Segmentectomy versus lobectomy in patients with stage I pulmonary carcinomaFiveyear survivalof.docx
《Segmentectomy versus lobectomy in patients with stage I pulmonary carcinomaFiveyear survivalof.docx》由会员分享,可在线阅读,更多相关《Segmentectomy versus lobectomy in patients with stage I pulmonary carcinomaFiveyear survivalof.docx(20页珍藏版)》请在冰点文库上搜索。
SegmentectomyversuslobectomyinpatientswithstageIpulmonarycarcinomaFiveyearsurvivalof
JThoracCardiovascSurg1994;107:
1087-1094
©1994Mosby,Inc.
GENERALTHORACICSURGERY
SegmentectomyversuslobectomyinpatientswithstageIpulmonarycarcinoma:
Five-yearsurvivalandpatternsofintrathoracicrecurrence
WilliamH.Warren,MD,L.PenfieldFaber,MD
Chicago,Ill.
SupportedinpartbytheJohnandJuneAntalekFoundation,theKoleFoundation,andcontributorstotheThoracicDiseaseResearchFund.
Addressforreprints:
WilliamH.Warren,MD,Suite218,1725WestHarrisonSt.,Chicago,IL60612.
Abstract
Onehundredseventy-threepatientswithstageI(T1N0,T2N0)non-small-celllungcancerunderwenteitherasegmentalpulmonaryresection(n=68)orlobectomy(n=105)from1980to1988.Fourpatientswerelosttofollow-up,buttheremaining169patientswerefollowedupfor5years.Survivalandtheprevalenceoflocal/regionalrecurrencewereassessed.Althoughnosurvivaladvantageoflobectomyoversegmentalresectionwasnotedforpatientswithtumors3.0cmindiameterorsmaller,asurvivaladvantagewasapparentforpatientsundergoinglobectomyfortumorslargerthan3.0cm.Therateoflocal/regionalrecurrencewas22.7%(15/66)aftersegmentalresectionversus4.9%(5/103)afterlobectomy.Areviewofhistologictumortype,originaltumordiameter,andsegmentresectedrevealednoriskfactorsthatwerepredictiveofrecurrence.Anadditionalresectionforrecurrencewasperformedinfourpatients.LobectomyisthepreferredoperativeprocedureforpatientswithstageItumorslargerthan3.0cm.Becausetherateoflocal/regionalrecurrencewashighaftersegmentalresections,diligentfollow-upofthesepatientsismandatory.(JTHORACCARDIOVASCSURG1994;107:
1087-94)
Overthepast30years,considerablecontroversyhasarisenovertheroleofsegmentalresectionsinthemanagementofstageI(T1N0,T2N0)lungcarcinoma.Althoughsegmentallungresectionhasbeendescribedasareasonableoptionforpatientswithcompromisedpulmonaryreserve,
1-3othershaveadvocateditforpatientsbelievedtobeabletotoleratealobectomy.
4-9Theappropriatenessofsegmentalpulmonaryresectionsmustbemeasuredbytheperioperativemorbidityandmortality,the5-yearsurvival,andtheprevalenceoflocallyrecurrentdisease.Despiteguidelinessuggestedtodistinguishalocalrecurrencefromasecondprimarypulmonarycarcinomaorasolitarypulmonarymetastasis,
10,11problemsofinterpretationpersist.
Forthepurposesofthisclinicalretrospectivestudy,wedefinedalocal/regionalrecurrenceasthedevelopmentofanadditionalcarcinomaintheipsilateralhemithorax(lungandmediastinum)within5yearsoftheresection,regardlessofthehistologicassessmentandtheexactlocationwithinthehemithorax.Althoughthisdefinitionisentirelyarbitraryanddoesnotattempttodistinguishincompletelyresectedtumorsfromsolitarymetastasesorsecondprimarycarcinomas,itis,nevertheless,objectiveandunambiguous.Theprevalenceoflocal/regionalrecurrencewasalsocomparedwiththeprevalenceofcarcinomadevelopinginthecontralateralhemithorax.
Thepurposesofthisclinicalreviewwere
(1)toevaluatesurvivalaftersegmentectomyandstandardlobectomyinthemanagementofstageIpulmonarycarcinomas,
(2)todeterminetheprevalenceoflocal/regionalandcontralateralrecurrence,and(3)toidentifyprognosticfactorsrelatingtosurvivalandtothedevelopmentoflocal/regionalrecurrence.
METHODSANDPATIENTS
Inthisretrospective,nonrandomizedstudy,theclinicalandpathologyfilesfrom1980to1988werereviewedforpatientswhohadundergoneeitherastandardlobectomyorasegmentalresection(definedastheresectionofoneormoreanatomicsegmentsofasinglelobebutlessthanalobe)forapathologicstageI(T1N0,T2N0)non-small-cellcarcinoma.Thechoiceoftheoperationwasatthediscretionofthesurgeon.Patientsbelievedtobeabletotoleratealobectomyunderwentasegmentectomyifthetumorwassmallandperipheral.Anatomicsegmentalresectionswereperformed.Theseoperationsinvolveddissectingoutthehilarstructuresandsecuringthebranchesofthepulmonaryartery,pulmonaryvein,andsegmentalbronchusindividuallyaspreviouslydescribed
1;patientshavingnonanatomic"wedge"resectionswerenoteligible.Alltumorswerelimitedtoonelobeandpatientswithtumorattheresectionmarginwereexcluded.Allpatientshadaregionalnodaldissection.PatientshavingasegmentectomywereincludedingroupIandthosehavingalobectomyingroupII.
Patientswereexcludediftheyhadsynchronoustumors,hadhadapreviousmalignanttumoratanysite,orhadreceivedpreoperativeorpostoperativeadjuvanttherapy.Allotherpatientswereeligibleirrespectiveoftheiroperativeriskfactors,pulmonaryfunction,orcardiacstatus.
Patientswereseenandexaminedatregularintervals,andchestradiographswereroutinelyobtained.Wheneverpossible,pathologicconfirmationofrecurrenttumorwasdonebeforetreatment.
Theprevalenceoflocal/regionalrecurrencewascomparedwiththeprevalenceofcarcinomadevelopinginthecontralaterallunginbothgroupsinthe5yearsoffollow-up.Tumorsthatrecurredassimultaneousbilateralcarcinomaswererecordedseparately.Tumorsthatrecurredsimultaneouslybothintheipsilateralhemithoraxanddistantlywereconsideredtobedistantrecurrences.
Noattemptwasmadetoassessoperativeriskfactorsofthesepatients.ItisacknowledgedthatsomeofthegroupIpatientshadlimitedcardiopulmonaryreserveandwerepoorcandidatesforalobectomy.
Thesurvivalstatisticsandtheprevalenceoflocal/regionalrecurrenceofthetwogroupswereanalyzedwithrespecttotumordiameterasmeasuredbythepathologist,histologicexamination,timeintervalbetweenresectionandappearanceoftherecurrence,andmanagementoftherecurrentcarcinoma.Becausetumordiameterwasconsideredtobepotentiallysignificant,resultswereanalyzedbytumorsize:
2.0cmorsmaller,2.1cmto3.0cm,andlargerthan3.0cm.
Tocomparethefrequenciesofthevariouscategoricaloutcomesinthetwogroups,weused
2testswhenallexpectedcellcountswerefiveormoreandFisher'sexacttestotherwise.Tocomparethetumorsizedistributionsofthetwogroups,weusedthetwo-samplettest.Toexaminetheeffectsofsinglevariablesonsurvival,ontime-until-recurrence,andondisease-freeinterval,weusedlogranktests.Tocheckformultivariateeffectsonthesesameeventtimes,weperformedstepwisefittingofCoxproportionalhazardsmodelsonthefollowingprognosticvariables:
age,sex,maximumtumordiameter,histologictype,andlocationofthetumor(upper,middle,lowerlobes)(rightversusleft).Forallstatisticaltests,thesignificancelevelwas0.05.TheplotsshowKaplan-Meierestimatesoftheproportionineachgroupaliveasafunctionoftimeaftertheoperation.AllstatisticalanalyseswereperformedintheSASstatisticalpackage(SASInstitute,Inc.,SASLanguageandProcedures:
Usage,Version6,1sted.,Cary,N.C.,SASInstituteInc.,1989).
RESULTS
Casehistoriesof173patientswerereviewed.Fourpatientswerelosttofollow-up.Oftheremaining169patients,66hadundergoneasegmentectomy(groupI)and103hadundergonealobectomy(groupII).TheagesofgroupIpatientsrangedfrom36to81years(mean63.9+9.8years);66.7%weremale.TheagesofgroupIIpatientsrangedfrom31to87years(mean63.8+9.9years);65.0%weremale.
TumorsingroupIwereclassifiedhistologicallyasadenocarcinoma(44cases),squamouscarcinoma(15cases),"mixed"phenotypes(5cases),andlargecellcarcinoma(2cases).TumorsingroupIIwereassessedtobeadenocarcinoma(53cases),squamouscarcinoma(35cases),"mixed"phenotypes(6cases),andlargecellcarcinoma(9cases)(
TableI).Thecompositionofthetwogroupswasnotstatisticallysignificantlydifferentwithrespecttohistologictype(p=0.16,Fisher'sexacttest).
TableI.Histologicdescriptionoftumors
Tumortype
GroupI(segmentectomy)
GroupII(lobectomy)
No.
%
No.
%
Adenocarcinoma
44/66
67
53/103
51
Squamouscarcinoma
15/66
23
35/103
34
"Mixed"phenotypes
5/66
8
6/103
6
Largecellcarcinoma
2/66
3
9/103
9
Overall,carcinomasingroupIweresmallerthanthoseingroupII(p<0.0001,Student'stwo-tailedtwo-samplettest)(
TableII).Ofthe66carcinomasingroupI,38were2.0cmorsmaller,13were2.1to3.0cm,and15werelargerthan3.0cmindiameter(3.1to6.5cm).Ofthe103carcinomasingroupIItumors,34were2.0cmorsmaller,10were2.1to3.0cm,and59werelargerthan3.0indiameter(3.1to16.0cm).
TableII.Distributionofcarcinomasaccordingtodiameter
Tumorsize(cm)
GroupI(segmentectomy)
GroupII(lobectomy)
Totals
No.
%
No.
%
<2.0
38/66
58
34/103
33
72
2.1-3.0
13/66
20
10/103
10
23
>3.0
15/66
23
59/103
57
74
Totals
66
103
169
Mean
2.23cm
3.28cm
Standarddeviation
0.97
1.71
Threepatients(2lobectomy,1segmentectomy)diedintheperioperativeperiodofpulmonaryembolus(n=1),myocardialinfarction(n=1),oradultrespiratorydistresssyndrome(n=1).
At5years,consideringtumorsofalldiameters,patientsundergoinglobectomy(groupII)haveastatisticallysignificantsurvivaladvantageoverpatientsundergoingsegmentectomy(groupI)(p=0.035)(Fig.1).However,acomparisonofpatientswithtumors2.0cmorsmallerandpatientswithtumors2.1to3.0cmindiametershowednostatisticallysignificantdifferencebetweengroupsIandII(p=0.24and0