E－vPalWuaItionofthePathologicalGradeinGliomas:ApplicationofDSC

E－vPalWuaItionofthePathologicalGradeinGliomas:ApplicationofDSC

WangYing１QIJianping２

WangYing１QIJianping２１,２DepartmentofRadiology,TonjiHospital,TongjiMedicalCollegeofHuazhongUniversityofScienceandTechnology,WuＧhan４３００３２,China;１DepartmentofRadiology,WuhangeneralhospitalofGuangzhoucommand,Wuhan４３００７０,ChinaCommunicationauthor:QIJianping,Email:qijp２k０１＠１６３．comFinancialsupportisfromNaturalScienceFoundationinChina’sHubeiprovincialscienceandtechnologyprogram(Numbered２０１２FFB０６８０９)【Abstract】objectiveAssessmentofDSC－PWIforgliomaspreoperativevalueofhistologicalgrade．MaterialsandMethods７６casesgliomasconfirmedbysurgeryandpathologywereanalyzed,３６casesoflow－grade(gradeI~II),４０casesofhigh－grade(gradeIII~IV),allcasesunderwentconventionalMRIplanescan,contrastMRIscanandDSC－PWI．MaximumrCBVvaluesofsolidpart(thesolidparttothecorrespondingcontralateralnormalwhitematterratio)arecacuＧlated．ComparisonofmaximumrCBVvaluesoftumordifferentlevelswasassessedbyttest．DrawingROCcurvetodeterminetheidealvalueofrCBVthresholdbeＧtweenhigh－gradeandlow－gradegliomas．ResultsThemaximumvalueofrCBVinthesolidpartofgliomasarerelatedpositivelywithpathologicallevels,themaximumrCBVvaluesofhigh－gradeandlow－gradegliomaswere０．８７４４±０．９３５、２．０４１５±１．５３１１respectively,significantdifferencewereshowedamongallgradesofgliomas(P＝０．００２)．Toidentifythehighlevelandlow－gradegliomas,１．３５astheidealthreshold,theAUCvalue０．７３９４,thesensitivityandspecificitywere６６．７％,７３．０％．ConclusionrCBVexaminationwillbeabletoassesstheleveloftumorpathology．PWIexaminationforpreoperativeevaluationofgliomashavea【Kheiyghweorrdvsal】ueofhistologicalgrade．glioma;DSC－PWI;rCBV【中图分类号】R４７【文献标识码】B【文章编号】１００８－６３１５(２０１５)１２－１２７９－０２Ingliomapathologicaldiagnosis,themicrovesseldensityoftumorisoneofmainstandardsforevaluationofpathologicalgrade[１,２]．Present,mostresearchesshowthatCBVisthemostvaluableparameterforevaluationofpaＧthologygrading．Therestparametershavelimitedvalue,thecorrelationofwhichwithgliomapathologygradingisn’tasgoodasCBV．However,mostreＧsearchesstillhaveproblems,suchaslimitedsamplequantity,lowcrediblestaＧtisticalresultsandsoon．Becauseofthis,weperformedtheresearchongliomapathologygradingbyevaluationofPWIexamination．１Materialandmethod１．１ResearchobjectSep．２００６toMar．２０１３,WuhangeneralhospitalofGuanＧgzhoucommandhasdoneDSC－PWIscantoalltheconsideredgliomapatientsthroughCTorMRIexamination．ThisprospectiveexaminationobtainedtheapＧprovalofhospitalacademiccommitteeandthepatients’informedconsent．Inalltheexaminations,gliomacasesconfirmedbypostoperativepathologyweretotal７６,astheanalysisobjectoftheresearch．Amongthem,therewere４６males,３０females,agebetween８－７２,averageageat４２．９３．Amongthe７６cases,７２caseswerethefirstepisodepatients,andanother４caseswerethepatientswhohadpostoperativerecurrenceafterhalfayearto１０years,whoseclinicalmanifestaＧtionsmainlywereheadache,nausea,vomit,megrim,limbactivitydisorder,epiＧlepsy,affectivedisorderandmemoryloss,etc．AccordingtoWHOclassification,amongthe７６cases,３６caseswerelow－level(I－II),ofwhich４caseswereIand３２caseswereII,and４０caseswerehigh－level(III－IV),ofwhich２５caseswereIIIand１５caseswereIV．１．２MRimagingequipmentandexaminationtechniqueAdoptedSignaHDxandHDe１．５TMRIfromAmericaGECompanytodoexamination,andalltheconventionalMRI,DSC－PWIscanadoptedhead８－channelphasedarraycoil．MRIplainscan:scanningsequencewasOSag－T１FLAIR,OAxi－T２WI,OAxi－T１FLAIR,OAxi－T２FLAIR．MRIenhancementscan:scanningsequenceinＧcludedOAxi－fsT１WI,OSag－fsT１WI,OCor－fsT１WI．DSC－PWIadoptedGRE－EPI:TE４０ms;TR２０００ms;rollangle９０°;vision２４;slicethickness６mm;interlamellarspacing２mm;excitationnumber１．Contrastagentwas４６９．０１mg/１５mlofgadopentetatemeglumine(magnevist),withtotalinjectionvolume０．２mmol/Kg,double－syringepowerinjectorelbowveinmassinjectionandinjectionrateat３ml/s．１．３Imagepost－processingTransmittedalltheDSC－PWIscanningdatatoGECompany’sADW４．４workstationforpost－processingbyitsbuilt－inFunctoolsoftwarepackage,andadjustedCBVvaluestobecolordrawings．ChosethemaximumCBVvalue’svoxelintumorsolidareaasROI,andpaidattentiontoeliminatethetumornecrosissacvariablearea．ThemaximumrCBVvalue＝themaximumCBVvalueoftumorsolidpart/brainparenchymaCBVvalueofcontralateralsymmetrypart．１．４StatisticalanalysisStatisticaltreatmentadoptedSAS９statisticalanalysissoftware．Thecomparisonbetweentwogroupsadoptedtcheckanalysis．DrewROCcurveandevaluatedcriteriabyusingareaunderthecurve．RelevantanalysisadoptedSpearmanrankcorrelation．Statisticalinspectionlevelɑ＝０．０５．２Result２．１ThecomparisonofmaximumrCBVvaluesofgliomasoildpartindifferentpathologicalgrades．ThevariationrangeofthemaximumrCBVvaluesoflow－level(I~II)gliomas(３６cases)solidpartwas０．５６２６to１．１８６１,average０．８７４４±０．９３５,seeingfigure１;ThevariationrangeofthemaximumrCBVvaluesofhigh－level(III~IV)gliomas(４０cases)solidpartwas１．５４５２to２．５３７８,average２．０４１５±１．５３１１,seeingfigure２．(tvalue３．９４,pvalue０．００２)Figure１female,４０yearsold,rightparietallobegradeIIdiffuseastrocytoＧma．AxialviewT１WI(drawinga);axialviewT２WI(drawingb);axialviewenＧhancementT１WI(drawingc);axialviewCBVdrawing(drawingd);axialviewaverageenhancementtimeMTEdrawing(drawinge);time－signalstrengthcurveTIC(drawingf)．Showirregularsolidmassinrightparietallobe,presenＧtingslightlylongT１slightlylongT２signal,containingspeckledlongT１longT２signal,slightlinearenhancement．LesionareainPWI－CBVdiagrampresentslightblue,black,indicatingCBVvalueislowerthancontralateralnorＧmalbrainparenchyma,rCBVvalue０．８１１．Figure２female,４４yearsold,leftfrontallobeanaplasticoligoastrocytomas．AxialviewT２WI(drawinga);axialviewenhancementT１WI(drawingb);DWIdrawing(drawingc),CBVdrawing(drawingd),CBFdrawing(drawinge),time－signalstrengthcurveTICdrawing(drawingf)．Showirregularcysticsolidmassinleftfrontallobe,presentingslightlylong,longT２signal,moderategarland－likeenhancement．DWIdrawingpresentsslightlyhighsignal．LesionareainPWI－CBVdiagrampresentsgreen－yellow－red,indicatingCBVvalueisobviouslyhigherthancontralateralnormalbrainparenchyma,rCBVvalue４．３３．Bystatisticaltcheckanalysis,themaximumrCBVvaluesofhigh－levelandlow－levelgliomasolidparthadsignificantdifference(P＝０．００２)(table１)．ThroughcorrelationanalysisofSpearman,gliomapathologicalgradeandthemaximumrCBVvalueoftumorsolidparthadpositivecorrelation(correlationcoefficientR＝０．５１７５９)．TheareaunderROCcurve(figure３),AUCvalue,is０．７３９４．Table１ThecomparisonofmaximumrCBVvaluesofgliomasoildpartindifferentpathologicalgrades．TherCBVvaluesofgliomasoildpartAverageLowerlimitUpperlimitStandarderrorPvalueLow－level０．８７４４０．５６２６１．１８６１０．９３５High－level２．０４１５１．５４５２２．５３７８１．５３１１０．００２Figure３ROCcurveofthemaximumrCBVvaluesofgliomasolidpart２．２ROCcurveanalysisofgliomapathologygradebytheevaluationoftumorsolidpart’smaximumrCBVvalueWithgliomapathologygradingasobservedobject,usingROCcurveanalysisoftumorsolidpart’smaximumrCBVtoidenＧtifythediagnosticefficiencyofhigh－levelandlow－levelglioma,obtainedareaundercurveAUCwas０．７３９４,showingthemaximumrCBVvaluesoftumorsolidparthadgooddiagnosticefficiencytotheidentificationofhigh－levelandlow－levelglioma(figure３)．Theidealidentifieddiagnosticthresholdoftumorsolidpart’smaximumrCBVvalueswas１．３５hours,andtheobtainedsensitivitywas０．６６６６７,specificity０．７２９７３．Table２listedthesensitivityandspecificityofgliomapathologicalgradebytheevaluationoftumorsolidpart’smaximumrCBVvalue．Table２Thesensitivityandspecificityofgliomapathologygradebytheevaluationoftumorsolidpart’smaximumrCBVvalueThemaximumrCBVvalueoftumorsolidpartSensitivitySpecificityAreaSEP９５％CILowerlimitUpperlimit１．３５０．６６６６７０．７２９７３０．７３９４０．０５７０．０００．６２７７０．８５１２３Discussions３．１ThefundamentalsofcheckandevaluationofgliomabyPWIIngliomapathologicalresearch[２],couldevaluategliomapathologicalgradebymicrovesseldensityandendotheliosisdegree,furthermoreobservedatypicalnucleus,mitosis,necrosis,etc．Tumorbloodvesselstraveledtortuously,thediameterofwhichcouldachieve４０um,andendotheliathicknesswasabout０．５um．ThediＧameterofnormalbraintissuebloodcapillarywasonly３－５um,andbloodvesselendotheliumthicknesswasabout０．２６um．ThegliomainearlyphasewasinhyＧpoperfusionandanoxicconditions,afterwardshypoxiainduciblefactorlɑinＧcreasedexpression,furtheradjustedandcontrolVEGFtoincreaseexpression,andVEGFstimulatedtheformationofimmaturetumorvessels[３]．Astumorgrowed,thenormalmicrovesselsintumorsurroundingtissuesfailintoittobeＧcometumorvessels．Astumorvesseldensityincreased,thegliomavesselvolumeincreased．Thehigherthegliomalevelwas,itsbloodvolumewashigher,andthusthegliomalevelofmalignancycouldbejudgedbythemeasureＧmentofbloodvolume．Thedocumentsshowthatthegliomagrademalignancyandprognosisarecloselyrelatedwithvascularizationdegree[４]．Theinvasivegrowthofgliomareliesonthenewvascularizationintumor．Thecountofmicrovesseldensity(MVD)isthequantitativeindexreflectingthetumornewvascularizationdegree．WuHonglinandothers[５]hashadcontrolstudyofCBVandtumorMVDonglioma,whichshowedthebothhadgoodcorrelation,andthoughtthatthevesseldiameterandquantityintumorcanleadtotheriseofCBVvalue．AccordＧingtothereportfromLawandothers[６],CBVobviouslyhasdirectiveproportiontogliomagrade(SpearmancoefficientofrankcorrelationR＝０．８１７)．rCBVistherelativeratiowhichputunaffectedcontralateralnormalbrainwhitematterCBVvalueasinnerreferenceobject．ChoosingrCBVnotCBVistoeliminatetheharmfuleffectcausedbythecerebralmiddlearterypositionerrorwhenchoosing,andpartiallycorrectstheerror[４]causedbyrecycleandcontrastagentleaking．３．２TheestimatedvalueofDSC－PWIexaminationtogliomapathologygradＧingThisarticlehasdoneresearchon７６casesofgliomas,usingtumorsolidpart’smaximumrCBVvalueregionevaluatingmethodtoevaluateDSC－PWIexaminationtogliomapathologygrading．Theresearchresultshowed:lowlevel(I~II)gliomasolidpart’smaximumrCBVvaluewas０．８７４４±０．９３５;high－level(III~IV)gliomasolidpart’srCBVvaluewas２．０４１５±１．５３１１,aＧmongthemthereweresignificantdifference(P＝０．００２)．BySpearmancorrelaＧtionanalysis,gliomapathologicalgradehadpositivecorrelationtotumorsolidpart’smaximumrCBVvalue,hintingthatDSC－PWIexaminationbymeasureＧmentoftumorsolidpart’smaximumrCBVvaluecontributedtoidentifylowlevelandhigh－levelglioma．Meanwhile,usingROCcurvetoanalyzetumorsolidpart’smaximumrCBVvaluefortheidentificationofdiagnosisefficacyoflow－level(I~II)andhigh－level(III~IV)gliomas．Theobtainedareaundercurve(AUC)was０．７３９４,furtherexplainedthattumorsolidpart’smaximumrCBVvaluehadquitehighdiagnosisefficacyfortheidentificationoflow－levelandhigh－levelgliomas．Whentheidealthresholdvaluewas１．３５,theobtainedsensitivitywas６６．７％,andthespecificitywas７３．０％．Theresearchresultofthisarticlewassimilarwiththemostliteratures,asKnoppandothers[７]hasresearchedfindingthattheaveragemaximumrCBVvalueoflow－levelgliomawas１．４４,whiletheaveragemaximumrCBVvalueofhigh－levelgliomawas５．０７,andtherewereobviousdifferencebetweenthetwo;Levandothers[８]thoughtthatthemaximumrCBVvalueofastrocytomahadsignificantpositivecorrelationtopathologicalgrade,andpointedthatoligoＧdendrogliomawouldlowerdowntheaccuracyofgliomagradeevaluatedbymaximumrCBVvalue,becausepartialoligodendrogliomaswerelocatedoncereＧbralcortexarea[９]．ThegliomamaximumrCBVvaluesofdifferentpathologicalgradesmeasＧuredbythisresearchexisteddiscrepancieswithaboveliteratures,andthereaＧsonsmayberelatedtofollowingfactors:①thenumberofresearchcasesweredifferent②researchsamplecategoriesweredifferent．Thisarticleincludedalltypesofgliomas,butsomeresearcheswerelimitedtoastrocytoma③MRequipＧment,contrastagentdosageandpostprocessingsoftwareweredifferent．Inaword,theangiogenesisconditionintumorcouldbelearnedbyDSCPWIexamination,andthemeasurementoftumorsolidpart’smaximumrCBVvaluecouldquantitativelyanalyzethemostobviouspartofangiogenesis,bythiswaytopre－judgegliomapathologygrade,furtherhelpedtoformulatereasonaＧbletreatmentplan．Bibliography[１]RosenBR,BelliveauJW,VeveaJM,BradyTJ．Perfusionimagingwith[NMRcontrastagents．MagnResonMed．１９９０,May,１４(２):２４９－２６５．２]VajkoczyP,MengerMD．Vascularmicroenvironmentingliomas．JNeuＧ[rooncol．２０００Oct－Nov,５０(１－２):９９－１０８．３]DeganiH,Chetrit－DadianiM,BoginL,FurmanHaranE．MagneticresoＧnanceimagingoftumorvasculature．ThrombHaemost．２００３Jan,８９(１):[２５－３３．４]

A(chenMG,JeltschM,KukkE,etal．VascularendothelialgrowthfactorDVEGF－D)isaligandforthetyrosinekinasesVEGFreceptor２(Flk１)andVEGFreceptor３(Flt４)．ProcNatlAcadSciUSA．１９９８,９５:５４８[－５５３．５]WuHonglin,etc．Glioma,MRcontrolstudyofperfusionimagingandmoＧlecularpathology[J]．JournalofClinicalRadiology,２００６,２５(２):１１２[－１１６．６]LawM,YangS,BabbJS,KnobbEA,GolfinosJG,ZagzagD,JohnsonG．ComparisonofcerebralbloodvolumeandvascularpermeabilityfromdyＧnamicsusceptibilitycontrast－enhancedperfusionMRimagingwith[gliomagrade．AJNR２００４,May,２５(５):７４６－７５５．７]KnoppEA,ChaS,JohnsonG,etal．Glialneoplasms:dynamiccontrastenＧ[hancedT２?－weightedMRimaging．Radiology．１９９９,２１１:７９１－１９８．８]LevMH,OzsunarY,HensonJW,RasheedAA,BarestGD,HarshGR,etal．Glialtumorgradingandoutcomepredictionusingdynamicspin－echoMRsusceptibilitymappingcomparedwithconventionalcontrastenhancedMR:confoundingeffectofelevatedrCBVof[oligodendrogliomas．AJNR．２００４,Feb,２５(２):２１４－２２１．９]DeganiH,Chetrit－DadianiM,BoginL,Furman－HaranE?.Magneticr(esonanceimagingoftumorvasculature．ThrombHaemost．２００３Jan,８９l):２５－３３．