ientnrolledandtreated

n-

icalinvestigationwasconductedaccordingtotheDeclarationofHelsinki

principles.

Generationof19-28zCAR-modifiedTcells

19-28zCARTcellswereharvested,transduced,formulated,andre-

leadaspreviouslydescribed(1,6,23).

AnalysisofcytokinesandCRPafter19-28zCAR

Tcellinfusion

PatientrumsampleswereanalyzedwiththeLuminexIS100system

andcommerciallyavailable39-plexcytokinedetectionassaysasde-

RESEARCHARTICLE

19February2014Vol6Issue224224ra258

o

n

F

e

b

r

u

a

r

y

2

1

,

2

0

1

4

s

t

m.

s

ci

e

n

c

e

m

a

g.

o

r

g

D

o

w

nl

o

a

d

e

d

f

r

o

m

scribed(1,6).TheDepartmentofLaboratoryMedicineatMSKCCud

rumtomeasurehigh-nsitivityCRPwiththeSiemensHighSensitiv-

ityCRPreagentkitontheADVIA1800,alsomanufacturedbySiemens.

Molecularstudiesofwholeblood,BM,andCSF

ThemalignantIgHrearrangementwasdetectedfromadiagnosticre-

-upBMaspirateswereprocesdtoextract7.5mg

ofgenomicDNAandsubmittedfordeepquencingatAdaptiveBio-

cesforthemalignantIgHrearrangementwere

thenudtointerrogatethehigh-throughputquencingoutputto

patients,monitoringforthemalignant

IgHrearrangementwasnotpossiblebecaunorelapdsamplewas

availableornoIgHrearrangementwasdetectedbecauthepatients’

IgHlocuswasgermline.

19-28zCARTcellsweredetectedbyqPCRand/ordeepquencing

R,genomicDNAwasisolatedfromtheappropriate

tissue,andaportionofthe19-28zCARconstructwasamplifiedas

described(6).Thedeepquencingprocessisinitiatedwithamulti-

plexPCRassaythatusmultiple,degenerateV

H

andJ

H

familypri-

mers(11).Asaconquenceofthedegeneratenatureoftheprimers,

thehigh-throughputquencingoutputalsoincludedquencesfor

themouanti-CD19IgHrearrangementassociatedwiththe19-28z

ore,wewereabletomonitorfor19-28zCARTcellper-

sistenceintheBMbyinterrogatingthehigh-throughputquencing

outputwiththeIgHrearrangementassociatedwiththe19-28zCAR.

Foralldeepquencingdata,if<7.5mgofgenomicDNAwassub-

mitted,resultswerenormalizedtotheoutputexpectedfrom7.5mgof

genomicDNA.

RelapdorrefractoryB-ALLdiagnosisand

clinicaloutcomes

B-ALLdiagnoswereconfirmedbypathologistsatMSKCConthe

basisofBMcellmorphology,flowcytometry,and/

standardcriteria(24,25),weclassifiedpatientoutcomesafterCART

cellinfusionasCR,molecularCR(CRm),CRwithincompleteplatelet

orneutrophilrecovery(CRi),MRD,ormorphologicresidualdia.

finedasthedis-

hould

berestorationofnormalhematopoiesiswithaneutrophilcount≥1000×

106/literandaplateletcount≥100,000×106/shouldbe<5%

rast,CRiisdefinedas

meetingthecriteriaforCRbutnothavingadequateplateletorneutro-

efinedaspatientsmeetingthecriteriaforCRor

CRi,butwithresidualdiadetectedbyqPCR,flowcytometry,or

-

trast,CRmcorrespondstopatientsinaCRorCRibutalsoconfirmedto

havenoMRD,thatis,MRD−,asdeterminedbyflowcytometryand/or

deepquencingand/logic

residualdiaisdefinedas≥5%blastsinaBMdifferential.

Statistics

QuantitativedatawereanalyzedwithttestsandANOVA,whenap-

alnumberofsamples,thestatisticaltest,andPvalues

7cytokinesassociatedwithsCRS

wereidentifiedbyscreening39cytokinesforstrongSpearmanrank

ordercorrelations(r≥0.4)betweencytokinemaxfoldchangeand

he,wethenlectedonly

,

wedevelopedthethresholdbarrierbyidentifyingthelowestmaxfold

curvesfor

CRPwereconstructedusingtheempiricalmethod,andthebestcut

pointwasidentifiedviatheYoudenindex(26).

SUPPLEMENTARYMATERIALS

/cgi/content/full/6/224/224ra25/DC1

cheme.

elsinpatientswithsCRS.

veforCRP.

mentschemeforpatientstreatedwithCARTcells.

sisandTcellproductioncharacteristics.

ionofBcellsandCARTcellsintheBMbydeepquencing.

temaximumcytokinevaluesafterCARTcellinfusion.

eevents.

llsintheCSFofpatientswithneurologicchanges.

REFERENCESANDNOTES

ens,,e,,,,o,ski,,

ska,z-Ojeda,,ewska,,,,,

,,erz,,lat,,ni,in,

CD19-targetedTcellsrapidlyinducemolecularremissionsinadultswithchemotherapy-

.5,177ra38(2013).

,,t,,,old,y,,

,,,,,Chimericantigenreceptor-modified

.368,1509–1518(2013).

derfer,,n,,,,

r-Stevenson,,,,,a,

ier,ter,,,cot,erg,B-cellde-

pletionandremissionsofmalignancyalongwithcytokine-associatedtoxicityinaclinicaltrial

119,2709–2720(2012).

in,ens,ère,Thebasicprinciplesofchimericantigenreceptorde-

Discov.3,388–398(2013).

,ens,,ère,in,HowdoCARswork?:Earlyinsights

munology1,1577–1583(2012).

ens,ère,,,,ski,,,o,

z-Ojeda,ska,,,lowski,an,nko,

lia,,,,,berg,,,,

ni,in,SafetyandpersistenceofadoptivelytransferredautologousCD19-

targetedTcel

118,4817–4828(2011).

,,,,,,,Tcellswith

chimericantigenreceptorshavepotentantitumoreffectsandcanestablishmemoryin

.3,95ra73(2011).

ng,ds,,s,,,t,,

,in,an,n,,one;Medical

RearchCounciloftheUnitedKingdomAdultALLWorkingParty;EasternCooperativeOn-

cologyGroup,Outcomeof609adultsafterrelapofacutelymphoblasticleukemia(ALL);an

MRCUKALL12/109,944–950(2007).

9.N.Gökbuget,,,ch,,A.Hüttmann,,r,

,e,h,tz,,,es,ann,t,

in,,r;GermanMulticenterStudyGroupforAdultAcuteLym-

phoblasticLeukemia,Outcomeofrelapdadultlymphoblasticleukemiadependsonre-

spontosalvagechemotherapy,prognosticfactors,andperformanceofstemcell

120,2032–2041(2012).

10.S.O’Brien,er,,,rg,ky,-Yehuda,,

,,r,,,,ine,o,,

,,h,erg,,der,man,,

er,jian,High-dovincristinesulfateliposomeinjectionforadvanced,

relapd,andrefractoryadultPhiladelphiachromosome-negativeacutelymphoblasticleu-

.31,676–683(2013).

,Immunoquencing:.

l.25,646–652(2013).

,field,C-reactiveprotein:.111,

1805–1812(2003).

RESEARCHARTICLE

19February2014Vol6Issue224224ra259

o

n

F

e

b

r

u

a

r

y

2

1

,

2

0

1

4

s

t

m.

s

ci

e

n

c

e

m

a

g.

o

r

g

D

o

w

nl

o

a

d

e

d

f

r

o

m

,jian,,,,S.O’Brien,,ki,

,g,Primaryrefractoryandrelapdadultacutelymphoblasticleuke-

mia:Characteristics,treatmentresults,86,

1216–1230(1999).

rg,,,,,,-Malaspina,,

,wski,,,poulos,Brink,,

,,R.J.O’Reilly,p,m,,s,Tcell–

depletedstemcelltransplantationforadultswithhigh-riskacutelymphoblasticleukemia:Long-

termsurvivalforpatientsinfirstcompleteremissionwithadecreadriskofgraft-versus-host

arrowTransplant.19,208–213(2013).

,,ández-Rivas,,,,court,dó,

,,ro,a,,,,,,

a,,;ProgramaEspañoldeTratamientoenHematologiaGroup,Out-

comeafterrelapofacutelymphoblasticleukemiainadultpatientsincludedinfourconc-

ologica95,589–596(2010).

,,,,,Chimericantigenreceptor–modified

.365,725–733(2011).

,,ito,l,,y,a,,

,-Desnoyers,a,,,,Chimericreceptors

containingCD137signaltransductiondomainsmediateenhancedsurvivalofTcellsand

.17,1453–1464(2009).

derfer,,,,r-Stevenson,n,

,d,,,,erg,EradicationofB-lineage

cellsandregressionoflymphomainapatienttreatedwithautologousTcellsgenetically

116,4099–4102(2010).

Stegen,,,,wski,,ng,

ic,-Maghami,,n,e-Pereira,,

PreclinicalinvivomodelingofcytokinereleasyndromeinducedbyErbB-retargetedhuman

Tcells:Identifyingawindowoftherapeuticopportunity?l.191,4589–4598(2013).

y,old,,er,t,,s,

,,,rald,,,,Cytokine

releasyndromeafterblinatumomabtreatmentrelatedtoabnormalmacrophageactiva-

121,5154–5157(2013).

ralingam,,,,lo-Cortes,r,altsis,

Cyt.

355,1018–1028(2006).

,,,in,CD19CAR-targetedTcellsinducelong-

termremissionandBcellaplasiainanimmunocompetentmoumodelofBcellacute

e8,e61338(2013).

an,ski,lowski,o,z-Ojeda,,,o,

ska,,in,ens,ère,Manufacturingvalidationofbi-

ologicallyfunctionalTcellstargetedtoCD19antigenforautologousadoptivecelltherapy.

ther.32,169–180(2009).

,t,y,T.Büchner,n,,er,

r,n,,-Coco,ze,,ann,,

B.Löwenberg,,,,ield;InternationalWorkingGroupfor

Diagnosis,StandardizationofResponCriteria,TreatmentOutcomes,andReportingStandardsfor

TherapeuticTrialsinAcuteMyeloidLeukemia,RevidrecommendationsoftheInternational

WorkingGroupforDiagnosis,StandardizationofResponCriteria,TreatmentOutcomes,andRe-

.21,4642–4649

(2003).

aum,lum,,l,eld,,,

,,S.O’Brien,,,,,n,End

point109,

1810–1816(2007).

26.M.Gönen,AnalyzingReceiverOperatingCharacteristicCurveswithSAS(SASInstitute,Cary,

NC,2007).

erz,,erz,,,,Developmentofanew

intensivetherapyforacutelymphoblasticleukemiainchildrenatincreadriskofearlyre-

72,3120–3130(1993).

ani,li,i,e-Schonk,ze,,o,

,li,,Adolescentandadultacutelymphoblasticleukemia:Prognostic

60,677–684(1982).

uez,,ich,,ie,arJr.,n

Jr.,32,69–75(1973).

Funding:NCI(M.L.D.,R.B.,I.R.,in),TerryFoxFoundation(R.B.),AmericanSocietyof

Hematology–AmosMedicalFacultyDevelopmentProgram(M.L.D.),AllianceforCancerGeneTher-

apy(in),MallahFoundation(in),MajorsFoundation(in,R.B.,andI.R.),

TheDamonRunyonCancerRearchFoundation(R.B.),theCarsonFamilyCharitableTrust(R.B.),the

WilliamLawrenceandBlancheHughesFoundation(R.B.),Kate’sTeam,n

oodwinandtheCommonwealthCancerFoundationforRearchandtheExperi-

mentalTherapeuticsCenterofMSKCC(in,R.B.,andI.R.).Authorcontributions:R.B.,

in,M.L.D.,ndeditedthemanuscript.M.L.D.,in,R.B.,andI.R.

conceptualizedtheoverallstrategyanddevelopeditsclinicaltranslationandimplementation.R.B.

rincipalinvestigator

cturingofTcells,releatesting,andqPCRacquisitionofclinicalsamples

wereperformedbyS.B.,J.S.,O.B.-O.,M.O.,J.Q.,T.W.,Q.H.,M.F.,H.S.,M.Y.,,Y.W.,andJ.S.;su-

pervidbyX.W.;ommanufacturing,flowcytometry,andqPCR

.R.R.B.,in,I.R.,M.L.D.,dandinter-

pretedtheresults.R.B.,M.L.D.,J.P.,K.C.,D.D.,S.S.C.,G.J.R.,H.Q.,E.H.,S.G.,edpatients

totheprotocoland/edandperformedmolecular

assaystoidentifythemalignantIgHclonotypeassociatedwiththeleukemiacellsofenrolledand

tedallpre-andposttreatmentBMaspiratesforevidenceofleukemia.

earchstudyassistantfortheprotocol

andassistedwithenrollment,sampleacquisition,

thedataassistantfortheprotocolandarrangedcollectionandprentationofthedataforthestudy

inginterests:7,446,190,which

coversin,

R.B.,erauthorsdeclarenocompetinginterests.

Submitted9December2013

Accepted24January2014

Published19February2014

10.1126/scitranslmed.3008226

Citation:,e,,o,,,,ski,

z-Ojeda,ska,,ewska,,,ot,f,

,,,nilla,,,sira,,

,,,,ni,,in,ens,

Efficacyandtoxicitymanagementof19-28zCARTcelltherapyinBcellacutelymphoblastic

.6,224ra25(2014).

RESEARCHARTICLE

19February2014Vol6Issue224224ra2510

o

n

F

e

b

r

u

a

r

y

2

1

,

2

0

1

4

s

t

m.

s

ci

e

n

c

e

m

a

g.

o

r

g

D

o

w

nl

o

a

d

e

d

f

r

o

m