www.elsevier.com/locate/radonline
CT-baseddelineationoflymphnodelevelsandrelatedCTVsinthenode-negativeneck:DAHANCA,EORTC,GORTEC,NCIC,
RTOGconsensusguidelines
´goirea,*,1,PeterLevendagb,1,KianK.Angc,JacquesBernierd,MarijelBraaksmab,VincentGre
VolkerBudache,CliffChaoc,EmmanuelCochef,JayS.Cooperc,GuyCosnardf,AvrahamEisbruchc,SamyEl-Sayedg,BahmanEmamic,CaiGrauh,MarcHamoiri,
´ReychlerkNancyLeec,PhilippeMaingonj,KarinMullerb,Herve
a´CatholiquedeLouvain,St-LucUniversityHospital,B-1200Brussels,BelgiumRadiationOncologyDept.andLaboratoryofRadiobiology,Universite
bRadiationOncologyDept.,ErasmusMedicalCenterDanieldenHoedCancerCenter,Rotterdam,TheNetherlands
cHeadandNeckgroup,RTOG,1101MarketStreet,14thFloor,Philadelphia,PA19107,USAdHeadandNeckgroup,EORTC,AvenueE.Mounier,83/11,B-1200Brussels,BelgiumeRadiotherapygroup,EORTC,AvenueE.Mounier,83/11,B-1200Brussels,Belgium
f´CatholiquedeLouvain,St-LucUniversityHospital,B-1200Brussels,BelgiumRadiologyDept.,Universite
gHeadandNeckDSG,NCIC-CTG,10AlcornAvenue,Suite200,Toronto,Ont.,Canada,M4V3B1
hDAHANCA,Norrebrogade44,8000AarhusC,Denmark
i´CatholiquedeLouvain,St-LucUniversityHospital,B-1200Brussels,BelgiumHeadandNeckSurgeryDept.,Universite
jGORTEC,RueC.Desmoulins,39,94805,Villejuif,France
h´CatholiquedeLouvain,St-LucUniversityHospital,B-1200Brussels,BelgiumMaxillo-FacialSurgeryDept.,Universite
Received4July2003;receivedinrevisedform18August2003;accepted12September2003
Abstract
Backgroundandpurpose:Theappropriateapplicationof3-DCRTandIMRTforHNSCCrequiresastandardizationoftheproceduresforthe
´goireetal.,andtheso-delineationofthetargetvolumes.Overthepastfewyears,twoproposals—theso-calledBrusselsguidelinesfromGre
calledRotterdamguidelinesfromNowaketal.—emergedfromtheliteratureforthedelineationofthenecknodelevels.Detailedexaminationoftheseproposalshoweverrevealedsomeimportantdiscrepancies.
Materialsandmethods:Withinthisframework,theBrusselsandRotterdamgroupsdecidedtoreviewtheirguidelinesandderiveacommonsetofrecommendationsfordelineationofnecknodelevels.ThisproposalwasthendiscussedwithrepresentativesofmajorcooperativegroupsinEurope(DAHANCA,EORTC,GORTEC)andinNorthAmerica(NCIC,RTOG),which,aftersomeadditionalrefinements,haveendorsedthem.Theobjectiveofthepresentarticleistopresenttheconsensusguidelinesforthedelineationofthenodelevelsinthenode-negativeneck.Resultsandconclusions:FirstashortdiscussionofthediscrepanciesbetweenthepreviousBrusselsandtheRotterdamguidelinesispresented.Thegeneralphilosophyoftheconsensusguidelinesandthemethodologyusedtoresolvethevariousdiscrepanciesarethendescribed.TheconsensusproposalisthenpresentedandrepresentativeCTVsthatareconsistentwiththeseguidelinesareillustratedonCTsections.Last,thelimitationsoftheconsensusguidelinesarediscussedandsomeconcernsaboutthedirectapplicationsoftheseguidelinestothenode-positiveneckandthepost-operativeneckaredescribed.q2003ElsevierIrelandLtd.Allrightsreserved.
Keywords:Radiotherapy;Headandneck;Nodelevels;Guidelines
1.Introduction
Theimplementationofthree-dimensionalconformalradiotherapy(3D-CRT)andintensity-modulatedradiationtherapy(IMRT)permitsfargreatercontrolofdose
*Correspondingauthor.1Bothauthorscontributedequallytothepaper.
0167-8140/$-seefrontmatterq2003ElsevierIrelandLtd.Allrightsreserved.doi:10.1016/j.radonc.2003.09.011
228´goireetal./RadiotherapyandOncology69(2003)227–236V.Gre
distribution,selectionandthedelineationoftarget
volumes.Thisnewcapabilityisparticularlyimportantforthemanagementoftumorsintheheadandneckregion,where,withfewexceptions(e.g.earlystagelaryngealandoralcavitytumors),radiationoncologistspreviouslyhavebeenusedtocomprehensivetreatmentofallnecknodelevels.Inrecentyears,however,ithasbeensuggestedthatmoreselectivetreatmentofthenecknodescouldleadtosubstantialreductioninthedoseinflictedoncriticalorgansatrisk,suchastheparotids,withoutjeopardizingloco-regionalcontrol[3,8].However,sub-optimalselectionanddelineationoftargetvolumescouldeasilyjeopardizetheclinicalimpactoftheexquisitedosedistributionsproduced.
Overthepastfewyears,severalauthorshaveadvocatedtheconceptoflimitedtreatment,i.e.selectiveneckdissectionorselectiveneckirradiation,forlimitedstagetumors(seereviewsinRefs.[2,6,7,9]).Itisbeyondthescopeofthisarticletodiscussthisissueatlength.But,comprehensivereviewoftheliteraturehasindicatedthatinthepreviouslyuntreatedneck,thelymphnodedrainageoftheoralcavity,larynxandpharynxfollowsasufficientlypredictablepatternthattheconceptofselectivetreatmenthasalegitimaterationale.Theapplicationofthisconcepthoweverrequiresstandardizationoftheterminologyandproceduresforbothneckdissectionandneckirradiation.In1991,theCommitteeforHeadandNeckSurgeryandOncologyoftheAmericanAcademyforOtolaryngology—HeadandNeckSurgeryproposedasetofdefinitionsofthevariouslymphnodedissectionprocedures[15].Theserecommendations,popularizedbyRobbins,werebasedonasystematicclassificationofthenecknodesintosixlevels,theboundariesofeachbeingdefinedbysurgicallyvisiblebones,muscles,bloodvesselsornerves.Theserecommen-dationsrecentlyhavebeenupdated,withrefinementsofsomeboundariesusingradiologiclandmarks,andfurtherdefinitionofsub-levels(e.g.IIa–IIb,Va–Vb)[16,17].Inthewakeoftheserecommendations,severalgroupshavetranslatedtheanatomicboundariesofthevariousnecknodelevelsonCT-orMR-scans[2,9,13,14,19,20].IntheRadiationOncologycommunity,twooftheseguideline
proposals—theso-calledBrusselsguidelinesfromGre
´goireetal.,andtheso-calledRotterdamguidelinesfromNowaketal.—appeartobethemostwidelyusedinclinicalpractice[9,14].TheRotterdamguidelineshavefurtherevolvedintoasimplifiedversionthattheirauthorsconsidermoreusableinaroutinepractice[20].DetailedexaminationoftheBrusselsandRotterdamrecommendations,however,revealssomeimportantdiscrepancies,preventinguniformdelineationofthetargetvolumesintheneckamongradiationoncologists.
Withinthisframework,theBrusselsandRotterdamgroupsdecidedtoreviewtheirguidelinesandderiveacommonsetofrecommendationsfordelineationofnecknodelevels[11].ThisproposalwasthendiscussedwithrepresentativesofmajorcooperativegroupsinEurope
(DAHANCA,EORTC,GORTEC)andinNorthAmerica(NCIC,RTOG),which,aftersomeadditionalrefinements,haveendorsedthem.Theobjectiveofthepresentarticleistopresenttheconsensusguidelinesforthedelineationofthenodelevelsinthenode-negativeneck.FirstashortdiscussionofthediscrepanciesbetweenthepreviousBrusselsandtheRotterdamguidelinesispresented.Thegeneralphilosophyoftheconsensusguidelinesandthemethodologyusedtoresolvethevariousdiscrepanciesarethendescribed.Theconsensusproposalisthenpresentedandrepresentativeclinicaltumorvolumes(CTVs)thatareconsistentwiththeseguidelinesareillustratedonCTsections.Last,thelimitationsoftheconsensusguidelinesarediscussedandsomeconcernsaboutthedirectappli-cationsoftheseguidelinestothenode-positiveneckandthepost-operativeneckaredescribed.
2.TheBrusselsandtheRotterdamguidelinesforlymphnodeleveldelineationinthenode-negativeneckTheBrusselsproposalwasanattempttotranslatepreciselytheconceptofnecknodelevelsasdefinedbyRobbinsonCTslicesusingsimilaranatomicboundaries,andtoextendtheconcepttonodesnotcoveredbyRobbins,i.e.theretropharyngealnodes.Afewanatomicboundariesoriginallydefinedbynerves,vesselsormuscleshadtobeadaptedtotakeintoaccountthelimitationsandadvantagesofCTscans.RobbinsoriginallydescribedthecraniallimitoflevelIIasthebaseofskull.Inreality,surgeonsusedtheinsertionoftheposteriorbellyofthedigastricmuscletothemastoidasthecraniallimitoflevelII;theBrusselsguidelinesinsteadusedthebottomedgeofthebodyofC1,whichiseasilyidentifiableonCTscan.Similarly,RobbinsdefinedthecaudallimitoflevelIIIasthepointatwhichtheomohyoidmusclecrossedtheinternaljugularvein(IJV);theBrusselsguidelinesinsteaddefinedthecaudallimitoflevelIIIasthebottomedgeofthecricoidcartilage.Lastly,Robbinsusedthespinalaccessorynerve(SAN)tosub-dividelevelIIintoIIa(anteriortoaverticalplanedefinedbythenerve)andIIb(posteriortothatplane).BecausetheSANcannotbeidentifiedonCTscans,theBrusselsguidelines,asproposedbySom,usedtheposterioredgeoftheIJVforthesubdivisionbetweenlevelsIIaandIIb[19].
TheRotterdamproposalscomprisedtwosetsofguide-lines,theoriginalonepublishedbyNowakin1999,anda‘simplified’versionpublishedbyWijersetal.laterinthesameyear[14,20].Theretropharyngeallymphnodeswereidentifiedinneitherofthesetwoproposals.Fortheoriginalguidelines,aradicalmodifiedneckdissectionwasper-formedonacadaverandthevariousboundariesofthenodelevelsweredemarcated.TheseboundarieswherethentranslatedontoasecondfrozencadaverwhichhadbeenCT-scannedandcutinto5mmthicksections.
´goireetal./RadiotherapyandOncology69(2003)227–236V.Gre229
Theboundariesofthesurgicallevelscouldthenbeprecisely
projectedontothematchedCTslices.
Althoughdevelopedwithsimilarobjectives,theBrusselsandtheoriginalRotterdamguidelinesdifferedsubstantially.Forexample,differencesexistedinthedefinitionsofthecranialborderoflevelII,theposteriorborderoflevelsII,III,IVandV,thecranialborderoflevelVandthecaudalborderoflevelVI.Readersarereferredtotheoriginalpublicationsforacomprehensivedescriptionoftheoriginalrecommendations[9,14].
ThesecondsimplifiedversionoftheRotterdamguide-lineswasdevelopednotonlytosubstituteboundarieswhichwereeasiertoidentify(e.g.vertebralbodies,salivaryglands,pharyngo-laryngeallumen)thantheoriginalanato-micalboundaries,butalsotoallowdelineationofdifferentnodallevelsonalimitednumberofCTslices,fromwhichtheallnecklevelscouldbereconstructedbyinterpolation.Thissimplifiedprotocolsubstantiallyreducedthecontour-ingtimeandallowedselectiveneckirradiationwithsimilarparotidglandsparingcomparedtotheoriginalRotterdamguidelines.However,differencesbetweenthesimplifiedversionoftheRotterdamguidelinesandtheBrusselsguidelineswereevengreaterthanbetweentheoriginalRotterdamandBrusselsguidelines(Fig.1).
3.GeneralmethodologyusedtoreachtheconsensusguidelinesforthedelineationofthenecknodelevelsInviewofthedifferencesobservedbetweentheBrusselsandtheRotterdamguidelines,amultidisciplinaryworkinggroup,includingmembersfromboththeoriginalBrusselsandRotterdamgroups,wascreatedtotrytocreateaunifiedsetofrecommendationsforthedelineationofthevariouslevelsintheclinicallyuninvolved,‘node-negative’neck.Subsequently,theworkinggroupwasenlargedtoincluderepresentativesofAmericanandEuropeancooperativegroups.Allofthephysicianswhocontributedtothecreationoftheseguidelinesarelistedasco-authorsofthismanu-script.Thegeneralprincipleswhichguidedtheactivitiesoftheworkinggroupwere(1)totranslateasaccuratelyaspossiblethesurgicalguidelinesintoradiologicguidelinesbasedonaxialCTsections,and(2)tominimizedifferencesininterpretationoftheguidelines,bydefininglessambiguousboundariesthanpreviouslydescribed.
Severalfactorsmotivatedthepaneltousethepreviouslydescribedsurgicalguidelinesastheirbasicframeofreference.First,perhapsmorethananywhereelse,achieve-mentsinheadandneckoncologyhaveresultedfromcomplementaryinteractionsofsurgeryandradiotherapy.Thiscomplementaritywillbecomeevenmorecriticalforfutureadvancesbasedon3D-CRTand/orIMRTsinceincreasinglymoreprecisedoseswillbedeliveredtoincreasinglymoreprecisetargetvolumes.Thispromptsustoadvocatefortheuseofasimilarlanguagetothatalreadyusedbysurgeonsformorethanadecade.Second,inproperly
Fig.1.ComparisonbetweenthesimplifiedRotterdam(leftsideoftheneck,Arabicfigures)andtheBrussels(rightsideoftheneck,romanfigures)guidelinesforthedelineationofthenecknodelevels.Onthetop,CTsliceatthelevelofthebasilaredgeofthemandible;levelsIa,Ib,IIandV,andretropharyngealnodes(RP)aredisplayed.Onthebottom,CTsliceatthelevelofthecricoidcartilage;levelsIV,VandVIaredisplayed.
selectedpatients,necknodedissectionperformedaccordingtostandardizedprocedures,removingonlyselectednodallevels,hasproducedhighratesofcontrolinthepathologi-callyassessednode-negativeneck,withoutpost-operativeradiotherapy[1].Thisobservationconfirmedthatthelocationsofthelymphaticareasatriskformicroscopicinfiltrationoftenarewell-defined,andretrospectivelyvalidatedtheuseofselectivetissuedissectionasaneffectiveprophylactictreatmentmodalityfortheneckofselectedpatients.Third,tosomeextentinthepast,theselectionandthedelineationofthetargetvolumesforheadandneckradiotherapyweredrivenmorebytechnicallimitationsthanbypatientanatomy.Thisleadtounavoidable,unnecessaryirradiationofnormaltissuesbearinglittleornoriskoftumorcellinfiltration,withthepotentialriskofacuteand/orlatecomplicationsoftreatment.Inthisframework,theuseof
230´goireetal./RadiotherapyandOncology69(2003)227–236V.Gre
aso-called‘surgicalreferencesystem’wasfelttobemore
appropriatethananaccuratetranslationof2Dirradiationtechniquesinto3Dvolumes.
Practically,withthehelpofheadandnecksurgeonsanddiagnosticradiologists,allthelandmarksusedtodelineatethevariousnodelevelsduringaneckdissectionwerelocatedonaxialCTslices.Thesurgicalboundarieswerecriticallyreviewedand,forsomeofthem(e.g.upperlimitoflevelV),newdatarecentlypublishedweretakenintoaccount[10].Tohavebettercorrespondenceofsomeoftheselandmarks(e.g.theupperlimitoflevelsIIandV,thelowerlimitoflevelIV)withanatomicstructureseasilyidentifiableonCTscans,radio-opaquevascularclipswereplacedduringnecknodedissectionsandvisualizedonCTstudiesperformedinthepost-operativeperiod.Whenlandmarksusedbysurgeonswerefeltsomehowambiguousand/orsubjecttointer-observervariation(e.g.thecaudallimitoflevelIV,theposteriorlimitoflevelVinthelowerneck),new,consensuslandmarkswereproposedbythepanel.
4.Theconsensusguidelinesforthedelineationofthenodelevelsinthenode-negativeneck
TheconsensusguidelinesforthedelineationoflevelsI–VIandtheretropharyngeallymphnodesarepresentedinTable1.Theboundariesrefertoapatientlyingsupinewithhis/herheadina‘neutral’position.Theterms‘cranial’and‘caudal’refertostructuresclosertothecephalicandpedalends,respectively.Theterms‘anterior’and‘posterior’werechosentobelessconfusingthantheterms‘ventral’and‘dorsal’,respectively.4.1.LevelsIaandIb
LevelIa(Fig.2B)isauniquemedianregionwhichcontainsthesubmentalnodes.Thelymphnodesarelocatedinatriangularregionlimitedanteriorlybytheplatysmamuscleandthesymphysismenti,posteriorlybythebodyofthehyoidbone,craniallybythegeniohyoidmuscleoraplanetangenttothebasilaredgeofthemandible,caudallybythehyoidbone,andlaterallybythemedialedgeoftheanteriorbellyofthedigastricmuscle.ThemediallimitoflevelIaisvirtual,astheregioncontinuesintothecontralaterallevelIa.NodesinlevelIadraintheskinofthechin,themid-lowerlip,thetipofthetongue,andtheanteriorfloorofthemouth[18].LevelIaisatgreatestriskofharboringmetastasesfromcancerarisingfromthefloorofthemouth,theanteriororaltongue,theanteriormandibularalveolarridge,andthelowerlip.
LevelIb(Fig.2BandC)containsthesubmandibularnodes.Itislocatedwithintheboundariesoftheanteriorandposteriorbellyofthedigastricmuscle,thestylohyoidmuscleandthebodyofthemandible.Itislimitedanteriorlybytheplatysmamuscleandthesymphysismenti,poste-riorlybytheposterioredgeofthesubmandibulargland,mediallybythelateraledgeoftheanteriorbellyof
thedigastricmuscle,andlaterallybythebasilaredgeandinnersideofthemandible,theplatysmaandtheskin.Craniallyitislimitedbythemylohyoidmuscleandthecranialedgeofthesubmandibulargland,andcaudallybyaplanecrossingthecentralpartofthehyoidbone.
Thesubmandibularnodesreceiveefferentlymphaticsfromthesubmentallymphnodes,themedialcanthus,thelowernasalcavity,thehardandsoftpalate,themaxillaryandmandibularalveolarridges,thecheek,theupperandlowerlips,andmostoftheanteriortongue[18].NodesinlevelIbareatriskofdevelopingmetastasesfromcancersoftheoralcavity,anteriornasalcavity,softtissuestructuresofthemid-faceandthesubmandibulargland.4.2.LevelsIIaandIIb
LevelII(Fig.2A–C)containstheupperjugularlymphnodeslocatedaroundtheupperone-thirdoftheIJVandtheupperSAN.Itextendsfromthebaseoftheskulltothecarotidbifurcation(surgicallandmark)orthecaudalborderofthebodyofthehyoidbone(clinicallandmark).LevelIIislimitedanteriorlybytheposterioredgeofthesubmandibulargland,theanterioredgeofthecarotidarteryandtheposteriorbellyofthedigastricmuscle,posteriorlybytheposterioredgeofthesternocleidomas-toid(SCM)muscle,mediallybythemedialedgeofthecarotidarteryandtheparaspinalmuscles(levatorscapulaeandspleniuscapitis),andlaterallybythemedialedgeoftheSCMandtheplatysma.Cranially,thepanelproposedtosetthecraniallimitoflevelIIatthecaudaledgeofthelateralprocessofthefirstvertebra,whichisaneasiestlandmarkthantheinsertionoftheposteriorbellyofthedigastricmuscletothemastoidwhichisthesurgicallandmark(Figs.3and4).Forretropharyngealprimarytumors,thecraniallimitoflevelIIshouldbeextendedtoincludethejugularfossa.Caudally,levelIIislimitedbythebodyofthehyoidbone.
LevelIIisfurthersubdividedintotwocompartments.Thelymphnodeslocatedanteriorlytoaverticalplanedefinedbytheupperone-thirdoftheSAN(surgicallandmark)areincludedinlevelIIa,whereasthelymphnodeslocatedposteriorlytotheSANareincludedinlevelIIb.Fromaradiologicalpointofview,theposterioredgeoftheIJVistakenastheboundarybetweenlevelsIIaandIIb.
LevelIIreceivesefferentlymphaticsfromtheface,theparotidgland,andthesubmandibular,submentalandretro-pharyngealnodes.LevelIIalsodirectlyreceivesthecollectinglymphaticsfromthenasalcavity,thepharynx,thelarynx,theexternalauditorycanal,themiddleear,andthesublingualandsubmandibularglands[18].ThenodesinlevelIIarethereforeatgreatestriskofharboringmetastasesfromcancersofthenasalcavity,oralcavity,nasopharynx,oropharynx,hypopharynx,larynx,andthemajorsalivaryglands.LevelIIbismorelikelyassociatedwithprimarytumorsoftheoropharynxornasopharynx,andlessfrequentlywithtumorsoftheoralcavity,larynxorhypopharynx.
´goireetal./RadiotherapyandOncology69(2003)227–236V.Gre
Table1
ConsensusguidelinesfortheradiologicalboundariesofthenecknodelevelsLevel
AnatomicalboundariesCranial
Ia
Geniohyoidm.,planetangenttobasilaredgeofmandibleMylohyoidm.,cranialedge
ofsubmandibularglandCaudaledgeoflateralprocessofC1
CaudalPlanetangenttobody
ofhyoidbonePlanethroughcentral
partofhyoidboneCaudaledgeofthebodyofhyoidbone
AnteriorSymphysismenti,platysmam.
Posterior
Bodyofhyoidbone
Lateral
Medialedgeofant.bellyofdigastricm.
Medialn.a.a231
Ib
Symphysismenti,platysmam.Posterioredgeofsubmandibulargland
IIa
IIb
CaudaledgeoflateralprocessofC1Caudaledgeofthebodyofhyoidbone
Post.edgeofsub-mandibulargland;ant.edgeofint.carotidartery;post.edgeofpost.bellyofdigastricm.Post.borderofint.jugularvein
Post.borderofint.jugularvein
Basilar
edge/innersideofmandible,platysmam.,skin
Medialedgeofsternocleidomastoid
Lateraledgeofant.bellyofdigastricm.Medialedgeofint.carotidartery,paraspinal(levatorscapulae)m.Medialedgeofint.carotidartery,paraspinal(levatorscapulae)m.Int.edgeofcarotidartery,paraspinal(scalenius)m.Medialedgeofinternalcarotidartery,paraspinal(scalenius)m.Paraspinal(levatorscapulae,spleniuscapitis)m.n.a.
Post.borderofthesternocleidomastoidm.Medialedgeofsternocleidomastoid
III
Caudaledgeofthebodyofhyoidbone
Caudaledgeofcricoidcartilage
IV
Caudaledgeofcricoidcartilage2cmcranialtosternoclavicularjoint
Postero-lateraledgeofthesternohyoidm.;ant.edgeof
sternocleidomastoidm.Anteromedialedgeofsternocleido-mastoidm
Post.edgeofthe
sternocleidomastoidm.Medialedgeofsternocleidomastoid
Post.edgeofthe
sternocleidomastoidm.Medialedgeofsternocleidomastoid
V
Cranialedgeofbody
ofhyoidbone
CTslice
encompassingthetransversecervicalvesselsbSternalmanubrium
Post.edgeofthe
sternocleidomastoidm.Ant-lateralborderofthetrapeziusm.
Platysmam.,skin
VI
CaudaledgeofbodyofthyroidcartilagecSkin;platysmam.
SeparationbetweentracheaandesophagusdRetro-pharyngealBaseofskull
Cranialedgeofthebodyofhyoidbone
Fasciaunderthepharyngealmucosa
Prevertebralm.(longuscolli,longuscapitis)
Medialedgesofthyroidgland,
skinandant.-medialedgeof
sternocleidomastoidm.Medialedgeoftheinternalcarotidartery
Midline
Midlinestructurelyingbetweenthemedialbordersoftheanteriorbelliesofthedigastricmuscles.
ForNPC,thereaderisreferredtotheoriginaldescriptionoftheUICC/AJCC1997editionoftheHo’striangle.Inessence,thefattyplanesbelowandaroundtheclavicledowntothetrapeziusmuscle.cForparatrachealandrecurrentnodes,thecranialborderisthecaudaledgeofthecricoidcartilage.dForpretrachealnodes,tracheaandanterioredgeofcricoidcartilage.
ba4.3.LevelIII
LevelIII(Fig.2D)containsthemiddlejugularlymphnodeslocatedaroundthemiddlethirdoftheIJV.ItisthecaudalextensionoflevelII.Itislimitedcraniallybythecaudaledgeofthebodyofthehyoidbone,andcaudallybythecaudaledgeofthecricoidcartilage.TheanteriorlimitistheposterolateraledgeofthesternohyoidmuscleandtheanterioredgeoftheSCMmuscle,andtheposteriorlimitistheposterioredgeoftheSCMmuscle.Laterally,levelIIIis
232´goireetal./RadiotherapyandOncology69(2003)227–236V.Gre
limitedbythemedialedgeoftheSCMmuscleandmedially
bythemedialedgeoftheinternalcarotidarteryandtheparaspinalmuscles(scalenius).
LevelIIIcontainsahighlyvariablenumberoflymphnodesandreceivesefferentlymphaticsfromlevelsIIandV,andsomeefferentlymphaticsfromtheretropharyngeal,pretrachealandrecurrentlaryngealnodes.Itcollectsthelymphaticsfromthebaseofthetongue,tonsils,larynx,hypopharynxandthyroidgland[18].NodesinlevelIIIareatgreatestriskofharboringmetastasesfromcancersoftheoralcavity,nasopharynx,oropharynx,hypopharynxandlarynx.4.4.LevelIV
LevelIV(Fig.2F)includesthelowerjugularlymphnodeslocatedaroundtheinferiorthirdoftheIJV.AccordingtoRobbins,itextendsfromthecaudallimitoflevelIIItotheclavicle[15].However,itappearsfromcriticalexaminationofsurgicalproceduresthatdissectionoflevelIVtypicallydoesnotgoallthewaydowntotheclavicleanddefinitelyneverreachesthemedialportionoftheclavicleatthelevelofthesternoclavicularjoint(Fig.3).Consequently,itwasagreedamongthepaneltosetthecaudallimitoflevelIV2cmcraniallytothecranialedgeofthesternoclavicularjoint.ThecraniallimitoflevelIVisthecaudaledgeofthecricoidcartilage.TheanteriorandposteriorlimitsarethesameasoflevelIII,i.e.theanteromedialedgeandtheposterioredgeoftheSCMmuscle,respectively.Laterally,levelIVislimitedbythemedialedgeoftheSCMmuscleandmediallybythemedialedgeoftheinternalcarotidarteryandtheparaspinalmuscles(scalenius).
LevelIVcontainsavariablenumberofnodesandreceivesefferentlymphaticsprimarilyfromlevelsIIIandV,someefferentlymphaticsfromtheretropharyngeal,pre-trachealandrecurrentlaryngealnodes,andcollectinglymphaticsfromthehypopharynx,larynxandthyroidgland[18].LevelIVnodesareathighriskofharboring
Fig.2.CTimagingofapatientwithaT1N0M0glotticSCC(seetumorinpanelD).Theexaminationwasperformedonadual-detectorspiralCT(ElscintTwin,Haifa,Israel)usingaslicethicknessof2.7mm,anintervalreconstructionof2mmandapitchof0.7.Contrastmediumwasinjectedintravenouslyatarateof2ml/swithatotalamountof100ml.SectionsweretakenatthelevelofthebottomedgeofC1(panelA),theupperedgeofC3(panelB),midC4(panelC),thebottomedgeofC6(panelD),thebottomedgeofC7(panelE),andmidD1(panelF).NecknodelevelsweredrawnoneachCTsliceusingtheradiologicalboundariesdetailedinTable1.EachnodelevelcorrespondstotheCTV,andthusdoesnotincludeanysecuritymarginfororganmotionorset-upinaccuracy.
Fig.3.Thickcoronal(top)andsagittal(bottom)reconstructionwithvolumerendering.Theexaminationwasperformedonadual-detectorspiralCT(ElscintTwin,Haifa,Israel)usingaslicethicknessof2.7mm,anintervalreconstructionof2mmandapitchof0.7.Radio-opaqueclipswereplacedduringtheneckdissectionprocedureatthecraniallimitoflevelsII(toparrowhead)andatthecaudallimitoflevelIV(bottomarrowhead).Theexaminationwasperformedwithin3–4weeksaftersurgery.
´goireetal./RadiotherapyandOncology69(2003)227–236V.Gre233
Fig.4.Thickcoronal(top)andsagittal(bottom)reconstructionwithvolumerendering.Theexaminationwasperformedonadual-detectorspiralCT(Elscint
Twin,Haifa,Israel)usingaslicethicknessof2.7mm,anintervalreconstructionof2mmandapitchof0.7.Radio-opaqueclipswereplacedduringtheneckdissectionprocedureatthecraniallimitoflevelsII(toparrowhead)andV(bottomarrowhead).Theexaminationwasperformedwithin3–4weeksaftersurgery.
metastasesfromcancersofthehypopharynx,larynxandcervicalesophagus.4.5.LevelV
LevelV(Fig.2C–E)includesthelymphnodesoftheposteriortrianglegroup.ThisgroupincludesthelymphnodeslocatedalongthelowerpartoftheSANandthetransversecervicalvessels.AccordingtoRobbins,levelVislimitedcraniallybytheconvergenceoftheSCMandthetrapeziusmuscles,andcaudallybytheclavicle[15].However,theuppermostpartoflevelVisdevoidofanyconstantlymphnodegroup.Insomecircumstances,afewlymphnodeslyingalongtheupperthirdoftheSANmaybefound,butthesenodesareactuallyincludedinlevelIIb[16].TheuppermostpartoflevelVcontainssuperficialoccipitallymphnode(s),andinconsistently,onesubfasciallymphnodeclosetotheoccipitalattachmentoftheSCMmuscle[18].Theselymphnodescollectlymphaticsfromtheoccipitalscalp,andthepost-auricularandnuchalregions.Theyarenotinvolvedinthedrainageofheadandneckcancersexceptofskintumors.Consequently,thecraniallimitoflevelVthatiscommonlyacceptedanddepictedhasbeenquestioned.Hamoiretal.hasrecentlyproposedtousethelowertwo-thirdsoftheSANasthecraniallimitoflevelV[10].Fromaradiologicalpointofview,ahorizontalplanecrossingthecranialedgeofthebodyofthehyoidboneappearsasareliablelandmarkforthecraniallimitoflevelV(Fig.4).ForthecaudallimitoflevelV,itappearsfromcriticalexaminationofneckdissectionprocedure,thatsurgeonsneverdissecttheneckfurtherdowntothecervicaltransversevessels.ItwasthusagreedbythepaneltosetthecaudallimitoflevelVatCTslicesencompassingthecervicaltransversevessels.Fortheotherboundaries,levelVislimitedlaterallybytheplatysmamuscleandtheskin,andmediallybythespleniuscapitis,levatorscapulaeandscaleni(posterior,medialandanterior)muscles.Anteriorly,itislimitedbytheposterioredgeoftheSCMmuscle,andposteriorlybytheantero-lateralborderofthetrapezius
muscle.ThislaterboundarymeansthatlevelVdoesnotextendposteriorlyallthewaytotheanterioredgeofthetrapeziusmuscle.Practically,avirtuallinejoiningtheantero-lateralborderofbothtrapeziusmusclescanbeusetosettheposteriorlimitoflevelV(Fig.2DandE).
LevelVreceivesefferentlymphaticsfromtheoccipitalandpost-auricularnodesaswellasthosefromtheoccipitalandparietalscalp,theskinofthelateralandposteriorneckandshoulder,thenasopharynxandtheoropharynx(tonsilsandbaseofthetongue)[18].LevelVlymphnodesareathighriskorharboringmetastasesfromcancersofthenasopharynx,oropharynx,subglotticlarynx,theapexofthepiriformsinus,thecervicalesophagusandthethyroidgland.4.6.LevelVI
LevelVI(Fig.2D–F),alsocalledtheanteriorneckcompartment,containsthelymphnodeslocatedinthevisceralspace:thepre-andparatrachealnodesincludingtheprecricoid(Delphian)nodeandtheperithyroidnodesincludingthelymphnodesalongtherecurrentlaryngealnerves.Itislimitedcraniallybythecaudaledgeofthebodyofthethyroidcartilage,caudallybythecranialedgeofthesternalmanubrium,anteriorlybytheplatysmaandtheskinandposteriorlybytheseparationbetweenthetracheaandtheesophagus.Thelaterallimitisthemedialedgeofthethyroidgland,theskinandtheantero-medialedgeoftheSCMmuscle.Fortheparatrachealandrecurrentnodes,thecraniallimitisthecaudaledgeofthecricoidcartilage.Forthepretrachealnodes,theposteriorlimitisthetracheaandtheanterioredgeofthecricoidcartilage(Fig.2E).LevelVIreceivesefferentlymphaticsfromthethyroidgland,theglotticandsubglotticlarynx,thehypopharynxandthecervicalesophagus[18].Thesenodesareathighriskorharboringmetastasesfromcancersofthethyroidgland,theglotticandsubglotticlarynx,theapexofthepiriformsinusandthecervicalesophagus.
234´goireetal./RadiotherapyandOncology69(2003)227–236V.Gre
4.7.Retropharyngealnodes
Retropharyngeallymphnodes(Fig.2AandB)liewithin
theretropharyngealspace,whichextendscraniallyfromthebaseoftheskulltothecranialedgeofthebodyofthehyoidbonecaudally.Thisspaceisboundedanteriorlybythepharyngealconstrictormuscles,andposteriorlybytheprevertebralfascia.Forthesakeofsimplicityandconsistency,thepanelproposedtousethefasciabelowthepharyngealmucosaastheanteriorlimit,andtheprevertebralmuscle(longuscolliandlonguscapitis)astheposteriorlimit.Laterally,theretropharyngealnodesarelimitedbythemedialedgeoftheinternalcarotidartery.Typically,retropharyngealnodesaredividedintoamedialandalateralgroup.Themedialgroupisaninconsistentgroupwhichconsistofonetotwolymphnodesintercalatedinornearthemidline.Thelateralgroupliesmedialtothecarotidartery.Themostsuperiorlymphnodeofthisgroup
isalsocalledthelymphnodeofRouvie
`re.Retropharyngealnodeinvolvementoccursinprimarytumorsarisingfrom(orinvading)themucosaoftheoccipitalandcervicalsomites,e.g.ofthenasopharynx,thepharyngealwallandthesoftpalate.Retropharyngealnodesarealsoatriskincaseofpharyngealtumorswithpositivenecknodesinotherlevelsintheneck[4,5,12].
5.Implicationsofnodallevelsforthecreationofclinicaltumorvolumes
Justasmodernheadandnecksurgeonsselectivelycandissectoneormorenodallevelssuccessfully,itseemslogicaltobelievethatmodernradiationoncologistsshouldbeabletoirradiatesimilarlyselectednodallevels.Atpresent,theabilitytoexaminesurgicalspecimenshistologicallyhasnoradiotherapeuticcounterpart,andthecriteriaandconfidenceforselectiveirradiationoflimitednodallevelsisthereforemorelimited.Whilethecriteriaforsuchselectivetherapylikelywillneedtoberefinedoverthecomingyears(andarenotthesubjectofthisreport),theCTVsthatwillneedtobeirradiatedtoencompassthevariousnodallevelscanbedefinednow.ExamplesofCTVsthatadequatelyencompassthedelineatednodelevelsareshowninFig.2.Acompleteatlasofcontrast-enhancedCTsectionsdepictingguidelineCTVsthatencompassthevariousnodelevelsfromthebaseofskulltothelevelofthesterno-clavicularjointshavebeenpostedontheDAHANCA(http://www.dshho.suite.dk/dahanca/guidelines.html),EORTC(http://groups.eortc.be/radio/EDUCATION.htm)andRTOG(http://www.rtog.org/hnatlas/main.htm)websites.
6.Discussion
Theconsensusguidelinespresentedpreviouslyreflectin-depthdiscussionsofapanelofEuropeanandAmerican
expertsfromvariousheadandneckdisciplines,i.e.radiationoncologists,radiologistsandsurgeons.TheirobjectivewastocreateasetofcommonrecommendationsforradiationoncologiststouseintheirdailypracticeforthedelineationofthevariousnecknodelevelsonCTsections.TheseguidelineswerethenpresentedtotherelevantmajorEuropeanandNorthAmericancooperativegroupsinradiationoncology(DAHANCA,EORTC,GORTEC,NCIC,RTOG),whichsubsequentlyendorsedthem.
Whenreadingtheseguidelines,thefollowinglimitationsmustbeclearlyunderstood
†Theseguidelinesdonotintendtogiveanyrecommen-dationfortheoptimaltreatmentstrategy(observationversusprophylacticirradiation)fornode-negativepatientswithaheadandneckprimary,ortheselectionofvariouslevelsthatrequiretreatment.Recentreviewpublicationshavebeguntoaddresstheseissuesfromthesurgicalandtheradiotherapeuticperspective[2,6,7,9].Inthefuture,additionalbasesforsuchdecisionswillbeforthcoming.Inthemeanwhile,physicianswillhavetoweightheavailabledataonpatternsoflymphnodeinfiltrationforvariouslocations,gradeandextentoftheprimary,theprobabilityoftreatmentmorbidity(e.g.higherriskofxerostomiaincaseofretropharyngealnodeirradiation),andtheabilitytoperformaneffectivesalvagetreatmentincaseofneckrecurrence.Wecurrentlytakenostand,insteadleavingthisdecisiontointerdisciplinaryheadandnecktumorboardsandinstitutionalpoliciesineverycenter.
†ThelevelsdelineatedinFig.2correspondtotheClinicalTargetVolumes(CTV),andthusdonotincludeanysecuritymarginfororganmotionand/orset-upuncer-tainty.ThemagnitudeofsuchsecuritymarginrequiredtogeneratethePlanningTargetVolume(PTV)willbebasedontheinfrastructureandexperienceofeachcenter.†Theseguidelinesaredefinedsolelyfortheirradiationof
node-negative,surgicallynaı
¨venecks,i.e.neckswithnodetectabletumordespiteadequateimagingassessment,andneckswithnoalterationoftheiranatomyduetoprevioussurgery.Atpresent,itisuncleariftheseguidelinescanbeextrapolatedtothenode-positiveneckand/orthepost-operativesituation.Thepanelagreedthatthegeneralprinciplesthatformthebasisoftheconsensusguidelinesstillhold,butthatadditionalrecommen-dationsshouldtakeintoaccounttheprobabilityoftumorspreadintoadjacentanatomicstructuresatrisk.
†Inthenode-positiveneck,animportantfactortoconsideristhepossibilityofcapsularruptureandextracapsularextension(ECE).TheriskofECEisdirectlyproportionaltothesizeofthelymphnode,typicallybeing20–40%fornodessmallerthan1cmindiameter,andabove75%forbulkynodesmorethan3cmindiameter(seereviewinRef.[2]).Thus,forpatientswithverysmall,butdetectable,nodes,itisreasonabletoconsiderthattheconsensusguidelinepresentedinthismanuscriptmay
´goireetal./RadiotherapyandOncology69(2003)227–236V.Gre235
stillhold.Forpatientswhohavenodesmorethan3cmin
diameter,itappearsthatadditionaladjacentstructuresatriskoftumorinfiltration(e.g.theSCMand/orparaspinalmuscles)shouldalsobeincludedintheCTV.Itisknownthatmuscularfasciasarestrongbarriersagainstmuscleinfiltration,andthatwhenthefasciahasbeendisrupted,thewholemuscleisatriskastumorcellseasilypropagateinthefattytissuealongthemuscularfibers.WhethertheentiremuscleshouldbeincludedintheCTV,oronlyaportionofitintheimmediatevicinityofthenode,isunknown.ButbecauseheadandneckIMRTisstillinitsinfancy,itdoesnotseeminappropriatetocoverthemusclemoregenerously,atleastuptoaprophylacticdose.Anotherimportantquestionforthenode-positiveneck,iswhetherthecraniallimit(towardsthebaseofskull)andcaudallimit(towardthesupraclaviculararea)oftheCTVshouldbeenlarged.Again,thereisnodefiniteanswertothisquestion,butitseemsreasonableincaseofinfiltrationoftheupperpartoflevelIItoincludethejugularfossaeintheCTV,and/ortoincludethesupraclavicularareaincaseoflowerneckinfiltration.SuchrecommendationsareingoodagreementwiththepatternofrelapseobservedafterIMRTtreatmentwhereselectiveCTVshavebeendelineated[3,8].
†Inthepost-operativesituation,itseemslogicaltotrytocoveratleasttheentireoperativebed,especiallyincaseofECE.Inaddition,additionalstructuresmayneedtobeincludedintheCTVbasedonthepathologicfindings.Forexample,incaseofECEwithinfiltrationofthefasciaoftheparaspinalmuscles,thesemusclesprobablyshouldbeconsideredathighriskandincludedintheCTV.Similarly,theCTVmayneedtobeenlargedtoincludethejugularfossaeincaseofnodalinvolvementofthecranialaspectoflevelII.Again,becauseheadandneckIMRTisstillinitsinfancy,itdoesnotseeminappropriatetobegenerousintargetvolumedelineationuntilmoredataareavailableonthepatternofrecurrenceafterselectivetreatment.
7.Conclusions
Complex3D-CRTandIMRTforthetreatmentofheadandneckcancersrequiresappropriateselectionandaccuratedelineationoftargetvolumesforsuccessfultreatmentdelivery.Webelievethattheconsensusrecommendationspresentedinthismanuscriptrepresentreasonablestatementsaboutthestate-of-the-artinthree-dimensionaldelineationofthevariousnodelevelsinthenode-negativeneck.TheserecommendationshavebeenendorsedbymajorEuropeanandNorthAmericancooperativegroupsinradiationoncology.Thenode-positiveneckandthepost-operativeneckpresentadditionalcomplexities;althoughsomerelatedalterationstotheconsensusguidelinesarediscussedinthismanuscript,furtherrefinementdefinitelywillbeneededin
thefuture.Inthemeantime,asIMRTforheadandnecktumorsisstillinitsinfancy,generousdelineationofthetargetvolumemightwellbeprudent.
Implementationoftheseguidelinesinthedailypracticeofradiationoncologyshouldcontributetoreducedtreat-mentvariationsfrompatienttopatientandhelptoconductmulti-institutionalclinicaltrialsorretrospectivestudies.Lastly,althoughguidelinesaredesignedtoapplytothevastmajorityofpatients,therewillalwaysbeindividualcasesforwhichsoundreasonsprecludetheiruse.Morethanever,oncologicknowledge,experienceandjudgmentarepre-requisitesforappropriateuseoftherecommendationsproposedinthismanuscript.
References
[1]ByersRM,WeberRS,AndrewaT,McGillD,KareR,WolfP.Frequency
andtherapeuticimplicationsofskipmetastasesintheneckfromsquamouscarcinomaoftheoraltongue.HeadNeck1997;19:14–19.[2]ChaoKS,WippoldFJ,OzyigitG,TranBN,DempseyJF.
Determinationanddelineationofnodaltargetvolumesforhead-and-neckcancerbasedonpatternsoffailureinpatientsreceivingdefinitiveandpostoperativeIMRT.IntJRadiatOncolBiolPhys2002;53(5):1174–84.
[3]ChaoKS,OzyigitG,TranBN,etal.Patternsoffailureinpatients
receivingdefinitiveandpostoperativeIMRTforhead-and-neckcancer.IntJRadiatOncolBiolPhys2003;55:312–21.
[4]ChongVF,FanYF,KhooJB.Retropharyngeallymphadenopathyin
nasopharyngealcarcinoma.EurJRadiol1995;21:100–5.
[5]ChuaDTT,ShamJST,KwongDLW,AuGKH,ChoyDTK.
Retropharyngeallymphadenopathyinpatientswithnasopharyngealcarcinoma.Acomputedtomography-basedstudy.Cancer1997;79:869–77.
[6]ClaymanGL,FrankDK.Selectiveneckdissectionofanatomically
appropriatelevelsisasefficaciousasmodifiedradicalneckdissectionforelectivetreatmentoftheclinicallynegativeneckinpatientswithsquamouscellcarcinomaoftheupperrespiratoryanddigestivetracts.ArchOtolaryngolHeadNeckSurg1998;124:348–52.
[7]EisbruchA,FooteRL,O’SullivanB,BeitlerJJ,VikramB.Intensity-modulatedradiationtherapyforheadandneckcancer:emphasisontheselectionanddelineationofthetargets.SeminRadiatOncol2002;12(3):238–49.
[8]DawsonLA,AnzaiY,MarshL,etal.Patternsoflocal-regional
recurrencefollowingparotid-sparingconformalandsegmentalintensity-modulatedradiotherapyforheadandneckcancer.IntJRadiatOncolBiolPhys2000;46(5):1117–26.
[9]GregoireV,CocheE,CosnardG,HamoirM,ReychlerH.Selection
anddelineationoflymphnodetargetvolumesinheadandneckconformalradiotherapy.Proposalforstandardizingterminologyandprocedurebasedonthesurgicalexperience.RadiotherOncol2000;56(2):135–50.Review.
[10]HamoirM,DesuterG,GregoireV,ReychlerH,RombauxP,Lengele
B.AproposalforredefiningtheboundariesoflevelV.IsdissectionoftheapexoflevelVnecessaryinmucosalcellcarcinomaoftheheadandneck.ArchOtolaryngolHeadNeckSurg2002;128(12):1381–3.[11]LevendagP,BraaksmaM,CocheE,etal.RotterdamandBrussels
CT-basednecknodaldelineationcomparedtothesurgicallevelsasdefinedbytheAmericanAcademyofOtolaryngology—HeadandNeckSurgery.IntJRadiatOncolBiolPhys2003;inpress.
[12]McLaughlinMP,MendenhallWM,MancusoAA,etal.Retro-pharyngealadenopathyasapredictorofoutcomeinsquamouscellcarcinomaoftheheadandneck.HeadNeck1995;17:190–8.
236´goireetal./RadiotherapyandOncology69(2003)227–236V.Gre
[13]Martinez-MongeR,FernandesPS,GuptaN,GahbauerR.Cross-sectionalnodalatlas:atoolforthedefinitionofclinicaltargetvolumesinthree-dimensionalradiationtherapyplanning.Radiology1999;211:815–28.
[14]NowakPJ,WijersOB,LagerwaardFJ,LevendagP.Athree-dimensionalCT-basedtargetdefinitionforelectiveirradiationoftheneck.IntJRadiatOncolBiolPhys1999;45:33–9.
[15]RobbinsKT,MedinaJE,WolfeGT,LevinePA,SessionsRB,Pruet
CW.Standardizingneckdissectionterminology.Officialreportoftheacademy’scommitteeforheadandnecksurgeryandoncology.ArchOtolaryngolHeadNeckSurg1991;117:601–5.
[16]RobbinsKT.Integratingradiologicalcriteriaintotheclassificationof
cervicallymphnodedisease.ArchOtolaryngolHeadNeckSurg1999;125:385–7.
[17]RobbinsKT,ClaymanG,LevinePA,etal.Neckdissection
classificationupdate:revisionsproposedbytheAmericanHeadandNeckSocietyandtheAmericanAcademyofOtolaryngology—HeadandNeckSurgery.ArchOtolaryngolHeadNeckSurg.2002;128(7):751–8.
[18]Rouvie
`reH.Anatomiehumainedescriptiveettopographique,6thed.Paris:MassonetCie;1948.p.226–30.
[19]SomPM,CurtinHD,MancusoAA.Animaging-basedclassification
forthecervicalnodesdesignedasanadjuncttorecentclinicallybasednodalclassifications.ArchOtolaryngolHeadNeckSurg1999;125:388–96.
[20]WijersOB,LevendagPC,TanT,etal.AsimplifiedCT-based
definitionofthelymphnodelevelsinthenodenegativeneck.RadiotherOncol1999;52(1):35–42.
因篇幅问题不能全部显示，请点此查看更多更全内容