thicknessmeizhai

RESEARCHOpenAccessClusteredprecursorsinbonemarrowsectionspredictearlyrelapseinpatientswithacutemyeloidleukemiawithinhematologicremissionYehuaYu,ZhentianWu,JingZhang,YuanmeiZhai,YinghuaYuan,SihongLiu,HuiWangandJunShi*AbstractBackground:Bonemarrow(BM)aspirationislargelyusedforrelapseassessmentinacutemyeloidleukemia(AML).
ItremainsunclearwhatrolesthatBMtrephinebiopsyplaysonrelapseassessment.
Methods:Bonemarrow(BM)sectionsduringcompleteremission(CR)from60acutemyeloidleukemia(AML)patientswereretrospectivelyanalyzed.
Computerimageprocessingtechnologywasperformedfordetectionofthedistancebetweenprecursorsandendosteum,anddensityofprecursorswasalsocalculatedunderlightmicroscopicimage.
Immunohistochemistrywasusedtoidentifytheimmunophenotypeofclusteredprecursors.
Results:Exceptforsingleanddoubleprecursors,thereexistedclusteredprecursorsof3-5cellsduringCR.
Here,wedemonstratedthatclusteredprecursors,butnotsingleanddoubleprecursors,wereusefulinriskfactorofrelapse.
Areaunderthereceivingoperatorcurve(ROC)wasof0.
007(CI95%,from0.
572to0.
851).
Usingastandardcut-offvalueof>4.
0/mm2forclusterdensity,earlyrelapsewasdetectedwithasensitivityof51.
5%andaspecificityof85.
7%.
MultivariateCoxregressionanalysisrevealedthatclusteredprecursorisanindependentriskfactorforearlyrelapse(AdjustedHR:0.
325,95%CI:0.
156-0.
679,p=0.
003).
Conclusions:Cumulatively,clusteredprecursorsinBMsectionsduringCRmayserveasanindependentriskfactorofearlyrelapseandpooroutcomeforAMLpatientsinclusterdensity>4.
0/mm2insections.
Earlyaggressiveinterventionsareneededtopreventhematologicrelapse.
Keywords:Bonemarrow,Biopsy,Relapse,AcutemyeloidleukemiaIntroductionInleukemia,theleadingcauseoftreatmentfailureisdis-easerelapse.
Salvagetreatmentsforrelapsedpatientsfre-quentlyleadtodismaloutcomesincludinglowlevelsofcompleteremission(CR)andshortoverallsurvival(OS)times[1-4].
Hitherto,thetraditionalapproachtoassesstreatmentresponseandfollow-uphasreliedoncountingblastcellsinbonemarrow(BM)smears.
BMsmears,however,mayoccasionallybedilutedbysinusoidalbloodandsubsequentlycannotprovidereliableinformationontreatmentresponseandfailtojudgeearlyrelapse[5].
ItmayevenbeinferiortotheBMimprintfoundintrephinebiopsyfortheevaluationofcellularity,atleastaccordingtoarecentreport[6].
Inaddition,BMaspir-ationonlyprovidesinformationoncellularityandlacksotherinformationsuchascellularlocalizationandmi-croenvironmentalstructure.
BMbiopsysectionsmightcompensatefortheseshortcomings[7].
Similartotheobservationsmadeforcellnumbers,celllocalizationmightalsoprovideimportantinformationapplicabletodiseasediagnosisandperhapsprognosisas-sessment.
Innormalhumansubjects,precursorsarerareandarefoundlocalizedneartheendosteum,andconsistof1-2cells.
Insomecasesofmyelodysplasiasyndrome(MDS),immatureprecursorsmightbelocatedintheintertrabecularregionandoccasionallyaggregateasclusterswhichareof3~5cells,suchclusterswereinitiallydefinedasabnormallocalizationofimmature*Correspondence:shijun7@hotmail.
comEqualcontributorsDepartmentofHematology,ShanghaiJiaoTongUniversityAffiliatedSixthPeople'sHospital,Shanghai,China2014Yuetal.
;licenseeBioMedCentralLtd.
ThisisanopenaccessarticledistributedunderthetermsoftheCreativeCommonsAttributionLicense(http://creativecommons.
org/licenses/by/2.
0),whichpermitsunrestricteduse,distribution,andreproductioninanymedium,providedtheoriginalworkisproperlycited.
Yuetal.
JournalofTranslationalMedicine2014,12:18http://www.
translational-medicine.
com/content/12/1/18precursors(ALIP)byTricot[8].
ALIPprominentlypre-sentedinhigh-riskMDSpatientsandthedelaypriortotransformingtoacutemyeloidleukemia(AML)wasmuchshorterascomparedwithALIP-negativeMDSpatients[8-10].
Itthussuggestedthatabnormallocalizationofprecursorscouldassistintheassessmentofpatientprognosis.
Inthepresentstudy,wefoundduringCRofAMLpa-tients,thereexistedclusteredprecursorsconsistingof3to5cellsintheintertrabecularregion.
SincetheyweremorphologicallyandanatomicallyanalogoustoALIPinMDS,wedefinedthemasALIP-likeclusters.
Thepre-ciseinformationthathistologicaldataprovidesduringCRislargelyunknown.
Here,weextendedourinvestiga-tionstoexploreahistologicalrelapseindicatorbycom-paringthefrequencyofALIP-likeclustersbetweenrelapseandno-relapsecases.
MaterialsandmethodsPatientsBetweenDecember2004andFebruary2013,115pa-tientswithdenovoAMLwereadmittedtotheHematologyDepartmentofShanghaiSixthPeople'sHospital,China.
PatientswerediagnosedaccordingtotheFABclassification.
CRandrelapsewerediagnosedaccordingtoChesonetal.
[11].
Thosepatientswhodidnotreceivestemcelltransplantationwereenrolledinthisstudy.
Caseswereexcludedfromthestudyifthey:(a)lackedcompleteclinicaldata;(b)presentedwithotherdiseasesthatmightimpactonchemotherapy;or(c)diedbecauseofotherdiseasebutnotofleukemia.
Withthoseexclusioncriteria,60patientswereenrolledinthisstudy.
Forinductiontherapies,AMLpatientsweretreatedwithoneanthracyclineagent(daunorubi-cin,idarubicin,epirubicinormitoxantroneatadailydosageof(40–60)mg/m2,7-8mg/m2,(60–90)or(8–10)mg/m2,respectively)for3daysandcytarabineat(100–150)mg/m2dailyfor7days.
Forconsolidationtherapies,AMLpatientsreceivedupto6cyclesofin-ductiontherapiesincludingonealternativeanthracy-clinewithsimilardosagementionedabove.
Alltrephinebiopsiesandadministrationswereperformedafterhav-ingobtainedinformedconsent.
Thisstudywasap-provedbytheinstitutionalEthicsCommitteesofourhospitalandconductedinaccordancewiththeethicalguidelinesoftheDeclarationofHelsinki.
Firstcheck-uppostinductionwasassessedbybonemarrowbiopsyonday21–28postinduction.
Concerningthephysicalcon-ditionofthepatients,thebiopsiesmightberetardeduntilperipheralbloodrecovery.
DuringCR,BMbiop-sieswereroutinelyperformedbeforeeveryconsolida-tiontherapies.
Afterconsolidationtherapies,bonemarrowwassurveilledbybiopsiesevery3monthsinthefirstyearandevery6monthsinthesecondyear.
BMsectionsandsmearsfromthepatientswereassayedunderpairedanalysisfromthetimeofdiagnosistothefollow-up.
CasesinCRweredividedintorelapseandno-relapsegroupsaccordingtopatientsrelapsedornotduringthefollow-up.
ThepathologistswhoexaminedtheBMsampleswerenotparticipantofthisstudyandwereinnocentaboutthegroupsituation.
Clinicalout-comewasassessedbyrelapse-freesurvival(RFS)andoverallsurvival(OS).
DeterminationofthedistancefromprecursorstoendosteumandprecursordensityonBMsectionsBMbiopsysampleswerecollectedfromAMLpatientsduringCRandfixedinBouinfixative.
Dehydrationwasperformedbyexposingspecimenstoascendingcon-centrationsofethanol.
AllsampleswereembeddedinHemapun865plastic.
Thecomponentoftheplasticandthedetailedembeddingtechniquehadbeenpreviouslydescribed[12].
BMsectionsof3μmthicknesswerestainedwithhaematoxilin-giemsa-acidfuchsin(HGF)forvisualizationofimmatureprecursorcells.
Specimenswereobservedbyusingofopticalmicroscopeimagingsystem(Olympus,Tokyo,Japan).
Precursorsweresub-groupedintosingle,doubleandclusteredcellsthatcon-sistedof≥3precursors.
Foreverycase,sectionswereobservedrandomlyby10fieldsofviewwith*400magni-fication.
Fordetectionoftheprecisedistancebetweenprecursorsandendosteum,computerimageprocessingtechnologywasperformed[13].
Thedistancedetectedbycomputeraspixelwastransferredintoμmbythealgo-rithmof1pixel=0.
2μm.
FordetectionofthedensityofprecursorsfoundinBMsections,ahemocytometerwasused,whereinthe16smallestsquarelatticesweresetas1mm2.
Immunohistochemical(IHC)stainingTostudythecellularityofALIP-likeclusters,IHCstain-ingwasperformed.
Theformalin-fixed,EDTA(PH7.
4)decalcificatedandparaffin-embeddedbonemarrowtis-sueswerecutatasectionthicknessof3μm.
Thestain-ingprocedurewasdoneinaccordwiththeinstructionsprovidedwithanUltraVisionQuantoDetectionSystemHRPDABkit's(ThermoScientific,TL-060-QHD,CA,USA).
Inbrief,sectionsweredeparaffinizedinxyleneandrehydratedbyexposureofthespecimenstogradedethanol.
Antigenretrievalwasperformedbysodiumcit-ratebufferwithpH6.
0at90°Cfor15min.
Mouseanti-humanCD34ClassIImonoclonalantibody(1:100diluted,M7165,Dako,Glostrup,Denmark),rabbitanti-humanCD117polyclonalantibody(1:200diluted,A4502,Dako,CA,USA),andmouseanti-humanmyeloperoxidase(MPO)antibody(1:8diluted,R-0405,Changdao,Shanghai,China)wereincubatedat4°Covernight.
Horseradishperoxidase(HRP)labeledsecondaryantibodyandtheYuetal.
JournalofTranslationalMedicine2014,12:18Page2of7http://www.
translational-medicine.
com/content/12/1/18substratediaminiobenzidine(DAB)wereusedaccordingtothekit'sinstructions.
Thensectionswereviewedunderthelightmicroscope(Olympus,Tokyo,Japan).
StatisticalanalysisMann–WhitneyUtestwasappliedtocomparethedens-ityofprecursorsbetweentherelapseandno-relapsecasesandtheirdistancetoendosteumineverycellgroup.
Chi-squaretestandFisher'sexacttestwereusedforcategoricaldata.
Areceiveroperatingcharacteristiccurve(ROC)analysiswasperformedandthecorre-spondingareaunderthecurve(AUC)wascalculatedtodeterminethevalueoftheprecursordensityinpredict-ingarecurrence.
SurvivalcurveswereconstructedusingtheKaplan–Meiermethodandthelog-ranktestwasusedtotestthedifferenceinRFSandOSbetweencaseswithdifferentdensitiesofALIP-likeclusters.
RFSwasdefinedasthetimebetweendiagnosisandtheonsetofthefirstrelapse,andOSwasdefinedasthetimebe-tweendiagnosisandtheoccurrenceofdeathorwhenlosttofollow-up.
CoxproportionalhazardregressionwasusedtomodeltheRFS.
Thedataweredescribedasmean±standarddeviation(SD),Pvalues≤0.
05wereconsideredstatisticallysignificant.
AllstatisticalanalyseswereconductedbyusingtheSPSS18.
0softwarepro-gram(StatisticalPackageforSocialScience,SPSSInc.
Chicago,IL.
,USA).
ResultsCharacteristicsofthepatientsAgesofpatientsintherelapsegroupandno-relapsegroupwere(51.
85±17.
96)yearsand(49.
86±16.
96)yearsrespectively.
Theratiosofmale/femaleinthetwogroupswere24/15and10/11respectively(Table1).
Therewerenostatisticaldifferencesinage,sexandFABsubtypesbetweenthetwogroups.
CellularityandlocalizationofprecursorsinBMhistologcalsectionsduringCRofAMLDuringCR,thereexitedsingle,doubleandclusteredprecursorsinBMsections(Figure1A).
Theexactdis-tancesfromthecenteroftheprecursorsorclusterstotheendosteumweredetectedbycomputerprocessingimage.
Clusteredprecursorswerelocatedfurtherthansingleprecursorsanddoubleprecursors(P4.
0/mm2)andhematologicrelapsewerecompared.
Asaresult,themediantimefromhistologicrelapsetohematologicrelapsewas8months(Figure2C).
Add-itionally,theprobabilityofRFSandOSwerecomparedbetweencaseswithclusterdensity>4.
0/mm2andthoselessthan4.
0/mm2.
AsshowninFigure3A-B,therewasapoorerprognosisinRFS(P=0.
0011)andOS(P=0.
045)forcaseswithclusterdensity>4.
0/mm2.
PrognosticassessmentofknownclinicalparametersIntheunivariateandmultivariateCoxregressionmodel,onlythefactorofclusterdensity>4.
0/mm2wasassoci-atedwithhigherrelapserisk(P=0.
002and0.
003,re-spectivelyTable3).
Othercandidateprognosticfactorsincludingbonemarrowprecursors,peripheralwhitebloodcellsandage,werenotassociatedwithhigherre-lapseriskinbothunivariateandmultivariatemodel(Table3).
Table1PatientdemographicsandclinicalcharacteristicsCharacteristicNo-relapsegroup(n=21)Relapsegroup(n=39)PvalueAge(years)0.
678Mean±SD49.
86±16.
9651.
85±17.
96Range20-8016-77Sex(male/female)10/1124/150.
414FABsubtype--0.
885M104M2814M358M446M524M611M712FAB:French-American-BritishClassification.
Yuetal.
JournalofTranslationalMedicine2014,12:18Page3of7http://www.
translational-medicine.
com/content/12/1/18ALIP-likeclusterswerecomposedofimmaturemyeloidprecursorsToclarifywhetherALIP-likeclusterwerecomposedofimmaturemyeloidprecursors,assayofstainingforMPOwasselectedtoconfirmthemyeloidlineage,andCD34andCD117werechosenforstainingimmatureprecur-sors.
Asaresult,allALIP-likeclusterswereMPOposi-tive.
However,theexpressionofCD34andCD117wasshowntobevariable.
AsshowninFigure4A,precursorsinsomeALIP-likeclustersexpressedbothCD34andABP=0.
0002P=0.
468P4/mm2Clusterdensity4/mm2Clusterdensity4/mm2)indicatedpoorersurvival.
Contrasttocaseswith4/mm2)carriedpooreroutcomeinbothRFS(A)andOS(B).
Yuetal.
JournalofTranslationalMedicine2014,12:18Page5of7http://www.
translational-medicine.
com/content/12/1/18afterchemotherapyinanAMLmicemodel[16].
Inthissetting,wehypothesizedthatALIP-likeclustersevolvedfromLSCharboringatsitesneartheendosteum,whileunderconditionsofadministrationretrieval,LSCmightexhibitdifferentiationundertheinfluenceofsomecyto-kineregulatorssuchasvascularendothelialcellgrowthfactor[17-19]andinsulin-likegrowthfactor[20].
Underthosesituations,theprogenycellsmigratetowardtheintertrabecularregion[21].
Inthecontextofthishy-pothesis,ALIP-likeclustersmightbeheterogeneousandhierarchicalwithmorematurepopulationlocatedininter-trabecularregion.
Contrarily,moreimmaturepopulationmightbelocatedclosertotheendostreum.
IHCstainingindicatedthattheprogenycellssharedvariableimmuno-phenotype,whereinsomesub-populationswereCD34/CD117positive,whiletheothersarenegative.
Theseob-servationssupportedthatanalogouswithLSC,thedaugh-tercellswerealsoheterogeneousandhierarchicalinbehavior.
Inthiscurrentstudy,weproposethatatavalueofgreaterthan4.
0/mm2ofALIP-likeclusterdensityinBMsectionswithinhematologicremissionisastronginde-pendentprognosticfactorforAMLafterachievingacompleteremission,andconsideritasanindicationofhistologicrelapseandearlyinterventionsmightbeconsideredasapre-requisiteinhistologicalrelapsetoimproveeventualclinicaloutcomes.
Here,wesuggestBMbiopsiesshouldbeconsideredwhileotherminimalresidualdiseasemeasurementssuchasreversetranscription-polymerasechainreaction(RT-PCR),flowcytometer,etc.
arenotavailable.
Nonetheless,therearesomelimitationsinthisstudy.
Firstly,cytogeneticandmolecularmarkeras-saysarenotroutinelyperformedatthishospital,compari-sonandcombinationwithotherMRDmeasurementssuchasRT-PCR,flowcytometer,etc.
[22,23]arelacking.
Secondly,wecouldnotexcludetheinfluenceofcyto-geneticsubtypesonrelapseandclinicaloutcome.
AfurtherprospectiveblindedstudyisrequiredtoassessTable3CoxregressionanalysisofRFSinAMLpatientsduringCRVariablesUnivariateanalysisMultivariateanalysisHR95%CIPvalueAdjustedHR95%CIPvalueALIP-likecluster*0.
3530.
183-0.
6800.
0020.
3250.
156-0.
6790.
003BMP1.
3810.
408-4.
6700.
6042.
3580.
598-9.
2900.
220PWBC1.
0111.
000-1.
0230.
0511.
0080.
997-1.
0200.
164Age1.
0080.
990-1.
0260.
4081.
0010.
980-1.
0220.
944Abbreviations:BMPBonemarrowprecursors,HRhazardratio,CIconfidenceinterval,PWBCperipheralwhitebloodcell.
*ALIP-likeclusterdensity>4/mm2versus<4/mm2.
MPOCD34CD117CD34CD117MPOABFigure4ALIP-likeclustersareofheterogeneousmyeloidprecursors.
AllALIP-likeclustersexpressedMPO.
SomeclustersexpressedbothCD34andCD117(A).
However,otherclustersexpressedneitherCD34norCD117(B).
Sectionswereobservedat*1000magnification.
ArrowsindicatedALIP-likeclusters.
Yuetal.
JournalofTranslationalMedicine2014,12:18Page6of7http://www.
translational-medicine.
com/content/12/1/18theclinicalsignificanceofthecut-offvaluesinre-lapseprediction.
ConclusionTakentogether,inthisretrospectivestudy,wefoundduringCRofAMLpatients,exceptforsingleanddoubleprecursors,thereexistedclusteredprecursorsof3~5cells.
TheclusteredprecursorsaremorphologicallyandanatomicallyanalogoustoALIPinMDS.
SimilarwiththeprognosticvalueofALIPinMDS,theseALIP-likeclustersinBMsectionsduringCRofAMLcouldalsobeusedasanindependentpredictorofearlyrelapseandpooroutcomeforAMLpatients.
Finally,weproposedthatwhileALIP-likeclusterdensityincreasedtooverthethresholdof4.
0/mm2,althoughinhematologicalre-sponse,urgentinterferenceshouldbeneededtopreventearlyrecurrenceandconsequentlyimprovetheOS.
CompetinginterestsTheauthorsdeclarethattheyhavenocompetinginterests.
Authors'contributionsYYperformedlaboratorytestsandcollectedpatientdataandanalyzedthedata.
ZWanalyzedthepatientdataandwrotethepaper.
JZ,YY,YZ,SLandHWperformedthelaboratorytests.
JSprovidedpatientdata,analyzedthedataandwrotethepaper.
Allauthorsreadandapprovedthefinalversionofthesubmittedmanuscript.
AcknowledgementsWethankLingLiandGui-TaoCao(HuadongNormalUniversity,Shanghai,China)forhelpinexploitingsoftwarefordistancedetection.
ThisworkwassupportedbytheNationalNaturalScienceFoundationofChina(GrantNo.
81170507)andtheShanghaiCommitteeofScienceandTechnology,China(GrantNo.
11140903700andNo.
12142201200).
Received:16August2013Accepted:17January2014Published:22January2014References1.
ZwaanCM,ReinhardtD,ZimmermanM,HasleH,StaryJ,StarkB,DworzakM,CreutzigU,KaspersGJ,InternationalBFMStudyGrouponPaediatricAML:Salvagetreatmentforchildrenwithrefractoryfirstorsecondrelapseofacutemyeloidleukaemiawithgemtuzumabozogamicin:resultsofaphaseIIstudy.
BrJHaematol2010,148:768–776.
2.
SanderA,ZimmermannM,DworzakM,FleischhackG,vonNeuhoffC,ReinhardtD,KaspersGJ,CreutzigU:ConsequentandintensifiedrelapsetherapyimprovedsurvivalinpediatricAML:resultsofrelapsetreatmentin379patientsofthreeconsecutiveAML-BFMtrials.
Leukemia2010,24:1422–1428.
3.
KurosawaS,YamaguchiT,MiyawakiS,UchidaN,SakuraT,KanamoriH,UsukiK,YamashitaT,OkoshiY,ShibayamaH,NakamaeH,MawatariM,HatanakaK,SunamiK,ShimoyamaM,FujishimaN,MaedaY,MiuraI,TakaueY,FukudaT:Prognosticfactorsandoutcomesofadultpatientswithacutemyeloidleukemiaafterfirstrelapse.
Haematologica2010,95:1857–1864.
4.
UhlmanDL,BloomfieldCD,HurdDD,PetersonBA:Prognosticfactorsatrelapseforadultswithacutemyeloidleukemia.
AmJHematol1990,33:110–116.
5.
HelgestadJ,RosthjS,JohansenP,VarmingK,stergaardE:Bonemarrowaspirationtechniquemayhaveanimpactontherapystratificationinchildrenwithacutelymphoblasticleukaemia.
PediatrBloodCancer2011,57:224–226.
6.
GongX,LuX,WuX,XuR,TangQ,XuG,WangL,ZhangX,ZhaoX:Roleofbonemarrowimprintsinhaematologicaldiagnosis:adetailedstudyof3781cases.
Cytopathology2012,23:86–95.
7.
IslamA:Proposalforaclassificationofacutemyeloidleukaemiabasedonplastic-embeddedbonemarrowbiopsysections.
LeukRes1993,17:421–427.
8.
TricotG,DeWolf-PeetersC,VlietinckR,VerwilghenRL:Bonemarrowhistologyinmyelodysplasticsyndromes.
II.
PrognosticvalueofabnormallocalizationofimmatureprecursorsinMDS.
BrJHaematol1984,58:217–225.
9.
VerburghE,AchtenR,MaesB,HagemeijerA,BoogaertsM,DeWolf-PeetersC,VerhoefG:Additionalprognosticvalueofbonemarrowhistologyinpatientssubclassifiedaccordingtotheinternationalprognosticscoringsystemformyelodysplasticsyndromes.
JClinOncol2003,21:273–282.
10.
SaadST,VassalloJ,ArrudaVA,Lorand-MetzeI:Theroleofbonemarrowstudyindiagnosisandprognosisofmyelodysplasticsyndrome.
Pathologica1994,86:47–51.
11.
ChesonBD,BennettJM,KopeckyKJ,BüchnerT,WillmanCL,EsteyEH,SchifferCA,DoehnerH,TallmanMS,ListerTA,Lo-CocoF,WillemzeR,BiondiA,Hidde-mannW,LarsonRA,LwenbergB,SanzMA,HeadDR,OhnoR,BloomfieldCD,InternationalWorkingGroupforDiagnosis,StandardizationofResponseCriteria,TreatmentOutcomes,andReportingStandardsforTherapeuticTrialsinAcuteMyeloidLeukemia:Revisedrecommendationsoftheinternationalworkinggroupfordiagnosis,standardizationofresponsecriteria,treatmentoutcomes,andreportingstandardsfortherapeutictrialsinacutemyeloidleukemia.
JClinOncol2003,21:4642–4649.
12.
XiaoL,LuxiS,YingT,YizhiL,LingyunW,QuanP:Diagnosisofunknownnonhematologicaltumorsbybonemarrowbiopsy:aretrospectiveanalysisof10,112samples.
JCancerResClinOncol2009,135:687–693.
13.
WuH,CaoG,LiL,ShiJ,InternationalConferenceonMultimediaTechnology:ICMT2010:IdentifyingImmaturePrecursorCellsinBoneMarrowPathologicalImagesBasedonDistanceTransformandInternalStructuresofCells.
Ningbo:IEEEXplore;2010.
14.
VoulgarelisM,GiannouliS,TasidouA,AnagnostouD,ZiakasPD,TzioufasAG:Bonemarrowhistologicalfindingsinsystemiclupuserythematosuswithhematologicabnormalities:aclinicopathologicalstudy.
AmJHematol2006,81:590–597.
15.
LoCelsoC,FlemingHE,WuJW,ZhaoCX,Miake-LyeS,FujisakiJ,CtéD,RoweDW,LinCP,ScaddenDT:Live-animaltrackingofindividualhaematopoieticstem/progenitorcellsintheirniche.
Nature2009,457:92–96.
16.
IshikawaF,YoshidaS,SaitoY,HijikataA,KitamuraH,TanakaS,NakamuraR,TanakaT,TomiyamaH,SaitoN,FukataM,MiyamotoT,LyonsB,OhshimaK,UchidaN,TaniguchiS,OharaO,AkashiK,HaradaM,ShultzLD:Chemotherapy-resistanthumanAMLstemcellshometoandengraftwithinthebone-marrowendostealregion.
NatBiotechnol2007,25:1315–1321.
17.
BellamyWT,RichterL,SirjaniD,RoxasC,Glinsmann-GibsonB,FrutigerY,GroganTM,ListAF:Vascularendothelialcellgrowthfactorisanautocrinepromoterofabnormallocalizedimmaturemyeloidprecursorsandleukemiaprogenitorformationinmyelodysplasticsyndromes.
Blood2001,97:1427–1434.
18.
CasalouC,FragosoR,NunesJF,DiasS:VEGF/PLGFinducesleukemiacellmigrationviaP38/ERK1/2kinasepathway,resultinginrhoGTPasesactivationandcaveolaeformation.
Leukemia2007,21:1590–1594.
19.
DiasS,HattoriK,ZhuZ,HeissigB,ChoyM,LaneW,WuY,ChadburnA,HyjekE,GillM,HicklinDJ,WitteL,MooreMA,RafiiS:AutocrinestimulationofVEGFR-2activateshumanleukemiccellgrowthandmigration.
JClinInvest2000,106:511–521.
20.
ChapuisN,TamburiniJ,Cornillet-LefebvreP,GillotL,BardetV,WillemsL,ParkS,GreenAS,IfrahN,DreyfusF,MayeuxP,LacombeC,BouscaryD:AutocrineIGF-1/IGF-1RsignalingisresponsibleforconstitutivePI3K/Aktactivationinacutemyeloidleukemia:therapeuticvalueofneutralizinganti-IGF-1Rantibody.
Haematologica2010,95:415–423.
21.
WuZ,YuY,ZhangJ,ZhaiY,TaoY,ShiJ:Clusteredimmaturemyeloidprecursorsinintertrabecularregionduringremissionevolvefromleukemiastemcellnearendosteumandcontributetodiseaserelapseinacutemyeloidleukemia.
MedHypotheses2013,80:624–628.
22.
PaiettaE:Minimalresidualdiseaseinacutemyeloidleukemia:comingofage.
HematologyAmSocHematolEducProgram2012,2012:35–42.
23.
HouriganCS,KarpJE:Minimalresidualdiseaseinacutemyeloidleukaemia.
NatRevClinOncol2013,10:460–471.
doi:10.
1186/1479-5876-12-18Citethisarticleas:Yuetal.
:Clusteredprecursorsinbonemarrowsectionspredictearlyrelapseinpatientswithacutemyeloidleukemiawithinhematologicremission.
JournalofTranslationalMedicine201412:18.
Yuetal.
JournalofTranslationalMedicine2014,12:18Page7of7http://www.
translational-medicine.
com/content/12/1/18