资源描述：
《Prognostic Model Predicting Metastatic Castration-Resistant Prostate Cancer Survival in Men Treated With Second-Line Chemotherapy》由会员上传分享，免费在线阅读，更多相关内容在学术论文-天天文库。
1、ArticlePrognosticModelPredictingMetastaticcastration-resistantProstatecancerSurvivalinMentreatedWithSecond-linechemotherapySusan Halabi,Chen-Yen Lin,EricJ. Small,AndrewJ. Armstrong,EllenB. Kaplan,Daniel Petrylak,CoraN. Sternberg,Liji Shen,Stephane Oudard,Johann deBono,Oliver SartorMa
2、nuscriptreceivedFebruary19,2013;revisedMay30,2013;acceptedAugust6, 2013.Correspondenceto:SusanHalabi,PhD,DukeUniversityMedicalCenter,2424ErwinRd,Ste11088,Durham,NC27710(e-mail:susan.halabi@duke.edu).BackgroundSeveralprognosticmodelsforoverallsurvival(OS)havebeendevelopedandvalidatedi
3、nmenwithmetastaticcastration-resistantprostatecancer(mCRPC)whoreceivefirst-linechemotherapy.WesoughttodevelopandvalidateaprognosticmodeltopredictOSinmenwhohadprogressedafterfirst-linechemotherapyandwereselectedtoreceivesecond-linechemotherapy.MethodsDatafromaphaseIIItrialinmenwithmCR
4、PCwhohaddevelopedprogressivediseaseafterfirst-linechemo-therapy(TROPICtrial)wereused.TheTROPICwasrandomlysplitintotraining(n = 507)andtesting(n = 248)sets.Anotherdatasetconsistingof488menpreviouslytreatedwithdocetaxel(SPARCtrial)wasusedforexter-nalvalidation.Adaptiveleastabsoluteshri
5、nkageandselectionoperatorselectednineprognosticfactorsofOS.A prognosticscorewascomputedfromtheregressioncoefficients.Themodelwasassessedonthetestingandvalidationsetsforitspredictiveaccuracyusingthetime-dependentareaunderthecurve(tAUC).ResultsThenineprognosticvariablesinthefinalmodelw
6、ereEasternCooperativeOncologyGroupperformancestatus,timesincelastdocetaxeluse,measurabledisease,presenceofvisceraldisease,pain,durationofhormonaluse,hemoglobin,prostatespecificantigen,andalkalinephosphatase.ThetAUCsforthismodelwere0.73(95%confi-denceinterval[CI] = 0.72to0.74)and0.70(
7、95%CI = 0.68to0.72)forthetestingandvalidationsets,respectively.ConclusionsAprognosticmodelofOSinthepostdocetaxel,second-linechemotherapy,mCRPCsettingwasdevelopedandexternallyvalidated.Thismodelincorporatesnovelprognosticfactorsandcanbeusedtoprovidepredictedprob-abilitiesforindividual
8、patientsandt
