SecondtransplantforacuteandchronicleukemiarelapsingafterfirstHLA-identicalsiblingtransplant
MEapen1,SAGiralt2,MMHorowitz1,JPKlein1,JEWagner3,M-JZhang1,MSTallman4,DIMarks5,
n7,CNBredeson1,RMartino8,RPGale9,MSCairo10,BMCamitta6,REChamplin2,ORingde
MRLitzow11andMdeLima2InternationalBoneMarrowTransplantRegistry,HealthPolicyInstitute,MedicalCollegeofWisconsin,Milwaukee,WI,USA;MDAndersonCancerCenter,Houston,TX,USA;3UniversityofMinnesota,Minneapolis,MN,USA;4NorthwesternUniversityMedicalSchool,Chicago,IL,USA;5BristolChildren’sHospital,Bristol,UK;6MedicalCollegeofWisconsin,Milwaukee,WI,USA;7HuddingeUniversityHospital,Huddinge,Sweden;8HospitalSanCreuISantPau,Barcelona,Spain;9CenterforAdvancedStudiesinLeukemia,LosAngeles,CA,USA;10ColumbiaUniversity,NewYork,NY,USA;and11MayoClinicRochester,Rochester,MN,USA
21Summary:
Treatmentoptionsforpersonswithleukemiarelapsingafterallogeneictransplantationarelimited.Weanalyzedtheoutcomeof279patientswithacuteandchronicleukemia,whorelapsedafterHLA-identicalsiblingtransplantationandreceivedasecondallogeneictransplant.Theinfluenceofpotentialriskfactorsontreatment-relatedmortality(TRM),relapse,treatmentfailure(relapseordeath)andoverallsurvivalaftersecondtransplantationwereassessedusingproportional-hazardsregression.Thecumulativeincidences(95%confidenceinterval)ofrelapseandTRMat5yearswere42(36–48)%and30(24–36)%,respectively.The5-yearprobabilitiesofbothoverallandleukemia-freesurvivalwere28(23–34)%.Inmultivariateanalyses,risksoftreatmentfailureandmortalitywerelowerinyoungerpatients(p20years)andpatientswhorelapsedafter6monthsfromfirsttransplantation.Risksofrelapsewerelowerinpatientswhorelapsedafter6monthsfromfirsttransplantationandincompleteremissionpriortosecondtransplantation.Risksofrelapsewerehigherafterreduced-intensityconditioningregimens.Anypotentialadvantageofusingadifferentmatchedrelateddonorforasecondtransplantationisnotsupportedbythesedata.Althoughage,diseasestatusandconditioningregimenareimportant,durationofremissionafterfirsttransplantationappeartobethemostimportantdeterminantofoutcome.
BoneMarrowTransplantation(2004)34,721–727.doi:10.1038/sj.bmt.1704645
Publishedonline23August2004
Keywords:secondtransplantation;leukemia;durationofremission;matchedrelateddonor;reduced-intensityconditioningregimen
Correspondence:DrMEapen,StatisticalCenter,InternationalBoneMarrowTransplantRegistry,MedicalCollegeofWisconsin,8701WatertownPlankRoad,POBox26509,Milwaukee,WI53226,USA;E-mail:meapen@mail.mcw.edu
Received27January2004;accepted2June2004Publishedonline23August2004
Leukemiarecursin20–70%ofpersonsafterallogeneictransplantation,withriskofrecurrencedependingprimar-ilyonthetypeofleukemiaanddiseasestatusattransplantation.1–3Prognosisofpatientswhorelapseispoorandanoptimalsalvagetherapyhasyettobeestablished.Currenttreatmentoptionsincludechemother-apy,donorleukocyteinfusions(DLI)andsecondallogeneictransplantation.Intensivechemotherapeuticregimenscaninduceremissioninsomepatientswithrecurrentleukemia,butremissiondurationsareshort,andmostpatientsdieofuncontrolledleukemiaorinfectiouscomplications.DLIresultsindurableclinicalandmolecularremissioninchronicmyelogenousleukemia(CML),especiallywhenusedforearlydisease.4–9ResultsofDLIforacuteleukemiarelapsingafterallogeneictransplantationarelessencoura-ging;only10–20%ofpatientsachievecompleteremission(CR)andremissionsareusuallybrief.6,10,11Moreover,acuteandchronicgraft-versus-hostdisease(GVHD)andinfectionsfrommarrowaplasiaand/orimmunosuppressivetherapyafterDLIareassociatedwithsubstantialmorbidityandmortality.4,9,11AnewtherapyforCMLisimatinibmesylate,atyrosinekinaseinhibitor.Intheshortterm,imatinibmaybeusedsuccessfullytoinduceandmaintainremissioninCMLrecurringafterallogeneictransplanta-tion.12However,follow-upofimatinib-treatedpatientsisshortandfurtherstudiesarerequiredtodocumentdurabilityofremissions.
Secondallogeneictransplantationmaybeeffectivesalvagetherapyinsomepatientswhorelapseafteraninitialtransplant.1,3,13–21Thusfar,moststudiesofsecondtransplantsforrecurrentleukemiaarelimitedbysmallnumbersofpatientswithheterogeneousprognosticfactors.Whileearlyrelapseafterfirsttransplantationandadvanceddiseasestatusatsecondtransplantationpredictpooroutcome,dataonotherfactorsthatmayaffectoutcomessuchasconditioningregimenanduseofadifferentdonorarefew.Tobetteridentifyriskfactorsthatinfluenceoutcome,westudied279recipientsofsecondtransplantsfromanHLA-identicalsiblingdonorforleukemiarecur-ringafterinitialHLA-identicalsiblingtransplantationand
OutcomesaftersecondallogeneictransplantationMEapenetal722reportedtotheInternationalBoneMarrowTransplantRegistry(IBMTR).
MethodsDatacollection
Dataonpatientsundergoingsecondallogeneictransplan-tationwereobtainedfromtheStatisticalCenteroftheIBMTR.TheIBMTRisavoluntaryworkinggroupofover400transplantteamsworldwidethatcontributedetailedinformationontheirallogeneictransplantstoaStatisticalCenterattheMedicalCollegeofWisconsin.Participatingcentersarerequiredtoregisterallconsecutivetransplantrecipients.TheIBMTRdatabaseincludesinformationon40–45%ofallogeneictransplantsperformedworld-widesince1970.Transplantrecipientsarefollowedlong-itudinally.Computerizederrorchecks,physicianreviewofsubmitteddata,andon-siteauditsofparticipatingcentersensuredataquality.
Inclusioncriteria
Thestudyincludespatients,withacutelymphocyticleukemia(ALL),acutemyeloidleukemia(AML)orCML,whoreceivedasecondallograftfromanHLA-identicalsiblingforrecurrentorpersistentleukemiaafterinitialHLA-identicalsiblingtransplantation,andtrans-plantedbetweenJanuary1,1990andDecember31,2000.Inall,279patientsfulfilledtheeligibilitycriteria(Table1).DatafromtheIBMTRsuggestthatapproximately6%ofpatients,withrecurrentleukemiaafterafirstmatchedrelateddonortransplantreceiveasecondmatchedrelateddonortransplantusingthesameoradifferentdonor.PatientswhoreceivedDLIforrecurrentleukemiaaftertheirinitialtransplantwereexcluded.TransplantteamsreportingtotheIBMTRarerequestedtoindicatewhetherpatientsreceivedabloodormarrowinfusionasasecondorsubsequenttransplant,orDLI.Inconsistenciesareresolvedbyeitheradatamangerorphysicianatthetransplantcenter.
Endpoints
Thestudyexaminedhematologicrecovery,acuteandchronicGVHD,treatment-relatedmortality(TRM),re-lapse,leukemia-freesurvival(LFS),treatmentfailure(relapseordeath,inverseofLFS)andoverallsurvivalaftersecondallogeneictransplantationforrecurrentleukemia.Theoutcomedatawerecompiledfromthedateofsecondtransplantationtodateoflastcontactordeath.Hematopoieticrecoverywasdefinedasachievinganabsoluteneutrophilcount(ANC)X500/mlandplateletsX20000/ml.Theincidenceofgrades2–4acuteGVHDwasevaluatedinallpatients;22,23chronicGVHDwasevaluatedinpatientssurviving90daysorlongeraftersecondtransplant.24TRMwasdefinedasdeathincontinuousremission;patientswerecensoredatrelapseor,forthoseinclinicalremission,atlastfollow-up.Relapsewasdefinedashematologicrecurrenceforbothacuteandchronic
BoneMarrowTransplantationTable1Characteristicsof279patientsreceivingasecondHLA-identicalsiblingtransplantforrecurrent/persistentdiseaseaftertheirinitialHLA-identicalsiblingtransplantVariable
1st2ndtransplanttransplantN(%)
N(%)DiseaseAML125(45)125(45)ALL72(26)72(26)CML82(29)82(29)Age(years)p11–20105155(18)(20)4056(14)(20)21–3065(23)64(23)430
108
(39)
119(43)KarnofskyscoreX90%229(83)154(55)Year,transplant1983–198998(35)—1990–1994124(44)188(67)1995–200057(20)91(33)Samedonoras1sttransplant—238(85)Donor–recipientsexmatchMale-male110(39)103(37)Male-female53(19)50(18)Female-male65(23)72(26)Female-female
51
(19)
54(19)Timefrom1sttransplanttorelapse,median—
14(1–155)
(range),months
Timefromrelapseto2ndtransplant,median—4(o1–48)
(range),months
DiseasestatusattransplantRemission212(76)144(52)Relapse67(24)135(48)Grafttype
Bonemarrow264(95)220(79)Peripheralblood
12(4)59(21)
Umbilicalcordblood
3(1)—MyeloablativeconditioningregimenCy/TBI7other134(48)63(23)TBI7other
16(6)27(10)Busulfan/Cy7other127(46)114(41)Melphalan7other2(o1)6(2)Cy/other
—4(1)Busulfan/other—20(7)ReducedintensityconditioningregimenaTBIregimen
—1(o1)Non-TBIregimen—44(16)GVHDprophylaxisNone
—38(14)CSA7other59(21)96(34)MTX7other
21(8)29(10)CSA/MTX7other148(53)99(35)Tdepletion7other41(15)3(1)Other4(1)4(1)Missing
6(2)
10(4)
Follow-upofsurvivors,median(range),93(6–147)
months
AML¼acutemyeloidleukemia;ALL¼acutelymphoidleukemia;CML¼chronicmyeloidleukemia;TBI¼totalbodyirradiation;CSA¼cyclosporine;MTX¼methotrexate;Cy¼cyclophosphamide;Ara-C¼cytosinearabinoside.aReducedintensityconditioningregimensweredefinedas:anyfludarabine-containingregimen;TBIo400cGy7otheragents;melphalanp150mg/m27Ara-C/etoposide;cyclophosphamide7Ara-C;busulfanp12mg/kg7Ara-C/etoposide;high-doseAra-C7etoposide.
leukemia;patientswhoneverachievedremissionaftersecondtransplantationwereconsideredtohavehadarecurrenceatday1.LFSwasdefinedassurvivalincontinuousCR;relapseordeathwereconsideredevents;andpatientssurvivingincontinuousCRwerecensoredatlastfollow-up.
Statisticalanalysis
ProbabilitiesofLFSandoverallsurvivalwerecalculatedusingtheKaplan–Meierestimator.25Cumulativeincidencerates(thechanceapatientwillhaveexperiencedaparticularcompetingriskpriortotimet,andwheredeathwithoutaneventisthecompetingrisk)werecalculatedusingstandardtechniquesforhematopoieticrecovery,acuteandchronicGVHD,TRMandrelapse.25ThestandarderrorofthesurvivorfunctionbyGreenwoodformulawasusedtocalculatethe95%confidenceintervals(CI).25PotentialprognosticfactorsforTRM,relapse,treatmentfailureandoverallmortalitywereevaluatedinmultivariateanalysesusingCoxproportional-hazardsregression.26Multivariatemodelswerebuiltusingastepwiseforwardselectionwithasignificancelevelof0.05.VariablesconsideredinmultivariateanalysesareshowninTable2.Whenevercategoriesofvariablesinitiallyclassifiedinto
Table2VariablestestedinCoxproportional-hazardsregression
models
Patient-relatedvariables
Ageat2ndtransplanta:p10vs11–20yearsvs420–30vs430years*Gender:malevsfemale*Disease-relatedvariables
Typeofleukemia:ALLvsAMLvsCML*
Diseasestatusat2ndtransplant:relapse/persistentdiseasevsremission*
Timefrom1sttransplanttorelapseb:p6vs46–12vs412months*Timefromrelapseto2ndtransplant:p3monthsvs43months*Transplant-relatedvariables1sttransplant
Diseasestatusat1sttransplant:relapse/persistentdiseasevsremission*
Conditioningregimen:non-TBIregimenvsTBI*
GVHDprophylaxis:TdepletionvsT-repleteallograft*Grades2–4acuteGVHDafter1sttransplantChronicGVHDafter1sttransplant2ndtransplant
Yearoftransplant:1995–2000vs1990–1994*Grafttype:peripheralbloodvsbonemarrow*
Donor:differentsiblingothervssamedonoras1sttransplant*Donor–recipientsexmatch:othervsfemale-male*
Conditioningregimen:TBIregimenvsreducedintensityvsnon-TBIregimen*
GVHDprophylaxis:CSA7MTX7steroidsvsothervsnone**Referencegroup.aAgeat2ndtransplant:therewerenodifferencesbetweenagegroupsp10and11–20years,and420–30and430years.Therefore,thefinalmodelwastestedas420yearsvsp20years*.bTimefrom1sttransplanttorelapse:therewerenodifferencesbetween46–12and412months.Therefore,thefinalmodelwastestedasp6vs46months*.
OutcomesaftersecondallogeneictransplantationMEapenetal723morethantwocategoriesshowednostatisticallysignificantdifferencesbetweencategories,categorieswerecollapsedtocreatethefewestpossiblenumbergroups(suchas:ageanddurationofremissionafterfirsttransplant).Ineachmodel,theassumptionofproportionalhazardswastestedforeachvariableusingatime-dependentcovariate.First-orderinteractionsbetweenvariableswereexaminedbyfittingaproportional-hazardsmodelandexamininginteractionbetweenvariables.AssociationsbetweenGVHDandleukemiarecurrenceaftertransplantationweretestedbymodelingacuteandchronicGVHDastime-dependentcovariateswithothersignificantvariablesinthefinalmultivariatemodelforrelapse.Allmultivariatemodelswereexaminedforcentereffectsbyusingrandomeffectsorfrailtymodel;27therewasnoevidenceofconfoundingbycentereffects.Completenessoffollow-upwasassessedusingaCstatistic,whichgivesameasureoftheproportionofallpotentialfollow-upinformationavailableforthisstudy.28AllP-valuesaretwo-sided.AllanalyseswerecarriedoutusingPROCPHREGinSASversion8.0(SASInstituteInc.,Cary,NC,USA).
Results
Patientandtransplantcharacteristics
Patient,diseaseandtransplantcharacteristicsatfirstandsecondtransplantationarepresentedinTable1.AllpatientsreceivedallograftsfromHLA-identicalsiblingdonors;85%receivedallograftsfromthesamedonorforbothtransplants.Grafttype,conditioningregimenandGVHDprophylaxisweredeterminedbytransplantteams.Inall,90%ofsecondtransplantsforrecurrentCMLwerebetween1990and1994.Bonemarrowwasthemostcommongrafttype.Nopatientreceivedanirradiationcontainingconditioningregimenforbothtransplants,and16%receivedareduced-intensityconditioningregimenfortheirsecondtransplant.Themedianfollow-upofsurvivorsis93(range6–147)monthsandthecompletenessoffollow-upis90%.27Atotalof90%ofsurvivorswerefollowedforatleast2yearsaftersecondtransplantation.
Hematopoieticrecovery
Mostpatientsachievedhematopoieticrecovery.Themediantimestoneutrophilandplateletrecoverywere16(range,6–82)daysand23(8–364)days,respectively.Thecumulativeincidences(95%CI)ofachievingANCX500/mlatday100andplateletsX20000/mlat1yearwere86(77–92)%and83(72–90)%,respectively.
AcuteandchronicGVHD
Thecumulativeincidenceofgrade2–4acuteGVHDatday100was29(24–35)%.ThecumulativeincidencesofchronicGVHDat1and5yearswere39(31–46)%and41(33–48)%,respectively.T-celldepletionforfirsttransplantationornothavingreceivedGVHDprophylaxisforthesecondwasnotassociatedwithoutcomesaftersecondtransplanta-tion.Tofurtherexaminetheassociation,ifany,
BoneMarrowTransplantationOutcomesaftersecondallogeneictransplantationMEapenetal724withGVHDprophylaxisatsecondtransplantationandoutcome,wecreatedfourcategories:samedonor,sameGVHDprophylaxisforbothtransplants,samedonor,differentGVHDprophylaxisforbothtransplants,differentdonor,sameGVHDprophylaxisforbothtransplantsanddifferentdonor,differentGVHDprophy-laxisforbothtransplants.Duringmodelbuilding,thisdidnotattainasignificancelevelof0.05.Nevertheless,thiswasforcedintothefinalmodelforTRM,relapse,treatmentfailureandoverallsurvival,andwedidnotobserveanassociation.
AmongpatientswhodidnotdevelopacuteGVHDafterfirsttransplantation,ratesofacuteGVHDaftersecondtransplantationwiththesameoradifferentdonorwere29%and26%,respectively.Incontrast,patientswhoreceivedallograftsfromthesamedonorforbothtrans-plantsanddevelopedgrade2–4acuteGVHDaftertheirfirsttransplantweremorelikelytodevelopgrade2–4acute
Table3Variable
TRM
Ageat2ndHSCT(years)p20420
Timefrom1stHSCTtorelapse(months)46p6
Relapse
Timefrom1stHSCTtorelapse(months)46p6Diseasestatusat2ndHSCTRemission
Relapse/persistentdiseaseConditioningregimenat2ndHSCTaNon-TBImyeloablativeregimenTBImyeloablativeregimenReducedintensityregimenTreatmentfailure(inverseofLFS)Ageat2ndHSCT(years)p20420
Timefrom1stHSCTtorelapse(months)46p6
Diseasestatusat2ndHSCTRemission
Relapse/persistentdiseaseOverallmortality
Ageat2ndHSCT(years)p20420
Timefrom1stHSCTtorelapse(months)46p6
aGVHDaftertheirsecond,relativerisk(RR)1.74,95%CI1.02–2.96,P¼0.04.AsimilartrendwasnotobservedamongpatientswhodevelopedacuteGVHDafterfirsttransplant,butreceivedthegraftfromadifferentdonorforthesecondtransplant.ChronicGVHDwasmorelikelyaftersecondtransplantinpatientswhoreceivedtheirallograftfromadifferentdonorratherthanfromthesamedonorforbothtransplants(61vs37%,P¼0.03),butthisdidnottranslateintoasurvivaladvantage(39vs28%,P¼0.2).
Treatment-relatedmortality
RisksofTRMwerehigherinpatientsolderthan20yearsofageandinthosewhorelapsedp6monthsaftertheirfirsttransplant(Table3).ThecumulativeincidencesofTRMat1and5yearswere26(21–31)%and30(24–36)%,respectively(Figure1).The1-yearcumulativeincidenceof
Factorsassociatedwithoutcomesaftersecondtransplantation
N
RR(95%CI)
P-value
9618321352
1.00
2.82(1.69–4.75)1.00
2.33(1.28–4.24)
o0.0001
0.006
213521441351449045
1.00
3.02(1.97–4.65)1.00
2.47(1.66–3.69)1.00
1.19(0.78–1.84)1.89(1.16–3.05)
o0.0001
o0.00010.030.420.01
9618321352144135
1.00
1.81(1.32–2.49)1.00
3.07(2.18–4.32)1.00
1.38(1.04–1.85)
0.0003
o0.0001
0.03
9618321352
1.00
1.94(1.42–2.64)1.00
3.47(2.49–4.83)
o0.0001
o0.0001
ReducedintensityvsTBImyeloablativeregimen*:1.58(0.95–2.62)P¼0.07.*Referencegroup.
BoneMarrowTransplantation%1.0 ,ecn0.8edicn0.6iRelapse
evi0.4talum0.2TRM
uC0.0
0
1
23
4
5
Years
Figure1CumulativeincidenceofrelapseandTRMaftersecondtransplantation.
1.0%0.8A y,til0.6BCibaDbEo0.4rPF0.2GH
0.0
0
1
23
4
5
Years
Figure2ProbabilityofLFSaftersecondtransplantation.(A)Agep20years,durationofremission46monthsandinCR;(B)agep20years,durationofremission46monthsandnotinCR;(C)age420years,durationofremission46monthsandinCR;(D)age420years,durationofremission46monthsandnotinCR;(E)agep20years,durationofremissionp6monthsandinCR;(F)agep20years,durationofremissionp6monthsandnotinCR;(G)age420years,durationofremissionp6monthsandinCR;and(H)age420years,durationofremissionp6monthsandnotinCR.
TRMinpatientswithneitheroftheseriskfactors(n¼71)was20(12–30)%;inpatientswithbothriskfactors(n¼36),thecumulativeincidencewas25(14–38)%.
Relapse
Thecumulativeincidencesofrelapseat1and5yearswere36(31–42)%and42(36–48)%,respectively(Figure1).Riskofrelapsewashigherinpatientswithashort(p6months)remissionafterfirsttransplant,thosewhowerenotinCRatsecondtransplantandthosewhoreceivedreduced-intensityconditioning(Table3).Grade2–4acuteGVHDaftersecondtransplantwasassociatedwithlowerriskofrelapse,RR0.50,95%CI0.27–0.93,P¼0.03.RiskofrelapsewassimilarinpatientswithandwithoutchronicGVHDaftersecondtransplant.
Treatmentfailure
ProbabilitiesofLFSat1and5yearswere38(32–43)%and28(23–34)%,respectively.Riskoftreatmentfailurewashigherinpatientsolderthan20yearsofage,inthosewhorelapsedp6monthsaftertheirfirsttransplantandinthosenotinCRatsecondtransplantation(Table3).Among
OutcomesaftersecondallogeneictransplantationMEapenetal7251.00.8% y,t0.6Ailibab0.4oBrP0.2C0.0
D
0
1
23
4
5
Years
Figure3Probabilityofoverallsurvivalaftersecondtransplantation.(A)Agep20years,durationofremission46months;(B)age420years,durationofremission46months;(C)agep20years,durationofremissionp6months;and(D)age420years,durationofremissionp6months.
patientswithnoneoftheseriskfactors(n¼71),the5-yearprobabilityofLFSwas49(36–60)%;amongthosewithalltheseriskfactors(n¼36),theprobabilitywas3(0.02–12)%(Figure2).
Overallsurvival
Probabilitiesofoverallsurvivalat1and5yearswere41(35–46)%and28(23–34)%,respectively.Riskofmortalitywashigherinpatientsolderthan20yearsofageandinthosewhorelapsedp6monthsaftertheirfirsttransplant(Table3).Amongpatientswithneitheroftheseriskfactors(n¼71),the5-yearprobabilityofoverallsurvivalwas51(38–62)%;amongthosewithbothriskfactors(n¼36),theprobabilitywas3(0.02–12)%(Figure3).Diseasestatusattransplantationwasnotassociatedwithoverallmortality,RR1.26,95%CI0.94–1.69,P¼0.12.
Intotal,202patients(72%)died.Recurrentleukemia(58%)wasthemostfrequentlyreportedprimarycauseofdeath.OtherreportedcausesincludedGVHD(11%),interstitialpneumonitis(12%),infections(9%),organfailure(5%),secondmalignantneoplasm(1%)andothercauses(4%).Themedian(range)timefromsecondtransplantationtodeathwas4(o1–100)months.
Discussion
Asecondallogeneictransplantforleukemiarecurringafterfirsttransplantmaybeareasonabletreatmentoptioninselectedpatients.However,identificationoffactorsinfluen-cingrelapseandmortalityaftersecondtransplantationanddeterminationofpatientgroupswhobestrespondtothistherapyhaveprovendifficult.UsingdatafromtheIBMTR,Mrsicetal1reportedoutcomesin114recipientsofsecondHLA-identicalsiblingtransplantsbetween1978and1989.Intheirstudy,relapseoccurringlaterthan6monthsafterfirsttransplant,diagnosisofCMLandCRatsecondtransplantwereassociatedwithabetteroutcome.Theprimaryobjectiveofthecurrentstudywastoidentifypotentialriskfactorsinamorecontemporaryperiod.Wefoundlengthofremissionafterfirsttransplant,andage,diseasestatusandconditioningregimenatsecondtrans-BoneMarrowTransplantationOutcomesaftersecondallogeneictransplantationMEapenetal726planttobeimportantindeterminingoutcomesaftersecondtransplant.AlthoughacuteGVHDaftersecondtransplantcorrelatedwithalowerriskofrelapse,wefoundnoevidencetosupporttheuseofadifferentmatchedrelateddonortoincreasethelikelihoodofdevelopingGVHDaftersecondtransplantation.
Ourdata,aswellasotherreports,indicatethatlengthofremissionafterfirsttransplantationisprobablythemostimportantvariablepredictingoutcomeaftersecondtrans-plant.1,3,15,17–21Relapse,treatmentfailureandmortalityarehigherinpatientswithearlyrelapse.Wedefinedearlyrelapseasoccurringwithin6monthsafterfirsttransplant.Interestingly,weobservedsimilaroutcomesforpatientswhorelapsed7–12monthsandthosewhorelapsedlateraftertheirfirsttransplant.Thiscontrastswithotherrecentreportssuggestingbetteroutcomesinpatientsrelapsinginthefirstyearaftertransplantation.3,20,21Similartootherreports,wedidnotobserveacorrelationbetweenintervalfromrelapsetosecondtransplantandtreatmentfailureormortality.1,3,17,20Asexpected,patientswithactiveleukemiaattransplantwerealsolesslikelytodowell,similartootherreports.1,3,19,20AgecorrelatedwithTRM,treatmentfailureandoverallmortality,probablytheresultofbettertoleranceoftherapyinyoungerpatients.Factorsgenerallyacceptedaspredictingoutcomesafterfirsttransplant,suchasdurationofpretransplantremissionforacuteleukemia,intervalfromdiagnosistofirsttransplantforchronicleukemia,anddiseasestatusatfirsttransplant,didnotcorrelatewiththeoutcomeaftersecondtransplant.Inaddition,thespecificconditioningorGVHDprophylaxisregimenforfirsttransplantdidnotseemtopredicttheoutcomeofthesecondtransplantation.
Incontrasttosomereports,outcomeswerenotdifferentamongthethreediseasegroups,ALL,AMLandCML.The3-yearprobabilitiesofLFSforAML,ALLandCMLwere27(19–35)%,30(20–42)%and32(23–43)%,respectively.Correspondingprobabilitiesofoverallsurvi-valwere27(19–35)%,30(20–42)%and33(23–44)%.Althoughthevariablefortypeofleukemia(AMLvsALLvsCML)didnotattainasignificancelevelof0.05,thisvariablewasforcedintoallfinalmodelsandtheoutcomeswereunchanged.Wedefinedleukemiarelapseashematologicalrecurrenceforbothacuteandchronicleukemia,andthismayexplainwhywedidnotobservedifferencesinoutcomebytypeofleukemia.Wedonothavedataoncytogeneticsorchimerism,chemosensitivevsrefractorydisease,andotherhematologicalmarkerssuchasthepercentofcirculatingblasts,andplateletcountatrelapseafterfirsttransplantation.Whilethesefactorsmaybeofprognosticsignificance,wewereunabletoexaminetheirprognosticsignificanceinthisdatasetandacknowledgethisasoneofthelimitationsofaregistry-basedstudy.
Mosttransplantstrategiesrelyheavilyonmaximallytolerateddosesofchemotherapy,withorwithoutradia-tion,toeradicatecancer,andallograftstorescuepatientsfromtherapy-inducedmarrowaplasiaaswellastoprovideimmune-mediatedeffects.Somereportssuggestthatcon-ditioningregimenswithtotalbodyirradiation(TBI)maylowermortalityinpatientswithacuteleukemiaaftersecondtransplantation.3Wedidnotobservethis.Onlyathirdof
BoneMarrowTransplantationpatientsinthisstudyreceivedaTBI-containingregimenandAMLwasthemostcommondiseaseinthisgroup.MostpatientswithALLandCMLreceivedaTBI-containingregimenfortheirfirsttransplant.Morerecently,reduced-intensityconditioningregimensandpostgraftingimmunosuppressiontocontrolgraftrejectionandGVHDhavebeenusedtoreduceregimen-relatedtoxicitieswhilepreservinggraft-versus-leukemia(GVL)effects.29,30Wefoundhigherratesofrelapseafterreduced-intensityconditioning.Thiswasindependentofthedurationofremissionafterfirsttransplantanddiseasestatusatsecondtransplant,suggestingthatdoseintensitymaybeimportantinthissetting.Thesefindingsshouldbeinterpretedwithcautionduetolackofhomogeneityofreducedintensityconditioningregimens,aswellasprobablebiasesinpatientselectionforthisrelativelynewformoftherapy.
WefoundnoevidencesupportingtheuseofadifferentHLA-identicaldonorforasubsequenttransplant.Risksofrelapseandmortalityweresimilarwhetherornotadifferentdonorwasused.Usingadifferentmatchedrelateddonor,whenavailable,forasubsequenttransplanttoincreasethelikelihoodofGVHDandsubsequentGVLeffectisnotsupportedbythesedata.Asonly15%ofthestudypopulationreceivedtheirsecondallograftfromadifferentHLA-identicalsiblingdonor,smallnumbersmayhavelimitedourabilitytodetectdifferences.
Whilesomestudiesinadultsreportsurvivaladvantagesafterperipheralbloodstemcell(PBSC)infusionforadvancedleukemia,31,32others3reportlowerrisksofrelapseusingbonemarrowforsecondtransplant.Wedidnotobserveacorrelationbetweengraftsourceandrisksofrelapseormortality.However,asonly21%ofthestudycohortreceivedPBSCgrafts,smallnumbersmayhavelimitedourabilitytodetectaneffect.
DLI,andmorerecently,imatinibmesylatehavebeenusedsuccessfullytomaintaindurableclinicalandmole-cularremissioninCML.4–9,12ThisstudypopulationincludespatientswithCMLinhematologicrelapse.AlthoughcurrentstandardofcareforrecurrentCMLareDLIand/orimatinibmesylate,forthosewhofailthistherapy,asecondtransplantmaybeanoption.Assuch,thedatapresentedheremayserveasareferencewhencounselingthesepatientswhohavefailedDLIand/orimatinibmesylate.
Aretrospectivereportsuchasourshaslimitationssuchasselectionbiasfortransplantation,choiceofconditioningregimenandourinabilitytoadjustforunknownorunmeasuredfactorsthatmayaffecttheoutcomeaftertransplantation.Thedatasuggest,withamedianfollow-upofover7years,thatsecondallogeneictransplantationproducessustainedremissioninaproportionofpatients.Althoughdiseasestatus,ageandconditioningregimenareimportant,durationofremissionafterfirsttransplantappearstobethemostimportantdeterminantofoutcome.Thereseemslittlepointinsubjectingpatientswithpost-transplantremissiondurationsofp6monthstoasecondaggressiveprocedure.Useofadifferentmatchedrelateddonortoimproveoutcomesafterasubsequenttransplantisnotsupportedbythesedata.Furthercontrolledstudiesarerequiredtodefinetheroleofreducedvsstandardintensityconditioningregimensinthissetting.
Acknowledgements
ThisworkwassupportedbyaClinicalResearchCareerDevelopmentAwardfromtheAmericanSocietyofClinicalOncology(ME),PublicHealthServiceGrantU24-CA76518fromtheNationalCancerInstitute,theNationalInstituteofAllergyandInfectiousDiseases,theNationalHeart,LungandBloodInstituteandtheCancerCenter,MedicalCollegeofWisconsin.
References
1MrsicM,HorowitzMM,AtkinsonKetal.SecondHLA-identicalsiblingtransplantsforleukemiarecurrence.BoneMarrowTransplant1992;9:269–275.
2KumarL.Managementofrelapseafterallogeneicbonemarrowtransplantation.JClinOncol1994;7:50–57.
3BosiA,LaszloD,LabopinMetal.Secondallogeneicbonemarrowtransplantationinacuteleukemia:resultsofasurveybytheEuropeancooperativegroupforbloodandmarrowtransplantation.JClinOncol2001;9:3675–3684.
4MackinnonS,PapadopoulosEB,CarabasiMHetal.Adop-tiveimmunotherapyevaluatingescalatingdosesofdonorleukocytesforrelapseofchronicmyeloidleukemiaafterbonemarrowtransplantation:separationofgraft-versus-hostdis-ease.Blood1995;86:1261–1268.
5BacigalupoA,SoraccoM,VassalloFetal.Donorlymphocyteinfusion(DLI)inpatientswithchronicmyeloidleukemiafollowingallogeneicbonemarrowtransplantation.BoneMarrowTransplant1995;19:927–932.
6CollinsRH,ShpilbergO,DrobyskiWRetal.Donorleukocyteinfusionsin140patientswithrelapsedmalignancyafterallogeneicbonemarrowtransplantation.JClinOncol1997;15:433–444.7DazziF,SzydloR,GoldmanJetal.Donorlymphocyteinfusionsforrelapseofchronicmyeloidleukemiaafterallogeneicstemcelltransplant:Wherewenowstand.ExpHaematol1999;27:1477–1486.
8DazziF,SzydloRM,CrossNCPetal.Durabilityofresponsesfollowingdonorlymphocyteinfusionsforpatientswhorelapseafterallogeneicstemcelltransplantationforchronicmyeloidleukemia.Blood2000;96:2712–2716.
9PorterDL,CollinsRH,HardyCetal.Treatmentofrelapsedleukemiaafterunrelateddonormarrowtransplantationwithunrelateddonorleukocyteinfusions.Blood2000;95:1214–1221.
10CarlensS,RembergerM,AschamJ,Ringde
nO.Theroleofdiseasestageintheresponsetodonorlymphocyteinfusionsastreatmentforleukemicrelapse.BiolBloodMarrowTransplant2001;7:31–38.
11LevineJE,BraunT,PenzaSLetal.Prospectivetrialofchemotherapyanddonorleukocyteinfusionsforrelapseofadvancedmyeloidmalignanciesafterallogeneicstemcelltransplant.JClinOncol2000;20:405–412.
12KantarjianHM,O’BrienS,CortesJEetal.Imatinibmesylatetherapyforrelapseafterallogeneicstemcelltransplantationforchronicmyelogenousleukemia.Blood2000;100:1590–1595.13BarrettAJ,LocatelliF,TreleavenJGetal.Secondtransplantsforleukemicrelapseafterbonemarrowtransplantation:highearlymortalitybutfavorableeffectofchronicGVHDoncontinuedremission.BrJHaematol1991;79:567–574.
14WagnerE,VogelsangGB,ZehnbauerBAetal.Relapseofleukemiaafterbonemarrowtransplantation:effectofsecondmyeloabalativetherapy.BoneMarrowTransplant1992;9:205–209.
OutcomesaftersecondallogeneictransplantationMEapenetal72715RadichJP,SandersJE,BucknerCDetal.Secondallogeneictransplantationforpatientswithrecurrentleukemiaafterinitialtransplantationwithtotalbodyirradiation-containingregimen.JClinOncol1993;11:304–313.
16MartinoR,BadellI,BrunetSetal.Secondbonemarrowtransplantationforleukemiainuntreatedrelapse.BoneMarrowTransplant1994;14:589–593.
17ChaingK-Y,WeisdorfDJ,DaviesSMetal.Outcomeofsecondbonemarrowtransplantationfollowingauniformconditioningregimenastherapyformalignantrelapse.BoneMarrowTransplant1996;17:39–42.
18BosiA,BacciS,MinieroRetal.Secondallogeneicbonemarrowtransplantationinacuteleukemia:amulticenterstudyfromtheGruppoItalianoTrapiantoDiMidolloOsseo(GITMO).Leukemia1997;11:420–424.
19KishiK,TakahashiS,GondoHetal.Secondallogeneicbonemarrowtransplantationforpost-transplantleukemiarelapse:resultsofasurveyof66casesin24Japaneseinstitutes.BoneMarrowTransplant1997;19:461–466.
20MichalletM,TanguyML,SocieGetal.Secondallogeneicheamtopoietictransplantationinrelapsedacuteandchronicleukemiasforpatientswhounderwentafirstallogeneicbonemarrowtransplantation:asurveyoftheSocieteFrancaisedeGreffedeMoelle(SFGM).BrJHaematol2000;108:400–407.21TomonariA,IsekiT,OoiJetal.Secondallogeneichematopoieticstemcelltransplantationforleukemiarelapseafterfirstallogeneictransplantation:outcomeof16patientsinasingleinstitution.IntJHematol2002;75:318–323.
22GlucksbergH,StorbR,FeferAetal.Clinicalmanifestationsofgraft-versus-hostdiseaseinhumanrecipientsofmarrowfromHLA-identicalsiblingdonors.Transplantation1974;18:295–304.
23PrzepiorkaD,WeisdorfD,MartinPetal.ConsensusconferenceonacuteGVHDgrading.BoneMarrowTransplant1995;15:825–828.
24AtkinsonK,HorowitzMM,GaleRPetal.Riskfactorsforchronicgraft-versus-hostdiseaseafterHLA-identicalsiblingbonemarrowtransplantation.Blood1990;75:2459–2464.25KleinJP,MoeschbergerML.SurvivalAnalysis:TechniquesofCensoredandTruncatedData.Springer-Verlag:NewYork,NY,2003.
26CoxDR.Regressionmodelsandlifetables.JRStatSocSerB1972;34:187–202.
27AndersenPK,KleinJP,ZhangM-J.Testingforcentereffectsinmulti-centersurvivalstudies:aMonteCarlocomparisonoffixedandrandomeffectstests.StatMed1999;18:1489–1500.28ClarkTG,AltmanDG,DeStavolaBL.Quantificationofthecompletenessoffollowup.Lancet2001;359:1309–1310.
29GiraltS,EsteyE,Albitaretal.Engraftmentofallogeneicprogenitorcellswithpurineanalog-containingchemotherapy:harnessinggraft-versus-leukemiawithoutmyeloablativether-apy.Blood1997;89:4531–4536.
30PawsonR,PotterMN,TheocharousPetal.Treatmentofrelapseafterallogeneicbonemarrowtransplantationwithreducedintensityconditioning(FLAG7Ida)andsecondallogeneicstemcelltransplant.BrJHaematol2001;115:622–629.
31ChamplinRE,SchmitzN,HorowitzMMetal.Bloodstemcellscomparedwithbonemarrowasasourceofhematopoieticcellsforallogeneictransplantation.Blood2000;95:3702–3709.32BensingerWI,MartinPJ,StorerBetal.TransplantationofbonemarrowascomparedwithperipheralbloodcellsfromHLA-identicalrelativesinpatientswithhematologiccancers.NEnglJMed2001;344:175–181.
BoneMarrowTransplantation
因篇幅问题不能全部显示，请点此查看更多更全内容