CorrelationofCullin1expressionwithclinicopathologicfeaturesinNSCLC
Statisticalanalysis
TheclinicopathologicdataofthepatientsaresummarizedinTable1.WeevaluatedtheassociationofCullin1andKi-67expres-sionwithclinicopathologicvariables.ForstatisticalanalysisoftheexpressionofCullin1andKi-67,thecarcinomaspecimensweredividedinto2groups,aspreviouslydescribed.ExpressionofCullin1wassigni cantlyassociatedwithhistologicaldifferen-tiation(P=0.002)andclinicalstage(P=0.010),buttherewasnorelationbetweenCullin1expressionandotherprognosticfactors,likeage,gender,tumorsize,histologicaltype,andlymphnodesta-tus(Table1).Furthermore,inmostspecimens,thehighexpressionofCullin1wassimilartothehighexpressionofKi-67(P=0.021).
StatisticalanalysiswasperformedusingtheSPSS13.0statisticalprogram.TheassociationbetweenCullin1andKi-67expressionandtheclinicopathologicfeatureswasanalyzedusingthe 2test.Foranalysisofsurvivaldata,Kaplan–Meiercurveswereconstructed,andthelog-ranktestwasperformed.MultivariateanalysiswasperformedusingCoxproportionalhazardsmodel.P<0.05wasconsideredstatisticallysigni cant.
Cullin1wassigni cantlyassociatedwiththesurvivalofNSCLCpatients
Results
CullinwashighlyexpressedinNSCLCtissues
TheexpressionofCullin1wasassessedineightpairedNSCLCandadjacentnon-tumoroustissuesbyWesternblotassay.WefoundthatCullin1proteinwashighlyexpressedinNSCLCtis-suescomparedwiththeadjacentnormaltissue(Fig.1AandB).Tocon rmtheexpressionofCullin1inNSCLC,weinvestigatedtheexpressionofCullin1byusingIHCon114samplesfrompatientswithNSCLC.Expectedly,Cullin1washighlyexpressedinpoordif-ferentiatedspecimenscomparedwithwelldifferentiatedones,whichhadthesametendencywithKi-67.RepresentativeexamplesofreactivityforCullin1andKi-67areshowninFig.2.Cullin1wasexpressedmainlyinthecytoplasmwhereasKi-67wasexpressed
Survivalinformationwasavailableforallthe114patients.Whenallvariableswerecomparedseparatelywithsurvivalstatus,Cullin1(P=0.033),Ki-67(P<0.001),histologicaldifferentiation(P<0.001),tumorsize(P=0.002),andclinicalstage(P=0.023)signi cantlyin uencedthepatients’survival(Table2).Inunivariateanaly-sis,theKaplan–MeiersurvivalcurvesshowedthathighCullin1expressioncorrelatedwithpoorsurvivalwithstatisticalsigni -cance(P<0.001;Fig.3).Inaword,multivariateanalysisusingtheCox’sproportionalhazardsmodelprovedthatCullin1(P=0.033;Table2)wasaprognosticfactorforpatients’overallsurvival.
Discussion
Inthisstudy,weshowedthatCullin1expressioncorrelatedwithseveralclinicopathologicalfactorssuchashistologicaldifferentia-tionandtumorsizein114NSCLCpatients.Therefore,Cullin1couldbeconsideredtoplayanimportantroleinpromotingNSCLCpro-gression,andtheidenti cationofthisproteinmightbehelpfulinpredictingoutcomeandimprovingprognosticmodels.

