北京潞河医院骨关节科科普号

膝关节骨性关节炎阶梯化治疗的新进展(2023)骨性关节炎（osteoarthritis，OA）是一种严重影响患者生活质量的关节退行性疾病，而膝关节骨性关节炎（kneeosteoarthritis，KOA）在临床最常见，主要表现为膝关节疼痛和活动受限。膝关节骨性关节炎是发病率最高、临床最常见、病程长、阶梯性明显、对个体和社会损害最大的骨关节炎之一。由于种种原因，目前我国各地区、各级医院骨科诊疗水平发展不均衡，关节疾病的诊疗水平参差不齐，对膝关节骨性关节炎的诊疗缺乏系统性的培训、全面深入的认识，难以对膝关节骨性关节炎患者严重程度进行恰当判断，易导致不适合治疗或诊疗延误。有鉴于此，查阅国内外最新文献，聚焦对膝关节骨性关节炎阶梯化治疗：基础治疗、药物治疗、修复性治疗和重建治疗四个层次，经过充分细致、广泛深入、独立客观、科学循证的文献分析，总结形成膝关节骨性关节炎阶梯化治疗的图文并茂、容易理解掌握的新进展。以期本新进展为医务人员对膝关节骨性关节炎阶梯化的治疗工作，提供科学、规范、有效的参考。新进展的全文请见PDF文档。

传统标准的4个X线片：AP,inlet,outlet,andJudetviews_肥胖患者骨盆和髋臼骨折的经皮治疗（2011）Percutaneoustreatmentofpelvicandacetabularfracturesinobesepatients BatesP,GaryJ,SinghG,ReinertC,StarrA.Percutaneoustreatmentofpelvicandacetabularfracturesinobesepatients[J].OrthopClinNorthAm,2011,42(1):55-67. 转载文章的原链接1：https://pubmed.ncbi.nlm.nih.gov/21095435/ 转载文章的原链接2：https://www.sciencedirect.com/science/article/abs/pii/S0030589810000726?via%3Dihub AbstractAbodymassindex(BMI)greaterthan30isbecomingincreasinglycommonintheUnitedStates.Surgeryforpelvicandacetabularfracturesinthispopulationisparticularlyproblematicbecauseconventionaltreatmentoftenrequireslargesurgicalexposures.Thesurgeryforboththesefracturesistechnicallydifficultbecauseofthevolumeofsofttissueandpronenesstocomplications.Woundproblemsandinfectionsareparticularlycommonafteropensurgeryinobesepatients,andtheseincreaselinearlywiththeBMI.Inthisarticle,wepresentasmallconsecutiveseriesover14monthsonobesepatientswhounderwentpercutaneoustreatmentoftheirpelvicoracetabularfractures. KEYWORDSPercutaneous，Acetabulum，Pelvicring，Fractures Definedasabodymassindex(BMI)greaterthan30,calculatedastheweightinkilogramstotheheightinmeterssquared,obesityintheUnitedStatesisbecomingincreasinglycommon.DatafromtheNationalHealthandNutritionExaminationSurveyobtainedin2007to2008showedthat33.8%oftheadultpopulationwasobese.1Thesteadyincreaseinthispopulationover20yearshasbeendescribedasanepidemic,2althoughrecentdatasuggestthattherateofincreaseappearstobeslowing,bothinadultsandchildren.1,3Theimplicationsofthisexpandingobesepopulationforthetraumaorthopediccommunityareenormousbecausethispopulationisadiscreetgroupthathasadifferentphysiologyfromthegeneraladultpopulation.Obesepatientshavehigherratesofpreexistingcomorbidities;theirmetabolicresponsetotraumaisdifferentandtheyhavehigherratesofperioperativecomplications,suchaswoundsepsisandvenousthromboembolism.4-7Inaddition,thereisevidencetosuggestthattheobesepopulationisatagreaterriskofpelvicinjuriesthanthegeneralpopulation.8Throughmultiplestudies,surgicaltreatmentoffracturesinvolvingthepelvisandacetabulumhasbeenshowntocarryamuchgreatermorbiditywhenperformedinobesepatients.9-13Forpelvicringinjuries,overallcomplicationratesofsurgeryhavebeenreportedtobeashighas54%,withwoundsepsisbeingbyfarthebiggestcomponent.10Intheacetabulum,ratesofwoundinfection,thromboembolism,andoperativebloodlossarealsoincreased2-fold,andtheseratesfollowalinearrelationshipwithbodymass,withoverallcomplicationratesreachingashighas63%inmorbidlyobesepatients(BMI>40).9,13Withsuchabloodcurdlingcomplicationprofilefromtraditionalsurgery,aless-invasivesurgicaloptionforthesechallengingfracturesinthishigh-riskpopulationispotentiallyveryattractive.Iftheinfectionratealonecouldbereduced,withoutalteringfunctionaloutcome,itwouldyieldahugebenefit.Formorethan10yearsatourinstitution,wehavebeenpercutaneouslytreatingallpelvicringinjuriesandmanyoftheacetabularfractures,particularlyinhigh-riskpatients,suchasthosewhoareobese.Inthisarticle,wepresentasmallconsecutiveseriesover14monthsonobesepatientswhounderwentpercutaneoustreatmentoftheirpelvicoracetabularfractures. PATIENTSANDMETHODSAretrospectivereviewwasperformedusingourhospitalsurgicaldatabase,afterinstitutionalreviewboardapproval.Overa14-monthperiod,betweenJanuary2008andMarch2009,theauthorsperformedpelvicoracetabularsurgeryon117consecutivepatients.AchartreviewofeachofthesesurgeriesrevealedacalculatedBMIof30ormorein38patients.Heightandweightdatawereunavailableinonepatient,butnootherhistoryofobesitywasnotedintheirchart.Ofthe38obesepatients,24hadapelvicringinjury,17hadanacetabularfracture,and3hadacombinationofboth.Overall,theauthorstreated16pelvicringinjuriesand7acetabularfracturesusingpercutaneoustechniques,makingatotalof23fracturesin20patients.ThefracturetypesaresummarizedinTable1.Fullmedicalrecordswereavailableforreviewinall20patients.  Table1SummaryofinjuriesinobesepatientsandtheirtreatmentAbbreviations:APC,anteroposteriorcompression;LC,lateralcompression.  Themeanageofthe20patientswas35andtheaverageBMIwasalso35,withonlyonepatientexceedingthethresholdof40forbeingmorbidlyobese(discussedlaterinCaseExample:Patient14).Therewere14menand6women.PelvicringinjurieswereclassifiedusingthesystemofYoungandcolleagues14andacetabularfracturesbythatofLetournelandJudet.15Primaryoutcomemeasureswerepostoperativecomplicationsrequiringrepeatsurgery,woundinfection,deepveinthrombosis(DVT),pulmonaryembolus(PE),andradiographicappearancebothimmediatelypostoperativelyandatfollow-up.Deepwoundinfectionsweredefinedasthoserequiringsurgicaldebridement. RadiographicReviewPatientswithpelvicringinjuriesunderwentanteroposterior(AP),inlet,andoutletviewsonadmissionandateachfollow-up,whereasthosewithacetabularfractureshadAPandJudetviews.FracturedisplacementwasmeasuredusingpixelcalibrationwithastandardrulerfromthePictureArchiveCommunicationSystem(PACS-MagicWeb,SiemensInc,NewYork,NewYork,USA).Measurementswereperformedby2independentobservers(P.B.andJ.G.)andameancalculated.Wheremultipleviewswereavailable,measuresshowingmaximaldisplacementwerechosen.Forthepelvicring,themethodtheauthorsusedformeasuringdisplacementwasthesameasthatbyLeFaivreandcolleagues.16Ahorizontallineisdrawnacrossthesuperiorendplateofthefifthlumbarvertebra.Byusingthislineasareferencefromwhichtomakeeitherhorizontalmeasurements(in-line)orverticalones(perpendicular),themaximumdisplacementoftheanteriorandposteriorringcouldbemeasured.Displacementswerecalculatedbycomparingnormalbonylandmarkstoeitherthereferencelineoraperpendiculartoit.Whentherewerebothpelvicringandacetabularfractures,bonylandmarksthatwerenotinvolvedwiththehipinjuryandwhosepositionwasfixedtotheconstantfragmentoftheiliumwerechosen.Thereductionpostoperativelyandatlatestfollow-upwasrecordedandgradedaccordingtothemethodofTornettaandMatta,17withexcellentgradebeingaresidualdisplacementof0to4mm;good,4to10mm;fair,10to20mm;andpoor,greaterthan20mm.Fortheacetabularfractures,choosingagradingsystemwasmoredifficult.TheauthorsusedthesystemofAndersonandcolleagues,18recentlydescribedfortheassessmentoffemoralheadmedializationafteramodifiedStoppaapproach.ThisapproachinvolvestakingalinefromthespinousprocessofL5downtothesymphysispubisandmeasuringthedifferenceinthedistancebetweenthislineandthecenterofthefemoralheadoneachside.Wehavegradedthisasgood(0-4mm),fair(4-10mm),andpoor(greaterthan10mm).Oursecondoutcomewasradiologicalsignsofarthriticchangeatfinalfollow-up,graded1to4(1,normalappearance;2,osteophytes;3,narrowedjointspace;4,boneonbone). SurgicalTechniqueGeneralconsiderationsForboththepelvicringandacetabulum,patientswerepositionedsupineonaradiolucentbedwiththeabdomenandipsilaterallowerextremitypreppedfree.Theweightlimitoftheoperatingroomtablewascheckedinsupermorbidlyobesepatients,andnitrousoxidewasavoidedwithanesthesiabecauseexcessivebowelgascanlimitfluoroscopicvisualization.Intra-abdominalcontrastwasflushedoutwherepossible.Patientswereparalyzedduringsurgery.Postoperatively,patientswererestrictedto3monthsoftoe-touch-weightbearingonthesideofthepelvicringinjuryoracetabularfracture. PelvicringThetechniqueweusedforpercutaneouspelvicringfixationhasbeenwelldocumentedinthepast16,19andinvolvestheuseofapelvicreductionframe.(Fig.1)Inshort,thissystemenablesonesideofthepelvistobestabilizedtotheoperatingtable,whiletheothercanbemanipulatedandfine-tunedwithahighdegreeofradiologicalaccuracyandcontrol.Oneofthestrengthsofthissystemisthatitallowsnear-anatomicreductionsofthepelvicring,withouttheneedforeitheropenapproachesormultipleassistants.OurstandardfixationwaswithtranssacraliliosacralscrewsinbothS1andS2,althoughthisvariedaccordingtopatientanatomy.Fortheanteriorpelvicring,thetechniqueoffixationdependedonthefractureconfigurationandthesofttissueenvelope.Althoughsomesymphysealdisruptionsweretreatedwithastandardplateandscrews,whensofttissueswerepoororthefracturewasopen,anexternalfixatorwasused.Inonecase,acerclagewirewasused.Pubicramusfracturesweregenerallytreatedwithcolumnscrewsinthisstudy.  Fig.1.(A)Imageofthepelvicreductionframethatallowsforclosedmanipulationofpelvicringinjuries.(B)ClinicalimagedemonstratingtheuseoftheStarrFrame.Theframehasbeenanchoredtotherighthemipelvis.Distalfemoralskeletaltractionwasusedontheleftfemurtoassistwithclosedreductionofthedisplacedlefthemipelvis.NotehowtheframeallowsforappropriateC-armpositioning.(CourtesyofStarrFrameLLC,Richardson,TX.)  AcetabulumScrewpathwaysforfixingacetabularfracturesarewelldescribedintheliterature20andarenotdiscussedhere.Reductionmaneuverscanbedividedintoclosedandopentechniques.Inanyperiarticularfracture,thereareusuallysomecapsularattachmentstothejointfragments,whichremainintactaftertheinjury.Ifpatientscanbeboughttotheoperatingroomwithinafewhoursoftheiraccident,simplefracturepatternssuchastransverseconfigurationscanbereducedbymanipulationofthehip.Themostcommonlyused,ofcourse,isin-linetraction,butforcedinternalorexternalrotation,flexion,andabductionareallmaneuversthathaveyieldedanatomicreductionsforus.Clearly,astimepasses,theseclosedreductionmaneuversbecomeincreasinglyunlikelytoworkandmostdisplacedacetabularfracturesrequiresomeformofopenreductionwithminiincision.Withthehipflexedup,weperformopenreductionswithminiincisionthroughasmalllateralwindow,approximately1to2cmbehindtheanteriorsuperioriliaccrest.A3-to5-cmincisionismade,andaftersharplydissectingofftheobliqueabdominalwallmuscleattachmenttothecrest,aCobbelevatorcanbepasseddowntheinnertabledirectlyontothefracturefragments.Inhighjuxta-andtranstectalinjuries,thefracturelinescanbedirectlypalpatedwithafingerthroughtheminilateralwindow,whichgivesanadditionalreadingforreductionoverfluoroscopyalone.SpecializedpelvicreductionclampshavebeendesignedanddevelopedbyCharlesReinert,whichhavesufficientexcursiontoallowclampingoffracturesaroundvoluminoussofttissueenvelopes(Fig.2).Theseclampsallowfortransversefracturestoberotatedandcompressed,foradisplaceddomefragmenttobesqueezedbackdown,andforquadrilateralplatemedializationtobekeyedin.Thecombinationoftraction,specificmanipulation,andpercutaneousclampsallowsustoachievefluoroscopicallyexcellentreductionsinmostofthesecases.Ifwethinkthatanexcellentreductioncannotbeachievedminimallyinvasively,wetreattheseinjurieswithopeninternalfixationviawell-describedapproaches.21,22Weshouldalsostressthatthisapproachisnotstraightforwardsurgery,particularlyinthosewhoareobese,andisatechniquethathasevolvedatourinstitutionoverthelast10to15years.Eachfracturepatternisuniqueandrequiresaslightlydifferentscrewconfigurationandreductionmaneuver.Weadvisesurgeonswhoarekeentotrythistechniquetobeginwithsimple,less-displacedfracturepatternsbeforemovingontomorecomplexones.  Fig.2.(A,B)Picturesofclampsandothermanipulativedevicescreatedforpercutaneousmanagementofpelvicringinjuriesandacetabularfractures.(CourtesyofStarrFrameLLC,Richardson,TX.)  Onceanacceptablereductionisobtained,thefractureisstabilizedwithlargefragment6.5-mmor7.3-mmcannulatedscrews.Thechoiceofscrewpathwaysisspecifictothefractureconfiguration. StatisticalAnalysisAllstatisticaltestsweretabulatedusingSASJMPv7software(SASInstituteInc,Cary,NC,USA).AStudentt-testwasusedtocomparemeansofcontinuousvariables.StatisticalsignificancewassetataPvalueoflessthan.05. RESULTSPelvicRingThe16pelvicfractureshadameanfollow-upof9.7months(3-24),andtherewerenodeaths.Onepatientrequestedtobetransferredoutofstatepostoperativelyandwassubsequentlylosttofollow-up.Themeaninitialdisplacementofthefractureswas22mm,andfinalreductionwasgoodorexcellentin15ofthe16patients,with7havingadisplacementof4mmorless.Noneofthefractureswentdownbyareductiongradebetweenpostoperativeandfollow-upradiographs.Therewasahighlysignificantdifferencebetweeninitialandpostoperativedisplacement(P=.0007)butnotbetweenthepostoperativeandfinaldisplacement(P=.54).PelvicringresultsaresummarizedinTable2.  Table2Pelvicringinjuries:summaryofresultsAbbreviations:F,female;LC,lateralcompression;M,male.  ComplicationsTherewerenoinfectionsinthisgroup,eithersuperficialordeep,andtherewerenopostoperativeDVTsorPEs.Threepatients,inwhomadequateprophylaxiswasimpossible,receivedtemporarycentralvenousfilters,butall3filtersweresubsequentlyremoved.Therewerenonewpostoperativeneurologicdeficits.Therewere2scheduledreturnstotheoperatingtheater,forsupplementaryanteriorfixation.Oneofthese(patient6)patientswasa34-year-oldmanwithAsevereAPcompression(APC)type3injury.Havingstabilizedhispelvicringposteriorly,theanteriorplatingwaspostponedby2weeksbecauseofanopenlaparotomywound,whichextendedverydistally.Oncehissofttissuesweremorecompliant,thesymphysiswasplated,withnofurthercomplications.ThesecondcaseisreportedlaterinthesectionPatient14:CaseExample.Theleastfavorableoutcomewasobservedina67-year-oldlady(theoldestinthisseries)withmoderatelydisplacedbilateralinsufficiencyfracturesandunilateralramusfracturesafterafalldownthesteps.HerBMIwas37,anditwasthoughtinitiallythatshecouldbetreatednonoperatively.However,asaresultofongoingpainandfurtherdisplacementobservedonserialradiographs,thepatientrequiredsurgeryat17days.Bilateralposteriorpercutaneousiliosacralscrewswerepassedafterclosedreductionwiththeframe,andpostoperativeradiographsshowedsomeimprovement(from15to11mmofdisplacement).However,herfollow-upradiographsshowedthatshehad“settled”backtothepreinjuryposition.Withadisplacementof16mmatfollow-up,thepatient’sresultwasgradedas“fair.”Therewerenocasesofheterotopicossification(Brookergrade2orworse)onthelastfollow-upradiographs,but3ofthe16patientswithpelvicringinjuriesdidundergoremovalofsymptomatichardwareafterbonyunion. AcetabulumThe7acetabularfracturestreatedpercutaneouslyhadameanfollow-upof9.1months(4-18months)withnonelosttofollow-upandnodeaths.Onepatienthadanundisplaced,transverse,posteriorwallfracture,butalltheotherswereinitiallydisplacedwithameanof17mm(9-22mm).PatientdatafortheacetabularfracturesaresummarizedinTable3.  Table3Acetabularfractures:summaryofresultsAbbreviations:ACS,anteriorcolumnscrew;F,female;LC-2,LC-2screw;Magic,magicscrew;M,male;SAS,supra-acetabularscrew.aFailureoffixationbetween2and4weeks.  ComplicationsTherewerenoinfectionsinthisgroup,eithersuperficialordeep,andnocasesofpostoperativeDVTsorPEs.Onepatient(describedlaterinCaseExample:Patient14)madeascheduledreturntosurgeryforsupplementaryanteriorfixationofherpelvicring,buttherewerenootherreturnstotheoperatingtheateracutely.Therewerenonewpostoperativeneurologicdeficits,andnoneofthepatientshaddevelopedheterotopicossification(Brookergrade2orworse)attheirlatestfollow-up.Immediatepostoperativereductionwasgraded“good”inallcases,buttherewasalossofreductionandsubsequentdevelopmentofarthritis(grade4)in2ofthe7casesrequiringtotalhipreplacement(patients17and19).Aged51and71years,thesepatientswereattheolderendofourgroup,andbothpatientshadcomplexfracturepatterns(T-typeandassociatedbothcolumn)inrelativelyosteopenicbonewithlargepreoperativemedializationofthehead(22and15mm).Inbothpatients,theinitialpostoperativepositionwasgood,withminimalmedicalization(2mminboth),noarticularstep,andnovisiblearticulargappinglargerthan1to2mmonanyofthe3views.Bothhadfailureoffixationbetween2and4weeksandwerereportedinthechartashavingbeenpartiallyweightbearing.Bythistimeitwasthoughtthatthesepatientswouldbebetterservedbydelayedhiparthroplastyratherthanrevisionfixation,andbothhavedonewellaftertheirhipreplacements.Discountingtheundisplacedfracture,forthe6displacedacetabularfractures,wehad2failuresoffixation,makinganoverallcomplicationrateof33%. CaseExample:Patient14Thepatientwasa21-year-oldpedestrianwhowasstruckbyacar.ShesustainedbilateralAPCinjuries,alongwithbilateraltransverseacetabularfractures,displacedontherightandundisplacedontheleft(Figs.3-5).Thepatientwasmorbidlyobese,withaBMIof45(thelargestofourseries).Afterapplyingtractionandabinderintheemergencyroom,shewastakentosurgeryonthedayofinjuryandthedisplacedacetabularfracturewasmanagedinitially.Througha5-cmlateralwindowandwiththehipflexed,aminimallyinvasivecollinearclampwaspasseddowntheinnertable,intothelessernotch.Withtractionanddeploymentoftheclamp,thefracturewasreduced.Thisreductionwasconfirmedbothbydirectpalpationwithafingerdowntheinnertableandbymultiplanarfluoroscopy.Thefracturewasreducedandstabilizedinthestandardfashion,withanteriorandposteriorcolumnscrews(6.5mmcannulated).  Fig.3.(A)PreoperativeAPpelvisand(B)axialcomputedtomographicscanshowingadisplacedtransverserightacetabularfractureandbilateralAPCpelvicringinjuriesina21-year-oldwoman.  Fig.4.(A)PostoperativeAP,(B)inlet,(C)outlet,and(D)Judetviewsshowsomeresidualwideningoftherightsacroiliacjoint,(E)butanotherwisewell-alignedpelvicringandacetabulum.  Fig.5.(A-D)Radiographsat4-monthsfollow-upconfirmthatthesacroiliacalignmenthasnotchangedandtheacetabulumshowsnosignsofarthrosis.  Thepelvicringwasthenreducedusingthereductionframe.Initially,theleft(leastunstable)sidewasreducedandthenfixedusinganS2iliosacralscrewgoingintothesacralbody.Theleftsidewasthenstabilizedtotheframeandtherightsidereducedontoit,usingtheframeelements.Onceagoodreductionwasachieved,finaldefinitivefixationwasplacedwithasingletranssacralS1screwandbilateralS2screws(allwere7.3mm).Atthispointitwasnotedthatthepatienthadaresidualsymphysealdiastasis,whichcouldbeonlypartiallycontrolledbyanexternalfixator.Itwasthoughtthatshewastoounstabletotoleratefurthersurgery,andtherefore,hersymphysealfixationwaspostponedby5days.ThroughaPfannenstielapproach,thesymphysiswasexposedbutcouldnotbeplatedbecauseofthenarrowramusbeingcompletelyfilledbya6.5-mmscrew.Thediastasiswasthereforecabledtogether,usingacablepasserthroughtheobturatorforamenunderdirectvision.Therewassomeresidualwideningof7mmoftherightsacroiliacjointafterthisprocedure,whichremainedlargelyunchangedatfinalfollow-up.Forbothacetabulumandpelvicring,thepatientwasgradedas“good”andshehadanotherwiseuncomplicatedfollow-up.Atlatestfollow-upthepatienthadsymmetricpain-freehipmovementsandwaswalkingwithoutaids.Therewasnoevidenceofearlyhiparthritis,butshewascomplainingofsomemildbacksymptoms. DISCUSSIONPelvicringinjuriesandacetabularfracturesaresevereinjuries,whicharecommonlytreatedwithreductionandinternalfixationvialargesurgicalapproaches.Complicationsoccurinthebesthands,butwiththeobesepopulation,theyhavebeenshowntoincreasesharply,inanalmostlinearfashionwiththesizeofthepatient.9Ofthere-portedcomplications,infectionandwoundbreakdownarebyfarthemostcommon.Forthepelvicring,Semsandcolleagues10recentlydescribedtheirexperienceoftreatingpelvicfracturesin48obesepatients,comparingtheiroutcomestoalargercohortof134patientswhowerenotobese.Theinvestigatorsfoundthatcomplicationratesinobesepatientswere54%comparedwith15%inthosewhowerenotobese.Theyfoundnotonlysignificantlyhigherwoundcomplicationsintheobesegroup(25%)comparedwiththenonobesegroup(5.9%)butalsoasignificantlygreaterrateoflossofreduction(31%vs6%).Theinvestigatorsstipulateclearlythatalltheinfectionsandwoundproblemsoccurredaroundopenexposures.Therewerenoinfectionsinthe18patientsreceivingonlypercutaneoussurgery.Thismirrorsourfindingsfromthe16pelvicringinjuriesdescribedinthisarticleinwhichnoneofthepatientsdevelopedinfections.Ofthose16patients,4underwentsymphysealplatingthroughaPfannenstielapproachaspartoftheirfixation.Wehaveincludedthesecasesinourpercutaneouscohortbecausethereductionwasalldoneclosedandtheposteriorpelvicringwasuniversallytreatedwithiliosacralscrewsalone.Semsandcolleaguesnoteda14%infectionratewithanteriorplatingofthesymphysis;inoursmallseriesof4,therewerenoinfections,although2weredelayedbecauseofeithernon-compliantsofttissuesoranestheticissues.Semsandcolleaguesalsofoundthatlossofreductionwasclearlyaproblemintheobesepatients(31%inobesepatientsvs6%innonobesepatients),particularlyintheOrthopaedicTraumaAssociationtypeCinjuries.However,theinvestigatorswereunabletocommentonwhethertherewasadifferencebetweenthepercutaneousgroupandtheopensurgicalgroupintermsofreductionloss,soitisunclearwhetheropenreductionoftheposteriorpelvisconferredanysignificantbenefitintermsofstability.Wenotedonelossofpositionbetweenthetimeofsurgeryandfollow-up(patient13)makingacomplicationrateof6%. Ourpreviouslyreportedexperienceofpercutaneouspelvicringfixationhasalsoshownlossofpositionratestobecomparablewiththereportedoutcomesofopenreductions.16SincethefindingsofGriffinandcolleagues,23whofounda13%failurerateofunilateraliliosacralscrewsinverticallyunstablesacralfractures,wehavechangedourposteriorfixationtotranssacralscrews,whichpassrightthroughthesacrumandanchorinthetricorticalboneofthesacralalaandopposingilium.Anecdotally,sincemakingthischange,ourfailureratesinverticallyunstablefractureshavedecreasedandwefeelthatthisseriesreflectsthat.Porterandcolleagues12recentlypublishedastudysimilarlydesignedtothatofSemsandcolleague.Theycompared186nonobesepatientswith102obeseones.Intermsofwoundcomplications,theobesegroupwasatmuchgreaterrisk,witha39%incidencecomparedwith19%inthenonobesegroup.Failuresoffixationwereapproximatelyequalbetweenthe2groups,runningataround7%.ThisfindingismoreinkeepingwithLefaivreandcolleagues’16recentarticleinwhichnocorrelationwasfoundbetweenobesityandfinalqualityofreduction.Fortheacetabulum,complicationratesafteropensurgeryrangefrom38%to63%acrossdifferentstudies.13,24BMIandcomplicationratesappeartohavealinearrelationship,withstatisticallysignificantincreasesinoperativetime,hospitalstay,andbloodlossreportedinmorbidlyobesepatients.13Naturally,thereareseveraltechnicaldifficultiesassociatedwithperformingcomplexacetabularreconstructionsinobesepatients.Thesurgicalapproachesareusuallymoredifficult,requiringlongerdeeperincisionstoachieveadequatevisualization.Standardpelvicinstrumentationmaybedifficulttoapplyaroundtheexcessiveenvelope,andthesheervolumeofsofttissuecanmakepreoperativeandintraoperativeimagingverydifficulttointerpret.Porterandcolleaguesrecentlystated,“.operativefixationofdisplacedacetabularfracturesinthemorbidlyobesepopulationmaypossessacomplicationratethatoutweighsthepotentialforasuccessfuloutcome.”Percutaneousmanagementofundisplacedacetabularfracturesiswelldescribedbyseveralauthorsandcommonlypracticedtoimprovepainandpreventdisplacementincertainfracturetypes.25,26Percutaneoustreatmentofdisplacedfracturesremainscontroversial,withonly1previousreportintheliterature.27Intheobesepopulation,aminimallyinvasiveoptionoffersseveraladvantages,includingshorteroperationtimes,minimalincisions,negligiblebloodloss,andverylowinfectionrates.However,notallacetabularfracturetypesareamenabletopercutaneousfixation,andinourinstitution,thosefracturesthatarenotamenablearetreatedwithstandardopentechniquesviawell-describedextensileapproaches.21,22Undisplacedandminimallydisplacedfracturesinthispopulationaretypicallytreatedpercutaneously.Displacedtransversefracturesandthoseinvolvingtheanteriorcolumnarealsosuitableforminimallyinvasivesurgeryifanexcellentreductioncanbeachieved.However,displacedfracturesoftheposteriorwallandposteriorcolumn,alongwithT-typevariantswithposteriordisplacementarenotconsideredcandidatesforthistypeofsurgery.Inthisstudywehavepresented7acetabularfractures,6ofwhichweredisplaced.Notably,therewerenoinfections,nodeaths,noneurologicinjuries,andnocasesofprovenDVTorPE.Atfollow-up,therewerenocasesofheterotopicossificationrequiringexcision,andnonerequiredhardwareremoval.Muchofthepubliclyexpressedconcernaroundthepercutaneoustreatmentofacetabularfractureshasbeentheabilitytobothachieveananatomicreductionandtomaintainit.Inourseries,allofthefractureshadgoodreductionsimmediatelypostoperatively,withminimalmedializationandgoodarticularcongruity.However,2ofthefractures(33%oftheinitiallydisplacedones)diddisplacebetween2and6weekswithnegativeconsequencesforthehipjoint.Bothofthesepatientswereolderthanthemean(aged51and71years),withosteoporoticfractureconfigurationsandmayhavehadsomecomplianceissues,allofwhichmayhavecontributedtothefailure.InPorterandcolleagues’13recentstudy,theyreporteda10%lossoffixationintheobesegroupthatwasstudied,whichisclearlysuperiortooursmallseries.However,thislossoffixationhastobecounterbalancedbytheir46%woundcomplicationrate,the5%ofnerveinjuries,and3deaths,nottomentionthe5%ofpatientsrequiringexcisionofheterotopicossificationand15%requiringremovalofhardware.Mayoreportedanobesesubsetof21patientsfromacohortof105.Hereportedacomplicationrateof38%,whichwasmainlypopulatedbyinfectionsandwoundproblems(24%),withfailureoffixationin5%.24Russellandcolleagues11describedseveralunusualcomplicationsintheirseriesof12over-weightand5obesepatients,aspartofamuchlargercohortof131acetabularfracturestreatedwithopensurgery.Ofthecomplications,24%wererelatedtopositioningofthepatients,andafurther24%developedwoundbreakdown.Lossoffixationintheentirecohortwasonlyseenin2earlyambulators.Ourseriescontrastsstronglywiththeseotherstudies.Excludingtheoneundisplacedfracture,ouroverallcomplicationratewas33%,composedofthe2fixationfailures.Wesawnowoundproblemsorinfections,noneurologicinjuries,noDVTsorPEs,andnoproblemswithheterotopicossification.Bothofourfailuresoffixationwereinolderpatientswithmoreosteoporoticfractureconfigurations,andbothwerereportedintheirchartstohavewalkedontheirfixationearly.Nonetheless,itwouldseemthatpercutaneouslyfixedacetabularfracturesarelessresistanttoearlyfailurethanthosetreatedopen.Theissueisthereforewhetherweareabletoreducetheseearlyfailuresbyeitherbetterpatientselectionorbettersupervisionofpostoperativeweight-bearingstatus.Thisarticlehasseverallimitations.Itisaretrospectivereview,withalltheinherentshortcomingsthatcomewiththat.Thenumbersaresmallcomparedwithotherstudies,withonly16pelvicringinjuriesand7acetabularfractures,3patientsbeinginbothgroups.Itisthereforedifficulttoraiseanymeaningfulstatisticalanalysis.Wealsohavenocontrolgroupagainstwhichtocompareourresults.Wedidtreat10acetabularfracturesinobesepatientswithopensurgery(seeTable1),butwedidnotthinkthatthiswasavalidcontrolgroupbecausethesepatientshadbeendeemedunsuitableforpercutaneoussurgeryatthetimeofadmission.Thesepatientshadadifferentsetofinjurytypes,notamenabletominimallyinvasivesurgery,andwefeltthatadirectcomparisonwouldbemisleading.BMIwascalculatedfromheightandweightdatarecordedbythenursingstaffinthepatients’charts.However,wecannotbecertainofwhethertheseindiceswereformallymeasuredorsimplyvolunteeredbythepatients,makingtheiraccuracyuncertain.Self-reportingofheightandweighthasbeenshowntobecommonlyinaccurate.28Ourfollow-upwasshort(around9months)and1ofthepatientswithpelvicringinjurieswaslosttofollow-up.However,alltheinjurieshadunitedbythetimeoftheirlastappointment.Radiologicalendpointsremainaproblemforbothpelvicandacetabularfractures.Radiographsarerarelyofidenticalrotationandviewingangle,whichmakesforpotentialerrorsinmeasurement.Fortheacetabulum,applyingtheclassicalgradingsystemofMatta29afterpercutaneoustreatmentisproblematicbecauseitemphasizesreductionofboththecolumnsandthearticularsurface.Percutaneousfixationwith6.5-mmscrewsusuallyobscuresthevisiblejointlineonatleast2views,makingthisanunreliableparameter.Also,whentreatedpercutaneously,thequadrilateralplateoftenremainsmediallydisplaced,despitethedomefragmentandtheheadbeingreduced.Wethereforeusedradiologicalparametersthatcanbereliablymeasuredandthatcorrelatewithfunctionaloutcome.Anotherlimitationisthatpercutaneoustechniqueshaveevolvedovertimeatourinstitution,withthewholesurgicalteamcompletelyfamiliarwiththesurgery.Translatingthisproceduretoanothercentermightnotdeliverthesameresults,particularlyintheshortterm.Onthepositiveside,allsurgerieswereperformedbythesame2surgeons,withacommonpre-andpostoperativemanagementprotocolandrehabilitationschedule.Overall,thisstudyislimitedinseveralways,particularlybysize,buttheoutcomesofoursmallcohortofpatientsaresocontrastingwiththehistoricalliterature,wefeelthatitisworthyofreporting.Itisalso,toourknowledge,thefirstdescriptionofpercutaneousacetabularfracturefixationbeingusedinanobesepopulation. SUMMARYWebelievethatwithcarefulpatientselection,percutaneoustreatmentofacetabularfracturesisagoodalternativetotraditionalopenmanagementinhigh-riskpatientgroups,suchasthosewhoareobese.Forpelvicringinjuries,thereductionframeisaneffectivewayofachievingbonyalignmentofeachhemipelvis,regardlessofpatientsize.Wehavefoundthatpercutaneousstabilizationhasnotresultedinatendencytoloosereduction.Withonly1fixationfailureinthisseriesandnoreportedcomplications,webelievethatpercutaneoustreatmentofpelvicringinjuriesintheobesepopulationisacompellingalternativetoopensurgery.

运动学对线(KA)全膝关节置换（TKA）的适用证、禁忌症、手术技术（2021）Kinematicalignmentintotalkneearthroplasty WeberP,GollwitzerH.Kinematicalignmentintotalkneearthroplasty[J].OperOrthopTraumatol,2021,33(6):525-537. 转载文章的原链接1：https://pubmed.ncbi.nlm.nih.gov/34414467/ 转载文章的原链接2：https://link.springer.com/article/10.1007/s00064-021-00729-4 AbstractinEnglish,GermanObjective:Theobjectiveofkinematicalignmentintotalkneearthroplastyistoimplanttheprosthesisaccordingtotheindividualjointline,legaxisandligamenttension. Indications:Kneeosteoarthritiswithfailureofnonsurgicaltreatmentaccordingtocurrentguidelines. Contraindications:Severedeformityorinstabilityrequiringaconstrainedkneeprosthesis.Necessityofintramedullarystems. Surgicaltechnique:Medialparapatellarapproachtotheknee.Resectionofthecruciateligaments,themeniscusandtheosteophytes.Femur-firsttechniquewithdistalresectionofthefemur,theintramedullaryguideisonlyusedfortheextension/flexionpositioningofthefemoralcomponent.Thepositioninginvarus-valgusisorientatedaccordingtothenativejointlineaftercorrectionofchondralwear.Thedistalresectionshouldbeequaltothethicknessoftheprosthesisconsideringthechondralwear(upto2mm)andthethicknessofthesawblade(1mm).Therotationofthefemoralcomponentissetaccordingtotheposteriorcondylaraxisunderconsiderationofchondralwear.Theamountofresecteddorsalboneshouldcorrespondtothethicknessofthedorsalcondylesoftheprosthesis.Thealignmentofthetibiaisparalleltotheindividualjointline.Thisenablesreconstructionoftheindividualphysiologicalslope,rotationandthevarus-valgusaxis.Extensionandflexiongaparecontrolled.Asymmetriesbetweenthelateralandmedialjointspacearecorrectedthroughavarusorvalgusrecutofthetibiaaslongasthesurgicalplanninghasnotbeenachieved.Thehip-kneeangleiscontrolled;however,theaiminkinematicalignmentistoreconstructtheindividualaxesandligamenttensionsandnotastraightlegaxis.Persistingasymmetriesinligamenttensionareadjustedbyclassicalsofttissuebalancingtechniques.Differencesbetweentheextensionandflexionspacesarecorrectedbyadaptingthetibialslope.Releaseoftheligamentsisusuallynotnecessary;sometimesastrippingofthedorsalcapsuleisperformed.Afterthetrialimplantation,theoriginalprosthesisisimplanted. Postoperativemanagement:Functionalrehabilitationwithweightbearingastolerated. Results:RandomizedstudiesshowedabetterfunctionintheKneeSocietyScoreandabetterrangeofmotionwithkinematicallyalignedprosthesescomparedtomechanicalalignment.Availablemeta-analysesalsoshowedbetterresultsforkinematicallyalignedknees.Thefirstmid-termresultsofthisnewtechniquewithafollow-upof10yearsshowasurvivalrateof97.5%oftheprosthesis. Keywords:Functionalresults;Kinematicalignment;Legaxis;Ligamentbalancing;Mechanicalalignment. IntroductoryremarksMAinTKAisthestandardtreatmentandaimsforaneutrallegaxiswithahip–kneeanglewithin3°ofvarusand3°ofvalgusanda90°orientationoftheimplanttothelegaxis[1,2].Manypatientshaveahip–kneeangleapproximating180°,butithasbeenshownthatuptoonethirdofmenhaveaconstitutionalvarusofmorethan3°.Inthesepatients,mechanicalalignmentofthekneeprosthesismightbeabnormalorundesirableandwillalmostcertainlyrequiresomedegreeofmedialsofttissuerelease[3].Thisispossiblyoneofthereasonswhyupto20%ofpatientswithmechanicallyalignedkneearthroplastiesaredissatisfied[4–6].Kinematicalignment(KA)isarelativelynewtechnique,whichStephenHowellandcoworkersfirstreportedin2008[7].Theaimofkinematicalignmentistoimplantatotalkneeprosthesisaccordingtotheindividualanatomyofeachpatientbyreconstructingthepre-arthritichip–kneeangle,thepre-arthriticjointlineobliquityandthusthenaturaltensionoftheligaments(.Fig.1).  Fig.1 aInkinematicalignmentthepre-arthritichipkneeangle(mechanicalaxis)andtheobliquepositionofthejointlinearereconstructed.Toachievethis,thelateraldistalfemoralangle(LDFA,mean87°)andthemedialproximaltibialangle(MPTA[medialproximaltibialangle])(mean87°)arereconstructed.bDuringwalking,themechanicalaxisofthelegisnotatrightanglestothefloorastheanklesarecloserthanthehips.Themechanicalaxishasanangleofapproximately3°totheverticalaxisduringwalking.Duetothisfact,thenativejointlineisparalleltothegroundduringwalking  Theexpectedbenefitsofkinematicalignmentareamoreindividualreconstructionoftheanatomyofthepatientleadingtoamorephysiologictensionoftheligamentsandapatient-specificindividuallegaxis.Thisshouldleadtoamorenaturalfeelingofthekneeleadinginahigherdegreetoaso-calledforgottenjointandwiththistoahighersatisfactionofthepatients.Sincethegoalofkinematicalignmentistoreconstructtheindividualkinematicsofeachpatient,thisarticlewillshowthetechniquewithamedialpivotingprosthesis.Thistypeofprosthesishasamedialdeepdishedinlayandalateralflatdesign.Amedialpivotkneedesignaimstoreconstructthephysiologicalmovementwithalateralroll-backinflexion[8,9]. SurgicalprincipleandobjectiveTheobjectiveofKAinTKAistoimplanttheendoprosthesisaccordingtotheindividualjointline,legaxisandligamenttensionineverypatient.Thisisachievedbyreconstructingthepre-arthriticsurfacegeometryaccordingtotheheightofthechondralsurface,asitwasonthefemoralandtibialsidebeforedevelopmentofosteoarthritis.Therotationofthecomponentsisreconstructedaccordingtotheindividualkneerotationasmeasuredbytheposteriorcondylaraxis. Advantages–Reproductionofthephysiologicallaxityoftheligamentsovertheentirerangeofmotion–Preservationoftheorientationofthejointline–Reconstructionofthephysiologicallegaxisasitwasbeforetheosteoarthritis,“pre-disease”state–Intraoperativecontrolofthealignmentbycalliper-basedmeasurementsoftheresectedboneandcartilage–Possiblyimprovedfunctioncomparedtomechanicalalignmentwithfasterrehabilitation Disadvantages–Modificationisnecessaryinkneeswithcongenitaloracquiredpathologicdeformities,e.g.patellamalalignment.–Traditionalinstrumentsandimplantsnotdesignedforkinematicalignment–Nodataonlong-termoutcomeandimplantsurvival Indications–Kneeosteoarthritiswithfailureofnonsurgicaltreatmentaccordingtocurrentguidelines[10] Contraindications–Generalcontraindicationsforkneereplacementsurgery–Localinfection–Severedeformityorinstabilityrequiringa(semi-)constrainedkneeprosthesis–Necessitytouse(long)intramedullarystems–Extra-articulardeformity(intheopinionoftheauthorstheplanedLDFAandMPTA(medialproximaltibialangle)shouldnotexceed83°);inthesecasesakinematicalignmentcanbepossiblyperformedwithaslightcorrectionoftheMPTAorLDFA,respectively,orincombinationwithanextra-articularcorrectiveosteotomy.–Relative:subluxationofthepatellawithdysplastictrochlea(eventuallythepositioningofthefemoralcomponentshouldbeinaslightexternalrotation)–Relative:unicompartmentalosteoarthritisoftheknee(possibleindicationforunicompartmentalkneeprosthesis) Patientsinformation–Generalrisksofsurgery,suchasneurovascularcomplicationsetc.–Infection–Persistenceofcomplaintsandpain–Implantloosening,fracture,instability,allergytoimplantmaterial,arthrofibrosis,limitedrangeofmotion,stiffness–Persistenceofaslightvarusorvalgusdeformity(asbeforethedevelopmentofosteoarthritis)–Hospitalization4–6days–Approximately3–6monthsrehabilitationperiod Preoperativeworkup–X-rayimagingofthekneeincludinglong-legstandingradiographs,anteroposteriorandlateralviewsandaxialpatellaview–Preoperativeplanningincludinganalysisofthehip–kneeangle,thelateraldistalfemoralangle(LDFA),MPTA,thetibialslope(.Fig.2)–Individualthree-dimensional(3D)cuttingblocksmaybeusedasanalternativetechnique(notshownhere).Inthiscase,acomputedtomography(CT)scanwith3Dplanningisperformedduringpreoperativeworkup.      Fig.2 Preoperativeplanning.Thehip–kneeangleis7°,thelateraldistalfemoralangle(LDFA)88.5°(1.5°valgus),themedialproximaltibialangle(MPTA)87°(3°ofvarus).ThekneeisplannedinkinematicalignmentwithreconstructionoftheLDFAandtheMPTA.Thiswillresultinvarusof1.5°,whichisthenativeconstitutionalvarusofthispatient.Aperpendicularlinedrawntothetibialjointline(inred)willhelpwiththeorientationoftheextramedullarytibialguideduringsurgeryinrelationtotheanklejoint.Inthiscase,thislinewillendatthesyndesmosis.IftheplanningshowsMPTAorLDFAoflessthan85°ormorethan95°,itshouldbecorrectedtostaywithintheselimits[11]  Instrumentsandimplants–Conventionalinstrumentsforkneereplacementsurgerywithextramedullaryreferencingofthetibia–Distalreferencingguidesforthedistalfemoralcut(.Fig.3)–Tibialrecuttingguides(varus–valgusandslope;.Fig.4)–Highpressureirrigation(jetlavage)–Patient-specificinstrumentsasanalternativetechnique  Fig.3 aThereare4differentreferencingguidesfortheorientationofthedistalfemoralcut.Thefemoralchondrallayerhasarelativelyconstantphysiologicalheightof2mm[12].Byusingthefirstguideshownontheleft,theamountofresectedboneandcartilagecorrespondstotheheightoftheprosthesis(9mm)andreconstructsthephysiologicalheightofthefemur.Theotherblocksareusedtoreconstructthephysiologicalheightofthefemurasitwasbeforetheosteoarthritisbegan.Theseconddisplayedblockwillbeusedinavalgusosteoarthritisinarightkneewithworncartilageonthelateralside.Ithasathickeningof2mmtowardsthebonysurfaceonthelateralsidecomparedtothe“unworn”guidetocompensateforthecartilagelossof2mm.Thisleadstoaresectionof7mmofbonelaterallyand9mmboneandcartilageonthemedialside.Thethirdblockisusedinavarusosteoarthritistoresect7mmofboneonthemedialsideand9mmofboneandcartilageofthelateralside.Thefourthblockisusedinthecaseofnomoreremainingcartilageonthemedialandonthelateralside.Thisleadsto7mmofboneresectiononbothsidestakingintoaccountthelostcartilageof2mmonbothsides.bThereferencingguidesareusedwiththeintramedullaryguide.Thisguideisonlyusedtosettheflexion/extensionposition,thevarus–valguspositionisnotfixedbythisintramedullaryguideduetotheoval-shapedsparewhole(arrow).Thevarus–valguspositionissetindividuallywiththereferencingguides.(WithpermissionfromMedactaInternational,CastelSanPietro,Switzerland)   Fig.4 Specialrecuttingblocksareavailableforthetibia.Afterresectionofthetibiatherecuttingblocksallowaresectionwith2°morevarusorvalgusangle,respectively,oranincreaseordecreaseofthetibialslopeby2°.(WithpermissionfromMedactaInternational,CastelSanPietro,Switzerland)  Anesthesiaandpositioning–Generalanesthesiawithorwithoutregionalanesthesiaorspinalanesthesia–Supineposition–Ifdesired,thetourniquetisappliedatapproximately20cmabovetheupperrimofthepatella;thetourniquetshouldbeusedasshortaspossible–Positioningofthelegwithkneerollandlegholder  SurgicaltechniqueThekinematicalignmentisshownforarightkneewithvarusosteoarthritiswithamedialpivotingknee.Thetechniquecanbeusedinthesamemannerforavalgusknee.However,accordingtotheauthors,theremainingvalgusaxisshouldbelimitedtoamaximumof3°(hipkneeangle)inthesecases.(Figs.5,6,7,8,9,10,11,12,13and14)  Fig.5Standardmedialparapatellarapproachtotheknee,identificationofthechondraldamage.Ifthereisremainingcartilageonthemedialside,completelyremovethecartilagebyusingaringcurette.Ifthelesswornsideshowspartiallyworncartilage,removeitcompletely.Removaloftheosteophytesandanteriorcruciateligament.Alsoassessforbonydefectsthatmustbetakenintoaccountforbonyresections  Fig.6 Theintramedullarycanalisopenedcentrallyapproximately7–10mminfrontoftheposteriorcruciateligament Fig.7 aPositioningofthedistalresectionguide.Theresectionguide(shownin.Fig.3)ischosenaccordingtothecartilagewear.Inthecaseofavarusosteoarthritiswithworncartilagemediallyasshownhere,the“worn”guideisusedonthemedialsideandthe“unworn”guideonthelateralside.Ifthemedialcartilageisonlywornpartially,ithastoberemovedcompletely.Ifthechondrallayeronthelateralcondyleisalsopartiallyworn,itmustbecompletelyremovedandconsequentlythe“worn/worn”guideisused.Theintramedullaryguideonlysetstheflexion/extensionangle.Thevarus–valguspositionofthecuttingguidecanbepositionedindependentlyoftheintramedullaryguideduetoanovalconnection.Therotationofthedistalresectionguideandtherespectivevarus–valgusguideisorientedwithregardtotheposteriorcondyles.Withthistechniquethevarus–valguspositionofthedistalresectionblockissetaccordingtotheindividualanatomyofthepatientundercorrectionofthecartilagewear.Afterthis,itisfixedwithpinsandtheintramedullaryguideaswellasthedistalresectionguideareremoved.bAftercheckingwiththeangelwingthedistalresectioncanbeperformedwiththedistalresectionblock.cTheamountofresectedboneiscontrolledwithacalliper.Theprosthesisusedinthiscasehasathicknessof9mmdistally,sotheamountofresectedboneontheunwornsideshouldbe8mm(+1mmthicknessofthebonetakenawaybythesawblade).Onthewornsidetherewillbe2mmlessasthecartilageisworn,sothethicknessshouldbe6mm(+1mmthicknessofthesawblade).Ifthemeasurementswiththecalliperdonotshowthedesiredresections,approximately0.5to1.0mmcanberesectedononeside(lateral/distal)byanadditionalrecuttingwiththesawbladethroughthesamecuttingblockposition  Fig.8 Afterthedistalresection,thechondrallayeroftheposteriorcondylesisexamined.Ifthisisintact(inmostofthecasesinvarusosteoarthritis,in55%ofthevalguscases),theposteriorreferencingfemoralsizerispositioned.Therotationissetat0°accordingtotheposteriorcondyles.Noexternalrotationisappliedasthegoalofkinematicalignmentisreconstructionoftheindividualjointlinesandthefemoralflexionaxis.Ifthechondrallayerisnotintact,itisremovedcompletelyanda2mmplatecanbefixedintheposteriorreferencingguidetoreplacetheworncartilage.Theguideisfixedwith2pinswithintheposteriorreferencingholesandthefemoralsizecanbedeterminedbypositioningthestylusontheanteriorcortex.Afterwardsthesizerisremovedandthepinsremaininplace     Fig.9aThe4-in-1cuttingblockofthedeterminedsizeispositionedonthepins.Thesurgeonshouldcheckthatthereisnoanteriornotching.Inthenextstep,thedorsalcondylesareresected,andtheheightiscontrolledwiththecalliper.bSincetheprosthesishasadorsalthicknessof8mm,theresectedblocksshouldhaveathicknessof7mm(+1mmsawblade)measuredwiththecalliper.Inthecaseofaworncartilagethethicknessshouldbe5mm(+1mmsawblade+2mmconsideringthelostcartilage).Ifthecorrectresultisobtained,the3remainingcutsareperformed.Otherwise,theblockisrepositionedandthecutsarecorrected.Ifthedifferencetothetargetissimilaronbothsidesby2mm,theblockcanberepositionedunderpreservationofthepinsinthedifferentholesoftheblock.Ifonesideiswrong,thepinofthesamesidehastoberemovedandtheblockisrepositionedandfixedwithonepininthedesiredposition  Fig.10 aThetibialresectionblockisplacedparalleltothenativetibialjointlineaftercompensationforcartilageandbonewear.ThetibialslopeandtheMPTAarereconstructed.Theangelwingcanbeusedtohelpassessthejointline.Asacontrolofthealignment,thepositionoftheextramedullarytibialguideabovetheanklejointcanbecomparedtothepreoperativeplanningandcorrectedifnecessary(Fig.2).Therotationissetaccordingtotheindividualphysiologicalrotation.Thisissetparalleltotheanterior–posterioraxisofthealmostellipticallyshapedboundaryofthearticularsurfaceofthelateraltibialplateau.Akagi’sline(inblue)canalsohelpintheorientation(linefromthetibialinsertionofthePCLtothemedialborderofthetibialtuberosity).Theresectionheight(8mmtarget)issetwithastyluspositionedonthelesswornsidenearthetibialtubercleand8mmofboneandcartilagenearthetibialtubercleisresected.bTheresectedtibiashouldbelaterallyandmediallysymmetrical,consideringcartilageandbonewear.Thetibialslopeshouldbereconstructed.Theamountofresectedboneismeasuredandshouldbe8mmclosetothetibialtuberclelaterallyandmedially Fig.11 Beforeanycorrectionofthetibialresectionisperformed,theextensionandflexionspacesaremeasured.Inkinematicalignment,theflexionspaceshouldbeslightlylooserlaterallytoreconstructthephysiologicalsituation.Thegoalisnottherestorationofarectangularspaceinflexion,buttoreconstructastablemedialflexiongapandaphysiologicallylaxerlateralflexiongapforlateralroll-back.Inextension,thespaceshouldberectangularandstable.Incaseofsymmetry,nofurtherresectionisnecessary.Incaseofdifferences,furtherstepsarerequiredasshownin.Table1  Table1  Fig.12Denervationofthepatellaandremovalofosteophytesandthelateralfacetofthepatella.Thedecisionwhethertoreplacethepatellaislefttothesurgeon.Afterthedesiredtensionoftheligamentshasbeenachievedthroughrecutsonthetibialside,thetrialscanbeinsertedandtherangeofmotion,tensionoftheligaments,andarticulationofthepatellaaretested.Thelegaxiscanalsobechecked;however,thegoalofKAisnottorestorealegaxisof180°,buttorestoretheprediseasestate(asdeterminedintheplanning,see.Fig.2)   Fig.13 aIfthecheckwiththetrialcomponentsgivesasatisfyingresultconcerningrangeofmotionand,balancingofthekneeand,bthemovementofthepatella,thepreparationofthefemurandthetibiaiscompleted.Thetibialrotationissetaccordingtothephysiologicalrotation.Theoriginalprosthesisisimplantedandafinalcheckisperformed.Inamedialpivotingknee,theroll-backonthelateralsidecanbeseeninflexion.Thepatellashouldbespontaneouslycenteredwithinthetrochleathepositionofthepatellaisalsochecked.Finally,thecapsuleandwoundareclosed  Fig.14PostoperativelonglegaxisafterimplantationoftheTKAwithkinematicalignment(samepatientasin.Fig.2):thehip–kneeangleis1.5°varus,theLDFA88.5°,theMPTA87°.Thusthepre-arthriticstatehasbeenreconstructed  Postoperativemanagement–Applicationofasterilewounddressingandanelastocompressivebandage–Aneventualintra-articulardrainisremovedafter24h–Removalofthesuturematerialbetweenthe12thand14thdayaftersurgery–Thrombosisprophylaxisaccordingtotheguidelines–X-raycontrolsinthesurgicaltheatre,after6weeksandafter1yeararerecommended.After6weeksweperformalonglegaxis,alateralandakneeskylineMerchantviewarerecommended.Afterthat,x-rays(anteroposteriorandlateralview)arerecommendedevery2years(Fig.14).–Postoperativephysiotherapy:functionalrehabilitationwithweight-bearingastolerated.Activeandpassivemotionwithinthepain-freerangeofmotion–Musclestrengtheningundersupervisionofaphysiotherapist Errors,hazards,complications–Generalsurgicalrisksasforanykneearthroplasty–Malpositioningofthefemoralcomponentinthecoronalplane.Preventionbymeticulousremovalofthechondrallayeronthewornside.Ifthechondrallayerispartlydamagedonthelesswornside,completeremovalofitandutilizationofthe“worn/worn”guide(Fig.7)–Malpositioningofthefemoralcomponentinthetransversalplane.Especiallyinpatientswithalateralizedpatella,orientationaccordingtothePCAcanleadtopatellarmalalignment.Inthesecases,apreoperativeCTwithmeasurementofthePCAinaccordancetotheTEAisrecommended,sinceaslightexternalrotation(between0and3°)ofthefemoralcomponentmightbenecessaryinthesepatients.–Malpositioningofthetibialcomponentinthecoronalplane.Withstandardextramedullaryreferencing,3°varuspositioningismoredifficulttoachievethan90°positioningtothetibialaxis.Thisriskcanbeminimizedbyconsequentplanningtopredicteventualaccentuateddeformities(Fig.2),bydrawingtheverticalaxistothejointlineintheplanningandtocontrolthisduringsurgerywiththepositionoftheextramedullaryalignmentguideinrelationtotheanklejoint(Fig.2),andbymeasurementoftheresectedbonewiththecalliper. ResultsInrecentyears,severalrandomizedcontrolledstudieshaveproposedthatpatientswithTKAthatwerekinematicallyalignedcomparedtomechanicallyalignedTKAshadabetterfunctionintheKneeSocietyScoreorintheOxfordKneeScore[14–17].Althoughsomeotherstudiesshowednosignificantbetween-groupdifference,noneshowedbetterfunctionalresultsinthemechanicallyalignedgroup[18,19],andthreemeta-analysesconfirmedsuperiorresultsinfunctionandrangeofmotioninkinematicallyalignedTKAgroupsofpatients[20–22].Recently,Ameta-analysiswaspublishedthatincludedonlyrandomizedcontrolledtrials.Intotal,theresultsof371kinematicallyalignedkneeswerecomparedto376mechanicallyalignedknees.Theauthorsshowedadifferenceof7pointsinthekneefunctionscore,abetterrangeofmotionof4°andabettercombinedKneeSocietyScore,whichallfavoredkinematicallyalignedkneesincomparisontothemechanicallyalignedTKAs.Thefollowuplastedbetween6and24months[23].The10-yearsurvivalrateofmorethan222kinematicallyalignedkneeswasreportedtobe97.5%[24].Concernhasbeenraisedthatadeviationofthe90°positioningofthetibialcomponenttothemechanicalaxiscouldleadtoearlyloosening[25].Midtermresultsdidnotshowahigherrevisionrateforvarusorvalguspositioningofthetibialcomponent[11,26,27].Currently,however,morethan5°deviationfromthe90°targetarediscussedcritically[28],andnodefinitivelimitsweredefinedregardinglegaxisandjointline.Ifduringplanningasevereextra-articulardeformityisrecognizedwiththeMPTAorLDFAexceeding83°,intheopinionoftheauthorsarestrictedkinematicalignmentshouldbeperformedwithapartialintra-articularcorrectionofthedeformityto83°MPTAorLFDA.InthecasesweretheMPTAorLDFAexceed80°,anextra-articularcorrectionosteotomyisadvocatedandafterthisakinematicimplantationcanbedonerespectingtheligamenttensionofthepatient.However,noclearlimitshavebeendefineduptonowandtheserecommendationsreflecttheopinionandexperienceoftheauthors.Inpatientswithalateralizedpatella,apositioningofthefemoralcomponentparalleltothePCAcouldleadtopersistinglateralizationofthepatella.Furtheroutcomeassessmentsinthesepatientsaremandatorybeforethetechniquecanbefullyrecommendedinthesepatients.Inconclusion,thekinematicalignmenttechniqueshowsexcellentfunctionalresultsthatseemtobeatleastasgoodasthoseoperatedwiththemechanicalalignmenttechnique.Themidtermdataareverypromising.However,long-termresultswithsurvivalratesof80%after25yearshavebeenpublishedformechanicalalignment,andfurtherstudiesarenecessarytoconfirmthepromisingclinicalresultsinkinematicalignmentinthelong-term[29].

功能对线全膝关节置换胫骨切除范围的扩大增加了股骨髁假体外翻和内旋的发生率（2024）WideningoftibialresectionboundariesincreasestherateoffemoralcomponentvalgusandinternalrotationinfunctionallyalignedTKAOllivierB,WakelinE,PlaskosC,VandenneuckerH,LuyckxT.WideningoftibialresectionboundariesincreasestherateoffemoralcomponentvalgusandinternalrotationinfunctionallyalignedTKA[J].KneeSurgSportsTraumatolArthrosc,2024,32(4):953-962转载文章的原链接1：https://pubmed.ncbi.nlm.nih.gov/38444096/转载文章的原链接2：https://esskajournals.onlinelibrary.wiley.com/doi/10.1002/ksa.12118AbstractPurpose:ThepurposeofthisstudywastoinvestigatetheinfluenceofincreasingthetibialboundariesinfunctionalalignmentonfemoralcomponentorientationinTKA.Methods:Aretrospectivereviewofadatabaseofrobotic‐assistedTKAsusingadigitaljointtensioningdevicewasperformed(BalanceBot®;Corin).Atotalof692TKAswithcorrectabledeformitywereincluded.Functionalalignmentwithatibia‐firstbalancingtechniquewassimulatedbyperformingananatomictibialresectiontorecreatethenativemedialproximaltibialanglewithincertainboundaries(A,87–90°;B,86–90°;C,84–92°),whileaccountingforwear.Afterbalancingtheknee,theresultingamountoffemoralcomponentoutliersinthecoronalandaxialplanewascalculatedforeachgroupandcorrelatedtothecoronalplanealignmentoftheknee(CPAK)classification.Results:Theproportionofkneeswithhighfemoralcomponentvarus(＞96°)orvalgus(＜87°)alignmentincreasedfrom24.5%(n=170)ingroupAto26.5%(n=183)ingroupBand34.2%(n=237)ingroupC(p＜0.05).Similarly,morekneeswithhighfemoralcomponentexternalrotation(＞6°)orinternalrotation(＞3°)wereidentifiedingroupC(33.4%,n=231)thaningroupB(23.7%,n=164)andA(18.4%,n=127)(p＜0.05).Therewasastatisticallysignificant(p＜0.01)overallincreaseinkneeswithbothfemoralcomponentvalgus＜87°andinternalrotation＞3°fromgroupA(4.0%,n=28)toB(7.7%,n=53)andC(15.8%,n=109),withCPAKtypeIandIIshowinga12.9‐and2.9‐foldincrease,respectively.Conclusion:Extendingthetibialboundarieswhenusingfunctionalalignmentwithatibia‐firstbalancingtechniqueinTKAleadstoastatisticallysignificanthigherpercentageofkneeswithavalguslateraldistalfemoralangle＜87°and＞3°internalrotationofthefemoralcomponent,especiallyinCPAKtypeIandII.LevelofEvidence:LevelIV.KEYWORDSfunctionalalignment,internalrotation,robotics,totalkneearthroplasty,valgusINTRODUCTIONChangesinprostheticdesign,alignmentstrategyandtheuseofpatient‐specificinstrumentation,navigationandroboticshaveallbeenproposedassolutionstofurtherenhancepatientsatisfactionafterTKA[3,12,28].Theuseofroboticsallowssurgeonstoaccuratelyplanthebonecutsandanticipatetheireffectonthebalanceoftheknee[37].Toolshavebeendevelopedthatassistinapplyingvarusandvalgusstressonthekneetomeasurethelaxityofthesurroundingligamentsandtheireffectontheflexionandextensiongaps[13].Thedifferenceinindividuallygeneratedstressbetweensurgeonsiseliminatedbytheinventionofdevicesthatapplyapredefinedforceonthemedialandlateralcompartmenttoreproduciblymeasurethesofttissuetensionthroughouttherangeofmotion,liketheCorinBalanceBot®[18,30].ModernTKAalignmenttechniqueslikeadjustedmechanical,kinematic,restrictedkinematic,functionalandinversekinematicalignmentallhavetheirownspecificgoalsandboundaries,whichtranslatesintodifferencesintheresultingbonecuts,softtissuereleases,jointlineobliquity(JLO)andcompartmentalforces[25,26,36].Thesestrategiesoftenacceptsomedegreeoftibialvarustobalancethekneeinextensionwithoutperformingligamentreleases[36].However,theeffectofthisobliquityatthetibialsideonthecoronalandaxialalignmentofthefemoralcomponentinfunctionalalignmentwithatibia‐firstbalancingtechniqueremainsmostlyunknown[17,34].Thegoalofthisstudyistoevaluatetheinfluenceofincreasingthetibialboundariesonfemoralcomponentorientationandoveralllimbalignmentinvarus,neutralandvalgusknees,whenaimingforequallybalancedmedialandlateralgapsinextensionandflexion.Thehypothesisisthatincreasingtheboundariesonthetibialsidewillleadtomorepronouncedvalgusandinternalrotationofthefemoralcomponentinkneephenotypeswithamoreobliquemedialproximaltibialangle(MPTA).MATERIALSANDMETHODSAretrospectivestudyofacommercialinternationaljointregistryofrobotic‐assistedTKAswasperformedafterobtainingethicalapproval(CorinRegistry,WCGIRB:120190312).Inclusioncriteriawereasfollows:caseswhichwereperformedusingaroboticsplatformwithadigitallycontrolledjointtensioner;correctablecoronaldeformity(definedasdeformitywithin±3°fromneutralhip‐knee‐ankle(HKA)angleorpreoperativekinematicswhichcouldachieveanalignmentof±3°HKAunderavarusorvalgusstress);aprecorrectioncoronaldeformityoflessthan10°fromneutral;preoperativejointbalanceusingadigitaltensionerrecorded;atibia‐firstworkflowandtibialresectionvalidated.Exclusioncriteriawereasfollows:entriesinthejointregistrywithanymissingroboticsdatarequiredtocomputethespecifiedalignmentandbalanceparameters;femur‐firstworkfloworrobotic‐assistedTKAwithouttheuseofadigitaljointtensioner.AflowchartofthestudyinclusioncriteriaandfinalTKAsincludedinthestudyisshowninFigure1.Atotalof692caseswerereturnedforanalysis.DatawereobtainedfromsixsurgeonsoperatingacrosstheUnitedStates(US).DemographicdataareunavailablefortheseTKAs.However,coronalandsagittaldeformityandleglength(definedashipcentretoanklecentre)werecapturedaspartoftheroboticsworkflow.FIGURE1Studyflowchart.Summaryofstudyflowchartandfinalnumberofkneesincludedinstudy.TKA,totalkneearthroplasty.SurgicalworkflowInallcases,amedialparapatellarposteriorcruciateligament(PCL)sacrificingtechniquewasperformedusingtheOMNIBoticsrobotic‐assistedplatformwiththeBalanceBot®digitalbalancingtool(Figure2a).Afterthekneewasopened,landmarkingwasperformedtodigitisethekneejoint,hipcentreandanklecentreasdescribedpreviously[10,11].Coronaldeformitywasrecordedbysupportingtheheelandplacingthekneeinfullextensionwithoutapplyinganexternalforce.Preoperativekinematicswererecordedbyextendingthekneethroughafullrangeofmotionwhileapplyingavarusandvalgusstresstotensionthesofttissueprofile.FIGURE2BalanceBot®digitalligamenttensioningdeviceandexampleintraoperativedataoutputavailabletothesurgeon.(a)ImageofBalanceBot®digitalligamenttensioningdevice.(b)Demonstrationofitsuseandpurpose:equalforcesareappliedtothemedialandlateralcompartmentswiththetensioningtoolinthejointwhiletheroboticsystemmeasuresthepositionofthetibiarelativetothefemurandcalculatesthemedialandlateralgapsthroughouttherangeofflexion.Allsurgerieswereperformedwithatibia‐firstgapbalancingmethod.Thetibialresectionwasplanned,executedandvalidatedwiththeroboticssystem.Aneutraltibialresectionwastargeted,ingeneral.However,surgeonsmadethefinalclinicaldecisiononthemostappropriatecoronalresection.Aposteriortibialslopeof3°wasplanned.Afterthetibialresectionwasvalidated,thedigitalbalancingtoolwasinsertedintothejointandatensionofbetween70and90Nwasapplied(Figure2b).Thekneewastakenfromflexiontoextensionwithcarebeingtakennottoapplyexternalcoronaloraxialrotationasdescribedpreviously[30].Gapsbetweenthetibialresectionplaneandfemoralcondyleswererecorded.Jointgapdatawereusedtoplanthefemoralresectionsandfinaljointbalance.Femoralresectionswereperformedusingtheroboticsplatform.TibialboundariesandfemoralcomponentalignmentsimulationsAllroboticdatawerethenimportedinRtosimulateresectionsaccordingtoafunctionalalignmenttechniquewithtibia‐firstbalancingunderdifferentboundaryconditionsforthetibialresection[22,35].Ananatomictibialresectionwasvirtuallyperformedfirst,byaligningthetibialresectiontothenativeMPTA,aftercorrectingforboneandcartilagewearasdescribedfurtherbelow.TheresectionanglewasthenlimitedtobewithinoneofthreedifferentboundariestothetibialmechanicalaxisdefinedusingtheMPTA:(A)87–90°,(B)86–90°and(C)84–92°.Thefemoralcomponentwassubsequentlyrotatedin0.5°incrementsthrougharangeofcoronalandaxialalignmentstoachieveabalancedjointmediolaterallyinextensionandflexion.TherelativepositionofthetibiaandfemurcapturedintraoperativelywiththeBalance-Bot®wasusedtosimulatemedialandlateralbalanceinflexionandextension.Balancewasdefinedaswhenthemediolateralgapswerewithin0.5mmat10°and90°flexion.Figure3illustratestheworkflowofthesimulatedtibialandresultingfemoralcuts.FIGURE3Functionalalignmentwithtibia‐firstbalancing.Simulatedboneresectionsoffunctionalalignmentwithtibia‐firstbalancing,startingwithananatomictibialresectionfollowedbyaparalleldistalandposteriorfemoralresectiontoachieveequalmedialandlateralgapsunderconstantappliedtension.WearassumptionandCPAKcalculationThelateraldistalfemoralangle(LDFA)andMPTAwerecalculatedfromthelandmarkedpositionsofthedistalcondylesandproximalplateauxonthefemurandtibia,respectively,aftercorrectingforcartilageandbonewear.CartilageandbonewearwereestimatedbycalibratingtheLDFAandMPTAtoachieveasimilarcoronalplanealignmentoftheknee(CPAK)phenotypedistribution,arithmeticHKA(aHKA)andJLOtoanotherUS‐basedpopulationof1946kneesin973patientsfromtheosteoarthritisinitiative(OAI)multicenterdatabase[14,32].Priorstudieshavemeasuredcartilagethicknessbetween1and3mmwithavarietyoftechniques[7,29].Medialandlateralwearthereforewasappliedtovarus(＞2°varus)andvalgusknees(＞2°valgus),respectively,in1mmincrementsfrom0to4mmtominimisethedifferenceinCPAKdistributionbetweencohorts.Posteriorlateralwearinvalgusknees(definedas＞2°valgus)wasapproximatedwith2mmofwearasdescribedbyNametal.[19].Thedistributionofwear‐correctedanduncorrectedCPAKphenotypesiscomparedtoliteratureandusedtosubdividethepopulationtodeterminekneesatmostriskoffemoralcomponentvalgusandinternalrotationusingfunctionalalignmentwithatibia‐firstbalancingtechnique.StatisticsDescriptivestatisticsarepresentedasamean±standarddeviationforallcontinuousvariablesandaspercentagesforcategoricalvalues.ContinuousvariableswerecomparedusingMann–WhitneyUtestswithanassumptionthatthevalueswerenotnormallydistributed.Categoricalvalueswerecomparedusingtheχ2test.Inallcases,acriticalpvalueof0.05wasused.AllstatisticalanalyseswereperformedinR.RESULTSCoronalandsagittalalignmentThecohortreportedacoronalandsagittaldeformityof2.6±3.5°varus(Figure4)and2.3±5.3°flexion,respectively.Themeanleglengthwas770±93mmlong.FIGURE4Distributionofcoronalalignmentofkneesasmeasuredbytherobotic‐assistedplatform.Coronalalignmentdistributionofthekneesincludedinthestudy(meanvalueof2.6±3.5°showninthefigure).CPAKdistributionUncorrectedandwear‐correctedCPAKphenotypedistributioniscomparedtoaUSpopulationinTable1.Tominimisethedifferencesbetweenthewear‐correctedlandmarksandaUSpopulation,acorrectionof3mmonthemedialproximaltibiaand2mmonthemedialdistalfemurwasappliedinkneeswith＞2°varusHKA,andacorrectionof3mmonthelateralproximaltibiaand2mmonthelateraldistalfemurwasappliedinkneeswith＞2°valgusHKA.ThedistributionofkneeswithineachCPAKphenotypeisshowninFigure5.Afterwearcorrection,thecohortreportsonaveragenodifferenceinMPTA,a1°morevalgusLDFA(p＜0.01),a1.3°morevalgusaHKA(p＜0.01)anda1.9°moredistalapex(p＜0.01),comparedtothereferenceUSpopulation.TheseresultsaredisplayedinTable2.TABLE1DistributionofCPAKphenotypes(previouslydescribedUSpopulationversusdatafromthisstudy).FIGURE5Coronalplanealignmentoftheknee(CPAK)phenotypedistributionofincludedknees.CPAKphenotypedistributionofincludedknees:proportionofkneeswithineachtypeafterwearcorrectioninthisstudyshownincolours,comparedtopreviouslydescribedUSpopulationpercentagesinblack.aHKA,arithmetichip‐knee‐ankleangle;JLO,jointlineobliquity;LDFA,lateraldistalfemoralangle;MPTA,medialproximaltibialangle.TABLE2BreakdownofCPAKconstituentmeasurements(previouslydescribedUSpopulationversusdatafromthisstudy).CoronalandaxialfemoralalignmentThedifferenceintibialcoronalboundariesbetweenthethreegroups(A,BandC)asdescribedaboveisshownrelativetothedistributionofcorrectedMPTAvaluesinFigure6.Theproportionofkneeswithintheboundaryincreasedby2.4‐foldfromgroupAtogroupC(p＜0.01).ThepostoperativeLDFA(relativetothemechanicalaxis)requiredtoachieveextensionbalancedecreasedfromgroupAtoB(89.5±3.4°vs.89.1±3.5°,p＜0.05)andgroupAtoC(88.8±4.0°,p＜0.01).However,nostatisticallysignificantchangewasobservedbetweengroupsBandC.AlthoughsmallchangesinmeanpostoperativeLDFAwerefound,theproportionofkneeswitheitherhighfemoralcomponentvarusalignment(＞96°)orvalgusalignment(＜87°)relativetothemechanicalaxisincreasedfrom24.5%(n=170)ingroupAto26.5%(n=183)ingroupBand34.2%(n=237)ingroupCasillustratedbyFigure7.Alldifferenceswerestatisticallysignificant(p＜0.05).Femoralrotationrelativetotheposteriorcondylaraxis(PCA)decreasedfromgroupAtoB(0.1±3.1°vs.−0.3±3.2°,p＜0.05)andgroupAtoC(−0.6±3.8°,p＜0.01)butnotbetweengroupBandC(p=n.s.)toachievebalanceinflexion.Similarly,theproportionofkneeswitheitherhighfemoralcomponentexternalrotation(＞6°)orhighinternalrotation(＞3°)increasedfrom18.4%(n=127)ingroupAto23.7%(n=164)ingroupBand33.4%(n=231)ingroupCasdemonstratedbyFigure7.Alldifferenceswerestatisticallysignificant(p＜0.05).FIGURE6Medialproximaltibialangle(MPTA)distributionofincludedkneesandbreakdowninthreetibialboundarygroups.DistributionofcorrectedMPTAofallincludedkneeswiththemeanvaluedisplayedassolidblackline.ThethreeappliedtibialcutboundarycriteriaareshownasA:87–90°(dashedline),B:86–90°(dottedline),C:84–92°(dotted‐dashedline).Theproportionofkneesfallingwithineachboundaryisdescribed.Alldifferencesbetweenthethreegroupsarestatisticallysignificant(p＜0.01).FIGURE7Distributionofresultingfemoralcomponentcoronalandaxialalignmentwithinthethreedefinedgroups.Distributionoffemoralcomponentcoronal(toprow)andaxial(bottomrow)alignmentrequiredtoachieveakneebalancedwithin0.5mmineachtibialboundarygroup.Bluecolumnsindicate‘safe’femoralcomponentalignmentsandredcolumnsindicateoutlieralignmentswithalateraldistalfemoralangle＞96°or＜87°(toprow)and＞6°externalor＞3°internalrotation(bottomrow).Theproportionofkneeswhichreportafemoralcomponentalignmentwithbothhighvalgus(LDFA＜87°)andhighinternalrotation(＞3°)increased4‐foldfromgroupA(4.0%,n=28)toB(7.7%,n=53)andC(15.8%,n=109).Alldifferenceswerestatisticallysignificant(p＜0.01).CPAKphenotypeimpactedtheproportionofkneeswhichrequirefemoralcomponentvalgusalignmentandinternalrotationtoachievebalancewhenapplyingfunctionalalignmentwithatibia‐firstbalancingtechnique.CPAKtypeIandIIshowedanincreaseof12.9‐and2.9‐fold,respectively,intheproportionofkneeswithfemoralcomponentvalgusalignment(LDFA＜87°)andinternalrotation＞3°asshowninFigure8(p＜0.01).CPAKtypesIIItoIXeithershowednochangeinproportionorhadtoofewkneesforanalysis.FIGURE8Proportionofkneeswithfemoralvalgusandinternalrotationaccordingtocoronalplanealignmentoftheknee(CPAK)phenotypeandgroup.Comparisonofproportionofkneeswithalateraldistalfemoralangle＜87°andfemoralcomponentinternalrotation＞3°betweengroupsA,BandCwithineachCPAKphenotype.StatisticallysignificantdifferencebetweenallthreegroupswithinthatCPAKphenotype(p＜0.05).aHKAalignmentdidnotchangebetweengroups:(A)1.4±3.4°varus,(B)1.5±3.4°varus,(C)1.3±3.6°varus(p=n.s.).DISCUSSIONThemainfindingofthisstudyisthatextendingtheboundariesofthetibialcutfromanMPTAof87–90°upto84–92°whenbalancingaTKAbyusingfunctionalalignmentwithatibia‐firstbalancingtechniqueresultsinastatisticallysignificanthigherpercentageofkneeswithanLDFA＜87°andinternalrotation＞3°onthefemoralside.ThemostpronouncedinfluencecanbenoticedinkneescorrespondingtotypeIoftheCPAKclassification(a12.9‐foldincrease),followedbytypeII(a2.9‐foldincrease)[14].Extendingtheboundariesonthetibialsidedoesnotcauseanalterationinthemeanoverallalignment,femoralrotationorJLObutcreatesmoreoutliervalues.AnexplanationforthisphenomenoncanbesoughtintheMPTAasanarticlebyMicicoietal.showsthatitsmeanvaluegraduallydecreasesfromvalgustoneutraltovarusknees[16].Hence,applyingatibia‐firstbalancingtechniquewillcopyamorepronouncedobliquityonthetibialside,asisespeciallythecaseinvarusknees,tothefemoralsideinboththecoronalandaxialplane.Therisksofcausingexcessiveinternalrotationofthefemoralcomponenthavethoroughlybeendocumentedintheliterature[1,4,24].Possibledisadvantagesthathavebeendescribedincludetheinductionofhigherquadricepsrequirementsaswellasmorepronouncedalterationsinnativetibiofemoralkinematicsandloadtransfer[9,15,20,33].Themostalarmingconcernthathasbeenraisedistheheightenedpotentialtoinducepatellofemoralmaltracking,whichremainsoneofthemostfrequentcomplicationsafterTKAandanimportantcontributortopatientdissatisfaction[2,5,27].Despitetheaforementioneddangersassociatedwithfemoralcomponentmalrotation,anarticlebyFluryetal.showsnoinfluenceofexcessivefemoraltorsiononpatellofemoralpainandinstabilitywhentheoverallalignmentremainsneutralanddenervationofthepatellahasoccurred[8].Arecentstudydeniestheisolatedinfluenceoffemoralcomponentmalrotationonpatellofemoralsymptomsandpatientsatisfaction[6].However,creatingexcessivevalgusonthefemoralsideinadditiontotheincreasedinternalrotationcanberegardedasanevenmoreunfavourableconditionforpatellarstabilityasaclinicalstudyrevealedasignificantlyincreasedpatellarbonetraceruptakewithvalgusalignmentofthefemoralcomponent[31].Sofar,concernsassociatedwithexcessivefemoralvalgusandinternalrotationhavemostlybeenraisedinvalguskneesonwhichkinematicalignmentisapplied[21].Thisarticledemonstratesasimilarriskinvaruskneeswhenextendedboundariesareappliedonthetibialsidewhileusingfunctionalalignmentwithatibia‐firstbalancingtechnique.Thepreliminaryresultsobtainedwiththisalignmentstrategydonotconfirmanincreasedincidenceofpatellofemoralcomplicationsorsymptoms[23,35].However,theriskmightbedisguisedbythesmallsamplesizesandrelativelyshortfollow‐upinthesestudies.Severallimitationsofthisstudycanbeidentified.First,itsfindingsarebasedonaretrospectiveanalysisofintraoperativelyobtainedcalculations.Assuch,possiblyrelevantdatalikedemographicinformationandmeasurementsfromstandingfull‐legradiographsareunavailable.However,overallbiasisreducedbythelargesamplesize(692cases)andthepresenceofdatafrommultiplesurgeons.Second,theresultsfromthisstudyareonlyapplicableonPCLsacrificingknees.NosolidconclusionscanbemadeaboutcruciateretainingimplantsastheanteriorcruciateligamentandPCLweresacrificedineachcase.Third,thekneesweredividedbasedontheCPAKclassificationbymakingseveralkeyassumptionstocorrectforboneandcartilagewear.However,theseassumptionswerebasedonresultsobtainedfrompreviouslypublishedpaperssurroundingthistopic.Themostimportantshortcomingofthisstudyistheabsenceofclinicaldatatosupporttheconcernsraisedinthisarticle.ThedesignofthestudydoesnotaccommodatethecollectionofinformationconcerningtheinvivoeffectofcreatingfemoralcomponentvalgusandinternalrotationonTKAsurvivalandpatientsatisfaction.Thisstudyaidssurgeonsinidentifyingpotentialrisksofextendingthetibialboundariesinfunctionalalignmentwithatibia‐firstbalancingtechnique.Creatingmorefemoralcomponentvalgusandinternalrotationcouldpotentiallyleadtopatellofemoralmaltrackingandpatientdissatisfaction.FutureclinicalstudiesexaminingthisalignmentstrategyshouldattempttodefinesafeboundariesbydividingpatientsbasedontheirCPAKclassificationandspecificallyfocusingonpatellofemoraltrackingandpatient‐reportedoutcomemeasures.CONCLUSIONExtendingtheboundariesofthetibialcuttoanMPTAof84–92°whenusingfunctionalalignmentwithatibia‐firstbalancingtechniqueinTKAleadstoastatisticallysignificanthigherpercentageofkneeswithanLDFA＜87°and＞3°offemoralcomponentinternalrotation,especiallyinCPAKtypeIandII.

术中X线8个透视图像（AP,inlet,outlet,iliacoblique,obturatoroblique,“downthewing,”obturatoroutlet,anda“quad”view）发现骨盆内髋臼螺钉穿透（2021）IntraoperativeRadiographicDetectionofIntrapelvicAcetabularScrewPenetration:LessonsLearnedFromOurTraumaColleaguesWilsonJM,PfledererJA,SchwartzAM,FarleyKX,ReimerNB.IntraoperativeRadiographicDetectionofIntrapelvicAcetabularScrewPenetration:LessonsLearnedFromOurTraumaColleagues[J].ArthroplastToday,2021,8:226-230.转载文章的原链接1：https://pubmed.ncbi.nlm.nih.gov/33937463/转载文章的原链接2：https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8079330/AbstractBackgroundIntraoperativevascularinjuryduringtotalhiparthroplastyrepresentsacatastrophiccomplication.Acetabularscrewplacementrepresentsonepossiblemodeofinjury.Thepurposeofthisstudywastoevaluatetheutilityofvariousfluoroscopicviewsinthedetectionofintrapelvicscrewpenetration.MethodsAradiopaquepelvisSawbonesmodelwasinstrumentedwithahemisphericalacetabularcomponent.Fourintrapelvicquadrantsweredefined.Screwswereplaced,3ineachquadrant,andimagedsequentiallyat3depths:0mm,5mm,and10mmpenetrated.Eightfluoroscopicimages八个透视图像wereused:anteroposterior,inlet,outlet,iliacoblique,obturatoroblique,“downthewing,”obturatoroutlet,anda“quad”view.Threeblinded,independentsurgeonsevaluatedtheimagesforintrapelvicscrewpenetration.Imageswereanalyzedinisolationandasa“triple-shotseries”consistingofthe“quad,”obturatoroutlet,andiliacobliqueviews.Sensitivityandspecificityvalueswerethencalculated.ResultsInisolation,the“quad”viewhadthehighestsensitivityforscrewpenetration(62%).Thetriple-shotserieswasfoundtobe100%sensitiveinall4quadrantsfordetecting10mmofscrewpenetration.Thespecificityoftheserieswasfoundtobe100%inallquadrantsexceptfortheposteriorsuperiorquadrantwhereitwas67%.Interobserveragreementapproachedperfection(Kappa≥0.947)betweenallsurgeons(P＜.001)whenusingthe3-viewseries.ConclusionsThisstudyisthefirsttoassesstheuseoffluoroscopyinthedetectionofintrapelvicpenetrationoftransacetabularscrews.Wefoundthata3-radiographseriesprovidedasensitiveandspecificmetricforthedetectionofintrapelvicscrewpenetration.Keywords:Acetabular,Totalhip,Fluoroscopy,Screw,Radiology,ArthroplastyIntroductionTHAiswellestablishedasoneofthemostsuccessfulproceduresperformedinmedicine[1].Assuch,itisunsurprisingthattheincidenceofthisprocedurecontinuestoclimb[[2],[3],[4]].WhilethecomplicationprofileofprimaryTHAisgenerallyfavorable,[5]thepotentialforbothintraoperativeandpostoperativecomplicationspersists.Whilenotuniversallyused,surgeonsrangefromselectivetogeneralizedusageofacetabularscrewaugmentationofpress-fitfixation.Theplacementoftransacetabularscrewsisrecognizedtohaveassociatedrisk,andcloseattentionmustbepaidtotheextrapelvicacetabularquadrantinwhichtheseareplaced[[6],[7],[8],[9]].Amultitudeofcomplicationshavebeenreportedwithaberrantscrewplacementincludingarteriallaceration,[10]arterialpseudoaneurysmformation,[[11],[12],[13],[14]]bladderinjury,[15]andnerveirritation,[16]amongothers[17].Arterialinjury,inparticular,duringTHAhasbeenreportedtooccurin0.08%-0.3%ofcases[[18],[19],[20],[21]].Inrecentyears,thedirectanteriorapproach,anditsassociateduseoffluoroscopy,hasbeenpopularized[22].Fluoroscopyisusefulinevaluatingcomponentposition[23,24]andmayalsobeusefulinassessingtrajectoryoftransacetabularscrews.However,theeffectivenessofintraoperativefluoroscopytodetermineacetabularscrewpositionhasnotbeenassessed.Priorinvestigationsregardingtheradiographicevaluationofacetabularscrewshavebeenlimitedinboththeviewsevaluatedandrelianceonformalradiography—makingthegeneralizationtointraoperativeevaluationdifficult[25].GiventheimplicationsofaberrantscrewpenetrationandthepotentialforcatastrophicacuteandchroniccomplicationsafterTHA,thediagnosticcapacityoffluoroscopymustbeclarified.Therefore,thepurposeofthisstudywastoevaluatethesensitivityandspecificityoffluoroscopicdetectionofintrapelvicacetabularscrewpenetration.Wehypothesizedthata3-imageserieswouldprovideasensitiveandspecificmeasure.MethodsSawbonesmodelForthepurposesofthisstudy,aSawbonespelvismodelwitharadiopaquecoatingwasused(Sawbones,SKU:1301-212;APacificResearchCompany,VashonIsland,WA).Theacetabulumofthepelviswasthendividedinto4quadrantsaspreviouslydescribedbyWasielewskietal.[8].Inbrief,wedrewalinefromtheanteriorsuperioriliacspinethroughthecenteroftheacetabulumandthenasecondlineperpendiculartothisfirstlinethatintersectedthefirstatthecenteroftheacetabulum,aspreviouslydescribed[8].Becausewewereprimarilyinterestedintheintrapelvicregionofscrewpenetration,wetranslatedthesewell-describedquadrantstotheintrapelvicanatomy.Inordertodothis,wefirstdrilledasmallholethroughthequadrilateralsurfaceatthecenteroftheacetabulum(atthecenterasdeterminedpreviously)tomarktheintrapelviccenteroftheacetabulum.Wedeterminedthecorrespondingintrapelvicquadrantsbydrawingalinefromtheanteriorsuperioriliacspinetothecenteroftheacetabulumandalineperpendiculartothisfirstline.WethenreamedtheacetabulumaswouldbeperformedinastandardTHA.Aftertheappropriatesizeandmedializationwasachieved,anuncementedhemisphericalacetabularcomponent(TridentII;StrykerOrthopedics,Mahwah,NJ)wasimpactedintopositionusingline-to-linepress-fittechnique.ExperimentaltrialsWethendrilledforandinsertedscrewsintoeachquadrantinasequentialmanner.Atotalof3screwswithindependenttrajectorieswereplacedineachquadrant.Thesescrewswereinsertedsequentially(1perfluoroscopictrial)tofacilitateeasierradiographicanalysis.Adepthgaugewasusedtomeasureidealscrewlength,and3screwlengthiterationswereconductedwitheachdrilledscrewpath.Forthefirst,thescrewwasplacedsuchthatthetipofthescrewwasflushwiththeintrapelvicbonysurface;thesecondwasplacedwith5mmofpenetration;andthethirdwasplacedwith10mmofpenetration.Penetrationwasmeasureddirectlyalongthescrewtrajectory,usingarulertoconfirmeachscrewhadpenetratedtheexactdesiredamount.Foreachscrewtrial,8radiographicimageswereobtained.Theseincludedthefollowing:anteroposterior(AP),inlet,outlet,obturatoroblique,iliacoblique,“downthewing,”“quadview,”andobturatoroutlet,aspreviouslydescribed[26].The“downthewing”(DTW),“quad,”andobturatoroutletviewsarecombinationimagesfrequentlyusedinpercutaneouspelvisandacetabularfixationprocedures.Inbrief,thequadviewprovidesanimagelookingdirectlyparalleltothequadrilateralsurfaceandisusefulfordetectingmedialwallpenetration.Thistypicallyisacombinationofsomeinletandsomeobturatoroblique.Theobturatoroutletisastandardobturatorobliquebutwiththefluoroscopytubetiltedintoanoutletorientation.The“downthewing”viewallowsthesurgeontovisualizethemedialandlateralcorticesoftheiliacwingandisobtainedbyrotatingthefluoroscopymachineaboutthepatient,towardthesideofinterestwiththeadditionofsomeinlet[26].Thepreciseanglesatwhichtheseviewsareobtainedwillvaryslightlyfrompatienttopatient(Figure1).Figure1Exampleofthetripleshotseriesofascrewplacedintheanterior-superiorquadrantwith10mmofscrewpenetration.(a)Iliacoblique,(b)quadview,and(c)obturatoroutlet.Thephotosofthemodelpelvistakenatthetrajectoryofthefluoroscopytubetoobtain(d)iliacoblique,(e)quadview,and(f)obturatoroutlet.Aftereachscrewwasinsertedandconfirmedtohavereachedthedesireddepthofpenetration,eachoftheaforementionedviewswasobtainedusingafluoroscopicmachineinastandardizedfashion(C-armfluoroscopicmachinemodel:OECEliteCFD;GEHealthcare,Chicago,IL).Thatis,eachscrewtrajectorywasimagedatotalof24times(8imagesat3differentdepths).Therefore,intotal,72imageswereobtainedofscrewsineachquadrant,and288imageswereobtainedandanalyzedintotal.Eachimagewasnumbered,andalogwaskeptsuchthatakeywasavailabledenotingthequadrant,view,andscrewpenetrationdepthofeachimage.Thereafter,3independentandblindedorthopedicsurgeonswereaskedtoevaluateall288imagesasseparateentitiesandbinarilydecidewhethereachimagescrewhaddefinitivelypenetratedtheintrapelvicbone.Weadditionallycompileda3-viewseriesofeachscrew(henceforthreferredtoasthe“tripleshotseries”)andaskedeachsurgeontoagainbinarilyanalyzethescrewpenetrationusingall3viewsinconjunction.Thisseriesconsistedoftheobturatoroutlet,iliacoblique,andquadviews(Figure1).Theseanswersweretabulatedcategorically.Intraobserverandinterobserverreliabilitieswerethencalculated.Inrarecasesofsurgeondisagreement,thesurgeonsundertookposthocreviewofthediscrepantimagetogetherandcametoaconsensusconclusion,perprecedence[24].Sensitivities,specificities,negativepredictivevalue,andpositivepredictivevalueswerethencalculated.Thesewerecalculatedforeachradiographicviewindividuallyaswellasforthe3-viewcombination.Subgroupanalysiswasperformedforeachquadrant.AllanalyseswereconductedusingMicrosoftExcel(MicrosoftCorporation,Redmond,WA).ResultsIndividualradiographicviewsInisolation,nosingleradiographicimagewasparticularlysensitive.Whenconsideringallscrews(inall4quadrants),the“quad”viewhadthebestsensitivityat0.62atboth5and10mmofscrewpenetration.Theiliacobliqueviewwastheleastsensitiveview(0.08and0.23at5and10mmofscrewpenetration,respectively).Nearlyallviewsweremoresensitiveforthedetectionofscrewpenetrationat10mmthanat5mm.Thequadviewwastheleastspecific(0.85),whiletheAP,DTW,Iliacoblique,andinletviewsallwere100%specific(Table1,Table2).Interobserverreliabilitywasfoundtobehighbetweenreviewers(Kappastatistic≥0.814,P＜.001).Table1Diagnosticperformanceofeachradiographicviewbypenetrationdepth.AP,anteroposterior;DTW,downthewing.Table2Diagnosticperformance—allscrewdepthscombined.AP,anteroposterior;DTW,downthewing;PPV,positivepredictivevalue;NPV,negativepredictivevalue.TripleshotseriesWhenourstandardizedtripleshotserieswasanalyzedasasinglemetric,thesensitivityandspecificityweresignificantlyimproved.Whenanalyzedbyquadrant,thetripleshotserieswasfoundtobe100%sensitiveinallbuttheposteriorinferior(67%)andposteriorsuperior(86%)quadrants.At5-mmscrewpenetration,theserieswas100%sensitiveintheanteriorquadrants,butonly33%sensitiveintheposterior-inferiorquadrant,and75%sensitiveintheposteriorsuperiorquadrant.At10-mmscrewpenetration,thetripleshotserieswas100%sensitiveandspecificinallbuttheposteriorsuperiorquadrant,whereitwasonly67%specific(Table3).Table3Diagnosticperformanceof3-viewradiographicseries.PPV,positivepredictivevalue;NPV,negativepredictivevalue.DiscussionIntrapelvicacetabularscrewpenetrationduringTHAisknowntobepotentiallyhazardouswithseriousandoftenimmediateassociatedcomplications[[10],[11],[12],[13],[14],[15],[16],[17],[18]].Whilemispositionedacetabularscrewstypicallyhaveacuteorsubacuteimplications,intrapelvicscrewpenetrationmayalsobeproblematicremotefromtheindexprocedure[14].Therefore,immediateidentificationofthiserrorisparamount.Withtherisingpopularityofintraoperativefluoroscopy,directfluoroscopicconfirmationofscrewplacementholdspromise.However,theliteratureregardingtheoptimalradiographicviewtodetectintrapelvicscrewpenetrationissparse[25].Theresultsofthisstudyindicatethatthenosingleradiographicviewisparticularlysensitiveforintrapelvicscrewpenetration.However,ofthe8studiedradiographicviews,themostsensitiveviewwasthe“quad”view.Whilethisviewachieved62%sensitivityforthedetectionof5and10mmofscrewpenetration,ithadthelowestspecificityofstudiedviewsat85%.Weanecdotallyfoundthatthequadviewwasbest,asmaybeanticipated,atdetectionofmedialwallpenetration.However,thisviewstruggledtodetectsupra-pectinealscrewpenetrationandsometimesprojectedpenetrationonscrewsthatwereintraosseous,butwithatrajectorytowardtheischialspineorgreatersciaticnotch.Therefore,a3-viewseriestoaddresstheseweaknesses(obturatoroutlet,iliacoblique,andquadviews)waschosenandevaluated.Thistripleshotseries,inourmodel,wasfoundtobe100%sensitivefor10mmofintrapelvicscrewpenetrationinall4acetabularquadrants.Thiscombinationwasalsofoundtobe100%specificinallbuttheposteriorsuperiorquadrant,whereitwas66%specific.Therefore,weproposeour3-viewseriesasaclinicallyapplicableintraoperativeassessmenttooltodetectclinicallymeaningfulintrapelvicscrewpenetration.Whilemanyofthecombinationviewsassessedinthisstudyarewellknownintheorthopedictraumaoperatingroomwheretheyarefrequentlyusedfortheplacementofpercutaneousscrews,[26]theutilityoftheseradiographicviewshasnotbeenassessedinarthroplasty[25].Priorliteraturehasfocusedonthecharacterizationofscrewpositionbeforeundertakingrevisionarthroplasty[25,27].Thesestudieseitherfocusedoncomputertomographyscans[27]orassessedonlytraditionalpelvicradiographicviews(AP,inlet,outlet,judets)[25].Inaddition,inonestudyassessingplainfilmradiography,metallicacetabularcomponentswerenotused,limitingitscontemporarygeneralizability[25].Finally,giventhepopularityoftheanteriorapproach,[22]andtheassociatedincreasingpopularityofintraoperativefluoroscopy,fluoroscopicevaluation(asopposedtoplainfilmradiography)isincreasinglyclinicallyrelevant.Giventhepotentialimplicationsofmispositionedacetabularscrews,weproposethatourtripleshotseriesbeusedintraoperativelytoconfirmacetabularscrewposition.Thisseriesallowsforrapiddetectionofscrewpenetrationintothetrueandfalsepelvises,usingthequadandobturatoroutletviews,respectively.Theadditionoftheiliacobliqueviewallowsforthedetectionofscrewpenetrationinthegreatersciaticnotchandalsohelpsdifferentiatebetweentruemedialwallpenetrationandscrewsthataresimplydirectedtowardtheischialspine.Thisseriesconsequentlywas100%sensitiveforscrewpenetrationinall4intrapelvicquadrantsat10mmofscrewpenetration.Ofnote,inbothanteriorquadrants,wealsofounda100%sensitivityforthedetectionofeven5mmofscrewpenetration.Theacetabularquadrantsystemiswelldescribed,andeachquadrantcontainsdistinctat-riskanatomicalstructures[8].Wasielewskietal.havepreviouslyestablishedthattheposteriorquadrantsarerelativelysafewhencomparedtotheanteriorquadrantswheretheexternaliliacandobturatorneurovascularbundleareatriskinthesuperiorandinferiorquadrants,respectively[8,17,28,29].Injurytothesevessels,particularlytheexternaliliac,areknowntooccur[7,9,11,17,20,21,[30],[31],[32],[33],[34],[35],[36]].Otherauthorshavecharacterizedthedistancetodangerfromtheacetabularsurface,andintheanterioracetabulum,thisdistanceisnearlyuniversally＜20mm[28].Similarly,intheposteriorquadrants,theneurovascularat-riskstructuresareknowntobeover10mmfromthebonysurface,whichisincontrasttotheanteriorquadrantswheretheat-riskstructurescanbeindirectcontactwiththebone[29].Thisisimportantasitindicatesthatour3-viewseriesprovidesadequatedetectionofclinicallyrelevantscrewpenetration(≤5mmintheanteriorquadrantsand≤10mmintheposteriorquadrants).Itshouldbenotedthatstandardviews,particularlyanAPpelvis,donotprovidethesamebenefit.Whilethisstudyprovidesnovelinformationandisofclinicalutility,therearesomelimitationswhichmustbeconsidered.First,wechosetousearadiopaqueSawbonesmodel.WhiletheuseofSawbonesinorthopedicresearchiswellestablished,[37,38]thedegreetowhichourresultsarereproducibleinvivowherepelvicmorphologymayvarycannotbedefinitivelydetermined.Inaddition,weusedahemisphericalacetabularcomponentanddidnottestotheracetabularcomponentdesigns,normorecomplexreconstructionoptions.Similarly,thecomponentpositioninourstudywasconsistent(45degreesabductionand15degreesanteversion),anditispossiblethatvariablypositionedacetabularcomponentsmayhaveimplicationsregardingthesensitivityofsomeradiographicviews.However,thiseffectisanticipatedtobeminimal.Finally,whileweimagedourpelvismodelonanoperatingroomtable,thisclearlyrepresentsidealimagingconditionsgiventhelackofsofttissueoverlyingthebonypelvis.Theimplicationsofthisare,again,unknown.However,giventhequalityofimagesobtainedbycontemporaryfluoroscopymachines,this,too,isanticipatedtohavelittleeffectonourresults.ConclusionsInconclusion,standardradiographicviewsofthepelvis(AP,inlet,outlet,judetviews)havelowsensitivityforthedetectionofintrapelvicscrewpenetration—evenat1cmofscrewpenetration.Wefoundthatthemostsensitiveviewforscrewpenetrationwasthequadview,althoughthisachievedasensitivityofonly62%inisolation.Themostimportantfindingofthisstudyisthata3-viewseriesofradiographicimages—obturatoroutlet,quadview,andaniliacoblique—is100%sensitiveforthedetectionofintrapelvicscrewpenetrationat5mmintheanteriorquadrantsandat10mmintheposteriorquadrants.Therefore,weproposethattheseimagesbeobtainedwhenplacingtransacetabularscrewstoconfirmscrewpositionandruleoutintrapelvicscrewpenetrationintraoperatively.Theanticipatednegligibleadditionaloperativetimeaddedwithourreliableandrepeatableprotocolalleviatestheuncertaintyandunactionablenatureofpostoperativeradiography.Therefore,thiscanhelpminimizescrewmispositionandpotentiallypreventdevastatingcomplications.

传统标准的4个X线片：AP,inlet,outlet,judets_全髋关节翻修术中螺钉位置的影像学评价（2004）Radiographicevaluationofscrewpositioninrevisiontotalhiparthroplasty GalatDD,PetrucciJA,WasielewskiRC.Radiographicevaluationofscrewpositioninrevisiontotalhiparthroplasty[J].ClinOrthopRelatRes,2004(419):124-129. 转载文章的原链接1：https://pubmed.ncbi.nlm.nih.gov/15021142/ 转载文章的原链接2：https://journals.lww.com/clinorthop/fulltext/2004/02000/radiographic_evaluation_of_screw_position_in.20.aspx AbstractInjurytointrapelvicstructuresduringremovalofscrewsinrevisionacetabulararthroplastyisanuncommon,yetpotentiallyseriouscomplication.Bicorticalscrewsareatgreatestriskforcausinginjuryduringremoval,especiallyifdirectedtowardintrapelvicvesselsandnerves.Complicationscanbeminimizedwiththoroughevaluationofscrewpositionbeforerevisionsurgery.Astudyofsevencadavericpelveswasdonetodetermineifplainradiographicviewsprovideusefulinformationregardingscrewposition.Ineachpelvis,bicorticaltransacetabularscrewswerefixedinallacetabularquadrants15mmlongerthanthemeasureddepth.Afterward,anteroposterior,inlet,Judet,andcross-tablelateralradiographicviewswereobtainedandintrapelvicdissectionsweredone.Radiographsandintrapelvicdissectionswerecomparedtodeterminescrewposition.Wefoundthattheobturatorandiliacoblique(Judet)viewsweremostusefulindefiningscrewposition.Theiliacobliqueviewclearlyrevealedscrewsthatviolatedthequadrilateralsurfaceandthereforeweredirectedtowardtheobturatorvesselsandnerve.Theobturatorobliqueviewrevealedscrewsthatviolatedtheanteriorcolumnandthereforeweredirectedtowardtheexternaliliacvessels.Thelateralviewadditionallyclarifiedsuchscrewsbydetermininggeneralanteriororposteriordirection. THAisoneofthemostcommonlydoneorthopaedicoperations.In2000,morethan152,000primarytotalhipand31,000revisiontotalhiparthroplastiesweredoneintheUnitedStates.8ThenumberofrevisionandrerevisionTHAsdoneannuallycontinuestoincrease.Cementlessacetabularfixation,oftensupplementedbytransacetabularscrewfixation,iscommonlydoneinprimaryandrevisionacetabulararthroplasties.Lacerationofmajorvesselsduringremovalofexistingacetabularcomponentsatrevisionarthroplastyhasbeenreported.16,19Beforerevisionacetabulararthroplasty,evaluationofintrapelvicscrewshasbeenrecommendedtocharacterizetheirrelationshiptovitalintrapelvicstructurestoguidesurgicalapproachandplan.2Varioussurgicalapproacheshavebeendescribedforremovalofintrapelvicacetabularcomponents.7,28Traditionally,angiographyhasbeenusedtodefinetherelationshipbetweenvitalintrapelvicstructuresandtotalhipacetabularcomponentsbeforerevisionTHA.15,17Fehringetal2describedtheuseofcontrast-enhancedcomputedtomography(CT)scanningofthepelvisasamorecost-effectiveandlessinvasivemethodofdefiningtheserelationships.AlthoughCTmaybeconsideredappropriateinevaluatingpatientsknowntohavesignificantintrapelviccementorlong,protrudingscrews,itmaynotbecost-effectiveasascreeningtoolforallpatientsscheduledtohaverevisionTHA.Thegoalsofthecurrentstudyweretodetermine(1)ifvariousradiographicviewsprovideusefulinformationregardingtransacetabularscrewposition,specificallyinidentifyingbicorticalscrewsthataredirectedtowardvitalintrapelvicstructures,and(2)todefinewhichradiographicvieworcombinationsofviewsaremostusefulintheevaluationofscrewpositioninprospectivepatientsrequiringrevisionsurgery. MATERIALSANDMETHODSSevenpelvesfromcadaversofmatureadults,includingthemuscles,nerves,abdominalcontents,andthevasculaturefromtheumbilicustothemiddlepartofthefemur,wereobtained.Thetransectedabdominalaortawasligatedwithasutureandtheninjectedwith120mLradiopaquesiliconeinjectioncompound. Arterialfillingwasverifiedbyobservingtherunofffromthedistalpartofthefemur.Thefemoralarteriesthenwereligatedbilaterally.Thevenoussystemwassimilarlyinjected,exceptinretrogradefashiontoavoidvalvularobstructiontofilling.Thesciaticnervewaslocatedjustdistaltothequadratusfemorismuscle,andaflexiblemetalguidewirewasplacedinthenerve. TheacetabularportionofaTHAwasdonethroughananterolateralincision.Theacetabulumwasreamed,andaproperlysizedacetabularcomponent(HPG-I;Zimmer,Warsaw,IN)wasusedasatemplatefortransacetabulardrilling.Thecupwasorientedin45°abductionand15°anteversion.Therotationofthecomponentshellintheacetabulumwasrandom.Withtheuseofaguide,holesweredrilledperpendiculartothecupandthroughtheinnertableoftheacetabulum.Theacetabularcomponentwasremovedandtheholesweremeasuredwithadepthgauge.Carewastakentopenetratethemedialboneminimallyduringdrillingandmeasurement.Screwsthatwere15mmlongerthanthemeasureddepthwereplacedthroughtheacetabularboneandintothepelvistoallowobservationofthescrewsonradiographsandanatomicdissections. Standardanteroposterior(AP),inlet,cross-tablelateral,andiliacandobturatoroblique(Judet)radiographsweretakenofeachpelvisafterinjectionofthelatexandplacementofthescrews.Computedtomographyscansofeachpelvisweremadebeforedissectiontoensurethataccuraterelationshipswerepreservedbetweenthescrewsandtheundisturbedvasculature,followedbythree-dimensionalreconstructionwiththeCTdata. Preciseanatomicdissectionsthenweredone.Thepelveswereevisceratedthroughamidlineabdominalincision,withcaretakennottodisturbtheparietalperitoneumcoveringtheiliacvesselsandtheinnerpelvicwall.Theperitoneumwascarefullydissectedfromtheexternaliliacvessels,obturatorvessels,andtheobturatorinternusmuscleliningthemedialportionoftheacetabulum.Screwsthathadpenetratedtheinnertableofthepelveswereseeninrelationtothenervesandvessels(Figs.1,2).Thesescrewsweretracedbacktotheacetabularsurfacetodeterminetheirspecificacetabularoriginandlocationwithintheacetabularquadrantsystem(Fig.3).30Observationofthepelvicvasculature,theflexiblemetalguidewireinthesciaticnerve,andthetransacetabularscrewsontheradiographsandCTscansmadeitpossibletodeterminetheanatomiccourseandproximityofthesciaticnerveandoftheexternaliliac,obturator,superiorandinferiorgluteal,andinternalpudendalvesselstospecificscrewsplacedthroughtheacetabulum.  FIGURE1.:Thisviewofanintrapelvicdissectionshowsscrewsprotrudingthroughthemedialwallofthepelvis.Screws1through4piercethequadrilateralsurfaceandareseeninrelationtotheobturatornerve,artery,andvein(smallarrowhead).Screws5and6piercetheanteriorcolumnneartheexternaliliacvein(labeled)andartery(largearrowhead).Screw10isdirectedposteriorlyawayfromtheobturatorandexternaliliacstructures.  FIGURE2.:Thisisaclose-up,intrapelvicviewofthequadrilateralsurfaceofthesamepelvisinFigure1.Theobturatornerve,artery,andveinaredelineatedclearlyandclosetothequadrilateralsurface.Screws1through4piercethissurfaceandaredangerouslyclosetotheobturatorstructures.  FIGURE3.:ThequadrantsaredelineatedtoshowtheoriginofscrewsintheacetabularquadrantsystemfrompelvisNumberone.  RESULTSTheAPviewclearlyrevealedscrewsthatviolatedthemedialborderofthepelvis(Fig.4).InobservingtheAPradiograph,Screws1,2,4,and10allseemedtoviolatethemedialborder.Screws3and11seemedequivocal.However,onintrapelvicdissection(Figs.1,2),Screw3,inadditiontoScrews1,2,and4allexitedthequadrilateralsurfacedangerouslyclosetotheobturatornerve,artery,andvein.Screw10exitedposteriorlyawayfromtheintrapelvicvasculatureandScrew11wasnotobservedduringintrapelvicdissection,indicatingthatitwasburiedposteriorlyinthemusculature.Intheothersixpelves,severalscrewsthatwerefoundtoviolatethequadrilateralsurfaceondissectionseemedequivocalornegativeontheAPview.Therefore,inthisstudy,theAPviewwasfairlyinsensitiveforrevealingscrewsthatviolatedthequadrilateralsurfaceofthepelvis(Table1).  Table1:SummaryofRadiographicViews  FIGURE4.:Radiographs,dissections,andCTscanningallowidentificationofexactscrewpositions.Screws1,2,4,and10seemtoviolatethemedialcortexofthepelvis.Screws3and11seemequivocal.Ondissection,Screws1through4allpiercedthequadrilateralsurface.Inthisandtheothersixpelves,severalscrewspiercingthequadrilateralsurfaceseemedequivocalorunicorticalontheAPview.  Theinletview(Fig.5),however,inallsevenpelves,consistentlyrevealedallscrewsthatviolatedthequadrilateralsurface.TheinletviewinFigure5clearlyshowedScrews1through4allexitingtheinnertableofbone.Ondissection(Figs.1,2),thesescrewsallexitedthequadrilateralsurfaceincloseproximitytotheobturatorstructures.Whentracedtotheacetabularsurface,Screw1wascentrallylocated,andScrews2to4originatedintheanteriorinferiorquadrant(Fig.3).However,Screws5and6(Fig.1),whichexitedtheanteriorcolumninthevicinityoftheexternaliliacarteryandvein,didnotseembicorticalontheinletview.Therefore,theinletviewwasnotusefulforscrewsthatviolatedtheanteriorcolumn.Thedissectionsconsistentlyshowedthatscrewsthatviolatedthequadrilateralsurfacecamedangerouslyclosetotheobturatornerveandvessels.  FIGURE5.:Screws1–4and10violatethemedialcortexofthepelvis.Ondissection,Screws1–4piercedthequadrilateralsurfaceincloseproximitytotheobturatorvesselsandnerve.Inallsevenpelves,screwsthatviolatedthequadrilateralsurfacewereclearlybicorticalontheinletview.  Theiliacobliqueview(Fig.6A)inallsevenpelvesprovidedthesameinformationaboutscrewpositionastheinletview(Fig.5).Inallcasesofscrewpenetrationofthequadrilateralsurface,theiliacobliqueviewsclearlyshowedthatthesescrewswerebicortical(Screws1–4,Fig.6A).Also,Screws5and6seemedtobenegativeforviolation,asnotedwiththeinletview.Therefore,theiliacobliqueviewclearlyshowedthescrewsthatviolatedthequadrilateralsurfaceofthemedialpelvis,butwasnotusefulforrevealingbicorticalscrewsthatviolatedtheanteriorcolumn.  FIGURE6.A–B.:(A)Screws1–4and10areclearlybicorticalinthisview.Ondissection,Screws1–4piercedthequadrilateralsurfaceincloseproximitytotheobturatorvesselsandnerve.Inallsevenpelves,screwsthatviolatedthequadrilateralsurfacewereclearlybicorticalontheiliacobliqueview.Theinlet(Fig.5)andiliacobliqueviewsessentiallyprovidethesameinformation.(B)Screws5and6areclearlybicorticalonthisview.Ondissection,thesescrewsviolatetheanteriorcolumnandcomedangerouslyclosetotheexternaliliacvessels.Inallsevenpelves,screwsthatviolatedtheanteriorcolumnwereseenasbicorticalontheobturatorobliqueview.  Theobturatorobliqueview(Fig.6B)consistentlyrevealedscrewsthatviolatedtheanteriorcolumnofthemedialwallofthepelvis.Ondissection,screwsthatexitedtheanteriorcolumn(Screws5and6)wereincloseproximitytotheexternaliliacvessels(Fig.1),afindingthatwascorroboratedbytheothersixpelves.Whentracedbacktotheacetabularsurface,Screws5and6werefoundtooriginateintheanteriorsuperiorquadrant(Fig.3).Ontheobturatorobliqueview(Fig.6B),Screws5and6clearlywereseenasbicortical,protrudingthroughthemedialwallofthepelvis.Therefore,theobturatorobliqueviewwastheonlyradiographicviewtoshowclearviolationoftheanteriorcolumn,andthuswasusefulforshowingbicorticalscrewsviolatingthispelvicsurface. Finally,thecross-tablelateralview(Fig.7)wasvaluableinshowingthegeneralanteriororposteriordirectionofscrews.Screws4to6seemedtobeanteriorlydirected,Screws9to11seemedposteriorlydirected,whereasScrews1to3,7,and8seemedneutralorequivocal.Screwsthatseemedposteriorlydirectedonthecross-tablelateralviewwereeithernotobservedduringtheintrapelvicdissectionordirectedawayfromtheobturatorstructuresandexternaliliacvessels(Screw10,Fig.1).Incontrast,screwsthatseemedanteriorlydirectedonthecross-tablelateralviewoftenwerefoundondissectiontoviolatetheanteriorcolumn(Screws4–6,Fig.1).Equivocalscrewsonthelateralviewoftenviolatedthequadrilateralsurface(Screws1–3,Figs.1,2,and7)orweredirectedproximallyintotheiliacwing(Screws7and8).Screws7to11originatedintheposteriorquadrants(Fig.3).Therefore,thecross-tablelateralviewwasvaluableforshowingthegeneralanteriororposteriordirectionofscrews,butwasnotusefulforrevealingbicorticalscrewsviolatingspecificpelvicsurfaces.  FIGURE7.:Screws4–6aredirectedanteriorly.Screws1–3,7,and8seemneutralorequivocal.Screws9–11aredirectedposteriorly.Thecross-tablelateralviewprovidedgeneralinformationabouttheanteriororposteriordirectionofscrews,however,itdidnotnecessarilyshowifscrewswerebicortical.  DISCUSSIONTherelationshipofacetabularanatomytovitalintrapelvicstructuresnotvisibletothesurgeonhasbeendescribedtominimizeinjuryduringacetabulararthroplasty.3,13,30,31Theexternaliliacvesselsandobturatorvesselsandnervehavebeenshowntobeatgreatestriskforinjurybecauseoftheircloseproximitytotheanteriorcolumnandquadrilateralsurface,respectively.Althoughrare,injuriestotheseandotherintrapelvicstructures(urinarybladder)duringTHAandrevisionTHAhavebeenreported.Directtraumahasoccurredwithplacementofinstruments,21cementincorporationoftheiliacvessels,26removalofextrudedcementduringrevision,17–19,27andmigrationofprostheses.1,6,9,14,20,24,25,29Injuryfromscrewfixationusingtransacetabularscrews4,5,10–12andduringremovalofexistingacetabularcomponentsduringrevisionTHA16,19alsohasbeenreported. Becauseofthedangerofinjurytointrapelvicstructuresduringscrewremoval,preoperativeevaluationofpatientsscheduledtohaverevisionTHAshouldincludeassessmentofscrewposition.AlthoughCTscanningandangiographyhavebeendescribedasameansofdefiningscrewpositionrelativetovitalintrapelvicstructures,2,15,17costandefficiencyissuesprohibitthemfrombeingusedasscreeningtoolsforallpatientsscheduledtohaverevisionsurgery.AlthoughstandardplainradiographsarerecommendedbeforerevisionTHAtoassessbonelossandquality,leglengthdiscrepancies,andfortemplatingtochoosepropercomponents,22,23itwasreportedthatplainradiographsareinadequateforevaluationofscrewposition.2WefoundthistobetruewhenusingonlythestandardAPview,whichwasnotextremelyusefulforshowingbicorticalscrewsthatpenetratedthequadrilateralsurface.However,theinletandiliacobliqueviewswerevaluableforshowingbicorticalscrewsthatpenetratedthequadrilateralsurfaceandtheobturatorobliqueviewwasusefulforshowingbicorticalscrewsthatpenetratedtheanteriorcolumn.Thecross-tablelateralviewadditionallyclarifiedthesefindingsbydelineatinganteriorlyorposteriorlydirectedscrews. Theuseofplainradiographsforevaluationofscrewpositionhaslimitations.Althoughtheseviewsarestandardandobtainedwithpreciselydefinedtechniques,slightvariationsmayexistbecauseofoperatorvariability.Additionally,radiographswillnotdefinetheexactrelationshipoftransacetabularscrewsrelativetospecificintrapelvicstructures.Forthesereasons,radiographicviewsaremoreappropriateforuseasscreeningtoolstodefinepatientswhomaybeatriskforinjuryduringscrewremoval;thatis,patientshavingbicorticalscrewsthatviolatethequadrilateralsurfaceoranteriorcolumn.PatientswhoaredefinedaspotentiallyatriskthenshouldhavemorespecifictestingsuchasCTscanningorangiographytomoreclearlydelineatescrewpositionrelativetospecificintrapelvicstructures,toguidesurgicalapproachandplan.2,7,15,17,28Thecurrentstudyusedfewcadavers,perhapsnotaccuratelyreflectingthetruerangeofanatomicvariation.Finally,thecurrentstudyevaluatedtransacetabularscrewsthatwereprotrudingintothepelviscavity15mmbeyondthemedialwall.Therefore,itisnotknowniftheseradiographicviewswillbeasusefulinshowingbicorticalscrewsthatprotrudefromthemedialwalllessthan15mm. Wasielewskietal30definedthesafeacetabularquadrantsforscrewplacementduringprimaryandrevisionTHA.Screwsfixedperpendiculartotheacetabularsurfaceintheposteriorquadrantsgenerallyareconsideredsafeandthosefixedperpendicularintheanteriorquadrantsgenerallyareconsideredunsafe.Ourdatacorroboratedthispreviousstudy.However,whenevaluatingprospectivepatientsforrevisionsurgery,screwsmaynothavebeenfixedperpendiculartotheacetabularsurfaceduringprimaryTHA.Forinstance,ascreworiginatinginthesafeposterosuperiorquadrantactuallymaypiercethequadrilateralsurfaceifdirectedanteriorly.Conversely,ascreworiginatingintheanteroinferiorquadrantactuallymaybeposteriorlydirectedintosafeareas.WhenusingJudetviewstoevaluatescrewpositioninprospectivepatients,bicorticalscrewpenetrationofthecriticalquadrilateralsurfaceoranteriorcolumnwillbeseenregardlessofscreworiginintheacetabularquadrantsystem. SurgeonswhodorevisionsurgerycommonlyusetheJudetviewstoevaluateacetabularbonelossandcolumnintegrityinprospectivepatients.Inthecurrentstudy,wefoundtheseviewskeyforscreeningpatientswithtransacetabularscrews,astheyprovidedthemostusefulinformationregardingthetwocriticalsurfacesofthemedialpelvis.TheiliacobliqueviewshowsscrewsthatviolatethequadrilateralsurfacetherebyplacingtheobturatorvesselsandnerveatriskwithscrewremovalduringrevisionTHA.Becausetheinletviewprovidesthesameinformation,itisnotnecessarytoobtainboth.TheobturatorobliqueviewshowsscrewsthatviolatetheanteriorcolumnandthereforeplacetheexternaliliacvesselsatriskwithscrewremovalduringrevisionTHA. Ifeithertheobturatororiliacobliqueviewsshowbicorticalscrews,thesurgeondoingtherevisionshouldobtainacross-tablelateralviewtodeterminegeneralanteriororposteriorscrewdirection.Ifclearlyanteriorlydirectedoncross-tablelateralviewscrewpositionrelativetospecificintrapelvicstructuresshouldbedelineatedusingCTscanning(asdescribedbyFehringetal2)orangiography,asthesescrewsmayviolatetheanteriorcolumn(Screws5and6,Figs.1,6B,and7).Ifneutralorequivocalonthelateral,thescrewsmayviolatethequadrilateralsurface,therebyalsorequiringadditionaldelineation(Screws1–3,Figs.1,2,6A,and7).Ifscrewsareclearlydirectedposteriorlyonthelateralviewtheylikelyaredirectedawayfromvitalintrapelvicstructures.Inthiscase,thesurgeonmayusediscretionregardingadditionaltesting. TheuseofsimpleplainradiographsmayprovidethesurgeonwhodoesrevisionTHAswithauseful,yetcost-effectiveandefficientscreeningtoolforallprospectivepatientswithtransacetabularscrews.Radiographsobtainedintheofficehavetheadvantageofbeingdirectlyreviewedbythesurgeon,atthetimewhenoperativeplanandimplantselectionarebeingdetermined.IfadditionaltestingwithCTscanningorangiographyisindicated,thesetestsmaybeaddedtothepreoperativeworkup.

多种因素影响高位髌骨的测量：滑车形态和膝关节矢状面平衡是髌股稳定性的决定因素（2024）DiverseFactorsAffectMeasurementofPatellaAlta:TrochlearMorphologyand SagittalKneeBalanceRepresentDefiningFactorsofPatellofemoralStability TanakaMJ.DiverseFactorsAffectMeasurementofPatellaAlta:TrochlearMorphologyand SagittalKneeBalanceRepresentDefiningFactorsofPatellofemoralStability[J].Arthroscopy,2024. 转载文章的原链接1：https://pubmed.ncbi.nlm.nih.gov/38484922/ 转载文章的原链接2：https://www.arthroscopyjournal.org/article/S0749-8063(24)00193-2/fulltext AbstractPatellaaltaisaknownriskfactorforpatellarinstability,contributingtoincreasedlengthchangesofthemedialpatellofemoralcomplex,andassociatedwithhighratesoffailureaftermedialpatellofemoralcomplexreconstruction.Distalizationthroughtibialtuberosityosteotomy(TTO)isasurgicaloptiontoaddresspatellaaltaduringpatellarstabilizationsurgery.However,distalizationhasbeenshowntohavegreatercomplicationratesthanothertypesofTTO,andthereforepreciseindicationsthroughproperassessmentsofpatellarheightareneeded.TheCaton-Deschampsindexisacommonlyusedmeasurementandisindependentofpatellartendonlength,allowingforassessmentofpatellarheightbeforeandafterdistalizingTTO.Additionaloptionsthatmayofferthisabilityarefemoral-basedmeasurementssuchasthepatellotrochlearindexandsagittalpatellarengagement,whichalsodonotrelyonthepositionofthetubercleandmaybetterrepresentthefunctionalengagementofthepatellofemoraljoint.However,femoral-basedmeasurementsdependontrochlearmorphology;theoretically,anadvantageoffemoral-basedmeasurementsisthattheymayreflectthefunctionalengagementofthepatellawithinthetrochlea.However,inknowingthecommonrelationbetweenpatellaaltaandtrochleardysplasia,thequestionbecomeswhetheranoverlapbetweenthepatellaandadysplasticproximaltrochleaonasagittalmeasurementtrulyrepresentsfunctionalengagementandstabilityofthepatellofemoraljoint.Measurementofpatellaaltaanddeterminingindicationsfordistalizationcanbeinfluencedbymeasurementtechniquesincludingmagneticresonanceimagingversusradiography,tibial-versusfemoral-basedmeasurements,andpositionalandmorphologicconsiderations.Increasedunderstandingoftrochlearmorphologyandsagittalkneebalanceastheyrelatetopatellaaltawillbeimportantfordefiningthefactorsthataffectpatellofemoralstability. Patellaaltaisaknownriskfactorforpatellarinstability.Biomechanicalstudieshavedemonstratedtheroleofpatellaaltaincontributingtoincreasedlengthchangesofthemedialpatellofemoralcomplex,1,2andclinicalstudieshavenoteditsassociationwithfailedmedialpatellofemoralcomplexreconstruction.3Distalizationthroughtibialtuberosityosteotomy(TTO)isasurgicaloptiontoaddresspatellaaltaduringpatellarstabilizationsurgery.However,distalizationhasbeenshowntohavegreatercomplicationsthanothertypesofTTO,4andtherefore,preciseindicationsthroughproperassessmentsofpatellarheightareneeded.TheCaton-Deschampsindexisacommonlyusedmeasurementowingtothebenefitthatthismeasurementisindependentofpatellartendonlength,therebyallowingforassessmentofpatellarheightbeforeandafterdistalizingTTO.Additionaloptionsthatmayofferthisabilityarefemoral-basedmeasurementssuchasthepatellotrochlearindexandsagittalpatellarengagement,whichalsodonotrelyonthepositionofthetubercleandfurthermoremaybetterrepresentthefunctionalengagementofthepatellofemoraljoint.Intheirstudy“NoCorrelationExistsBetweenTibial-andFemoral-BasedMeasurementsofPatellaAltainaPopulationWithChronicPatellofemoralPainorInstabilityUndergoingPatellaDistalization,”Snow,Singh,Rix,andHaikal5aimedtoassesstherelationbetweentibial-andfemoral-basedmeasurementsofpatellaalta.In71kneeswithpatellaalta,tibialmeasurementsincludingtheInsall-SalvatiindexandCaton-Deschampsindexwerecomparedwithfemoral-basedmeasurementssuchasthepatellotrochlearindexandsagittalpatellarengagement.Becausefemoral-basedmeasurementsaredependentontrochlearmorphology,theauthorsalsoconsideredtrochlearlength,asdeterminedbythelateralcondyleindex(LCI),andkneeflexionangleandtheirinfluenceonthesemeasurements.Snowetal.5foundnocorrelationbetweenthetibial-andfemoral-basedmeasurementsofpatellaalta.Additionally,theyreportednocorrelationofthefemoral-basedmeasurementswithtrochlearlengthasmeasuredbytheLCI.Weappreciatetheauthorsadditionallymentioningourstudyontrochlearlength,6inwhichweshowedthatadysplastictrochleahasincreasedlength,incontrasttothedecreasedlengthnotedpertheLCIinthestudybyBiedertetal.7Althoughtheseapparentdifferencesintrochlearlengthmaybedueinparttovariationsinmorphology,aswellaslocationofmeasurement,anotherfactorthatshouldbeconsideredistheanteriorizationoftheproximaltrochlea.This,incombinationwithtrochlearlength,canleadtotheappearanceofasupratrochlearprominence,whichisanotherknownfactorinpatellarinstability.Thetheoreticaladvantageoffemoral-basedmeasurementsisthattheymayreflectthefunctionalengagementofthepatellawithinthetrochlea.However,inknowingthecommonrelationbetweenpatellaaltaandtrochleardysplasia,8thequestionbecomeswhetheranoverlapbetweenthepatellaandadysplasticproximaltrochleaonasagittalmeasurementtrulyrepresentsfunctionalengagementandstabilityofthepatellofemoraljoint.Snowetal.5havetouchedonanimportanttopicasweconsiderthemultipleradiographicmethodsandmeasurementswecurrentlyusetoassesspatellaalta.Magneticresonanceimagingversusradiography,tibial-versusfemoral-basedmeasurements,andpositionalandmorphologicconsiderationscanallinfluencewhenandhowwemayconsideradistalizationproceduretoaddresspatellaalta.AlthoughtheLCImaynothaveinfluencedthefemoral-basedmeasurementsinthisstudy,increasedunderstandingoftrochlearmorphologyandsagittalkneebalanceastheyrelatetopatellaaltawillbeimportantfordefiningthefactorsthataffectpatellofemoralstability.Futurestudiesareneededtounderstandtheoptimalmeasurementtechniquetoreproduciblyidentifypatellaaltainkneesduringthemanagementofpatellarinstability.

骨性关节炎的2D与3DMRI临床应用与研究(2023)2Dversus3DMRIofosteoarthritisinclinicalpracticeandresearch WalterSS,FritzB,KijowskiR,FritzJ.2Dversus3DMRIofosteoarthritisinclinicalpracticeandresearch[J].SkeletalRadiol,2023,52(11):2211-2224. 转载文章的原链接1：https://pubmed.ncbi.nlm.nih.gov/36907953/ 转载文章的原链接2：https://link.springer.com/article/10.1007/s00256-023-04309-4 AbstractAccuratelydetectingandcharacterizingarticularcartilagedefectsiscriticalinassessingpatientswithosteoarthritis.Whileradiographyisthefirst-lineimagingmodality,MRIisthemostaccurateforthenoninvasiveassessmentofarticularcartilage.MultiplesemiquantitativegradingsystemsforcartilagelesionsinMRIweredeveloped.TheOuterbridgeandmodifiedNoyesgradingsystemsarecommonlyusedinclinicalpracticeandforresearch.Otherusefulgradingsystemsweredevelopedforresearch,manyofwhicharejoint-specific.Bothtwo-dimensional(2D)andthree-dimensional(3D)pulsesequencesareusedtoassesscartilagemorphologyandbiochemicalcomposition.MRItechniquesformorphologicalassessmentofarticularcartilagecanbecategorizedinto2Dand3DFSE/TSEspin-echoandgradient-recalledechosequences.T2mappingismostcommonlyusedtoqualitativelyassessarticularcartilagemicrostructuralcompositionandintegrity,extracellularmatrixcomponents,andwatercontent.Quantitativetechniquesmaybeabletolabelarticularcartilagealterationsbeforemorphologicaldefectsarevisible.Accuratedetectionandcharacterizationofshallowlow-gradepartialandsmallarticularcartilagedefectsarethemostchallengingforanytechnique,butwherehighspatialresolution3DMRItechniquesperformbest.Thisreviewarticleprovidesapracticaloverviewofcommonlyused2Dand3DMRItechniquesforarticularcartilageassessmentsinosteoarthritis. KeywordsMRI ·Knee ·Osteoarthritis ·Quantitative ·3Dimaging IntroductionOsteoarthritisisacommonwholejointdiseasewitha30–50%prevalence.Osteoarthritisaffectsapproximately240millionpeopleworldwideand32millionintheUSA[1,2].Typicalsymptomsincludejointpain,stiffness,lossofjointfunction,andprogressivelydecreasingmobility,contributingtosubstantialmorbidityandmortality.Thehallmarkofosteoarthritisisthedisintegrationoftheosteocartilaginouslayers,includingsubchondralbonealterationsandprogressivelossofarticularcartilage[3,4].Theroleofimaginginassessingosteoarthritisisbasedondiagnosis,grading,andpredictionofdiseaseprogression,aswellasstudyingthediseaseforresearchpurposes[5–7].Radiographyisthefirst-lineimagingmodalitytovisualizeandgradestructuralalterationsofosteoarthritisandtotrackdiseaseprogression[1].Computedtomographyofferscross-sectionalassessmentofarticularsurfacesbutcannotdirectlyvisualizearticularcartilage.Computedtomographymaybeobtainedinrecumbentnon-weight-bearingpositionoruprightwithweight-bearingtoimprovetheassessmentofjointspacenarrowing[8].MRIcandirectlyvisualizearticularcartilageinanoninvasivefashioninstagesbeforeradiographicallydetectablejointspacenarrowingoccurs.Duetothesuperiorsoft-tissuecontrastandhighspatialresolution,MRIalsoprovidesthemostaccuratewholejointassessment[9–14].ArticularcartilageassessmentswithMRImayincludediagnosingmorphologicandcompositionalalterationsandmonitoringclinicaloutcomesofsurgicalandconservativetreatmentsinclinicalpracticeandresearch[11,12].CurrentMRItechniquesfocusonthemorphologicandcompositionalevaluationofarticularcartilageusingtwo-dimensional(2D)andthree-dimensional(3D)pulsesequencetechniques[11,15–18].Thisreviewarticleprovidesapracticaloverviewofcommonlyused2Dand3DMRItechniquesforarticularcartilageassessmentsinosteoarthritis. ArticularcartilagestructureThearticularcartilageofdiarthrodialjointsrepresentsahighlyspecializedconnectivetissuewithlimitedabilitiesforintrinsichealingandrepair[3].Theprincipalpurposeofhealthyarticularcartilageistoprovideasmoothandlubricatedsurfaceforarticulationandreducethefrictionalcoefficientassistinginthetransmissionofloads[19].Thehealthyhyalinearticularcartilage(meanthickness:2–4 mm)comprisesanabundantextracellularmatrixandsparselydistributedmetabolicallyactivechondrocytes[9].Infurtherdetail,theextracellularmatrixconsistsofwater(ca.70–80%ofthewetweightofarticularcartilage),collagen,proteoglycans,andtoalesserextent,non-collagenousproteinsandglycoproteins[11].Thesecomponentsprimarilyaimtoretainwaterwithintheextracellularmatrix,maintainingthespecializedmechanicalpropertiesofhyalinearticularcartilage[9].Thearticularcartilagecanbedividedintofourdistinctzones—superficial(tangential)zone,middle(transitional)zone,deepzone,andcalcifiedzone—witheachzonehavingadifferentcompositionandpurpose[9].Thesuperficialzoneisathinlayeroftightlypackedcollagenfibersalignedparalleltothearticularsurfaceandflattenedchondrocytesatahighdensity[3,9,19].Thechondrocytesinthesuperficialzoneprimarilysynthesizehighcollagenlevels,makingitthezonewiththehighestwatercontent[3,19].Thepurposeofthiszoneistoprotectthedeeperlayersfromtheshearforcesduringjointarticulation[9,19].Whendisrupted,themechanicalpropertiesofcartilagechange,contributingtothedevelopmentofosteoarthritis[19].Themiddlezoneconnectsthesuperficialzonewiththedeepzoneandcomprisesobliquelyorganizedcollagenfibrils,proteoglycans,andchondrocyteswithsphericalmorphologyatalowdensity.Themiddlezoneprotectsagainstshearandcompressionstress[9].Thedeepzoneisperpendiculartothearticularsurfacecollagenfibrilswithchondrocytesarrangedincolumnsparalleltothecollagenfibers.Thiszoneprovidesthehighestresistancetocompressiveforcesduetotheperpendicularorientationofthecomponents.Furthermore,thedeepzonehasthehighestamountofproteoglycanandthelowestconcentrationofwater[3,9].Thecalcifiedzoneanchorsthecollagenfibrilsofthedeepzonetothesubchondralbone,thus,fasteningthecartilagetotheadjacentbone[3,9].Theabsenceofbloodvessels,lymphatics,andnerveswithinthehyalinearticularcartilagehinderitsregenerationcapabilities[3,9,10,19].ThelayeredappearanceofthearticularcartilageonMRIisassumedtobebasedonthedifferenceinwatercontentbetweenthesuperficialanddeeplayers(Fig. 1)[10].  Fig.1Coronalintermediate-weightedturbospin-echoMRimage(repetitiontime,4000 ms;echotime,35 ms)ofthekneedemonstratesthecharacteristicgradientlayeringofintactarticularcartilageonintermediate-weightedturbospin-echoMRimages.Thegradientlayeringischaracterizedbylowsignalintensitiesnearthetidemark(arrows),graduallyincreasingsignalintensitiestowardsthearticularsurface,andabsentfluidbrightsignalswithinthearticularcartilagesubstance.Thedarklinealongthearticularcartilagesurfaceindicatesthelaminasplendens  MorphologicarticularcartilageassessmentMRIhasmultiplesemiquantitativegradingsystemsfordegen-erativeandtraumaticcartilagelesions[20],mostlyderivedfromarthroscopicclassifications.ThemostcommonclassificationsusedinclinicalpracticearethemodifiedOuterbridgegradingsystem,themodifiedNoyesgradingsystem,andtheInternationalCartilageRepairSociety(ICRS)classification(Table 1)[11–13,21,22].Thetwomajorcharacteristicsassessedinallthreegradingsystemsarethedepthofcartilagelesionsandsubchondralboneinvolvement[22,23].  Table1CommonMRIgradingsystemsforarticularcartilagedefects[21–24]   MRIassessmentofarticularcartilagemaybeperformedwithdifferentfieldstrengths[25,26].However,3 TMRIwithdedicatedsurfacecoilsandoptimizedpulsesequencesmaybepreferredasitpermitshigherspatialresolutionwithclinicallyfeasibleacquisitiontimes,whichmayimprovethediagnosticperformanceofdetectingearly,small,andlow-gradearticularcartilage,suchasfissuresandcoaptedflapsthataresubjecttothepartialvolumeaveragingeffects[10,12,27].MorecomplexMRIgradingsystemsforcartilagedefectshavebeendevelopedforresearch[13,28],someofwhicharejoint-specific,includingtheKneeOsteoarthritisScoringSystem(KOSS),MRIOsteoarthritisKneeScore(MOAKS),andWholeOrganMRScore(WORMS)[13,28–30]. 2DMRI2DFSE/TSEpulsesequencesThemostoftenusedMRIsequencestoevaluatejointabnormalitiesinclinicalroutineandforresearchpurposesare2DT1-,T2-,protondensity-,andintermediate-weightedimagingsequencesthatcanbeacquiredwithorwithoutfatsuppression[11,31–33].Comparedtostandard2Dspinechosequences,2D-FSE/TSEtechniquesacquiremultipleechoesduringeachrepetition,thussignificantlyshorteningexaminationtime[11,34].Furthermore,2DFSE/TSEpulsesequenceshaveexcellentSNR,highin-planespatialresolution,andversatiletissuecontrastthatallowsfordifferentiationbetweenanatomicalstructures[11,27,31,32,35].Morerecentlyinducedaccelerationtechniques,deeplearning-basedimagereconstructionandsuper-resolutionalgorithmspermitfasterMRacquisitionswithhigherspatialresolutionacquisitionsinlessthan1 minperpulsesequence(Fig. 2),enablingcomprehensivejointexaminationsinunder5 minacquisitiontime[36–40].  Fig.2Deeplearningsixfoldacceleratedcoronalfat-suppressedintermediate-weighted(A)andsagittalfat-suppressedT2-weighted(B)MRimagesofthekneedemonstrateamedialmeniscustear(arrows)oftheposteriorsegment  Alimitationof2DFSE/TSEsequencesistheacquisitionofrelativelythicksliceswithanisotropicvoxelsandsmallsectiongapsinbetweentheslices,whichmaypromotethepartialvolumeeffectandmaythereforeobscurefinedetail,suchasarticularcartilagefissuresandlow-gradeshallowsurfacedefects[11,32,35,41,42].Additionally,reformatting2Ddatasetsintootherplanesoftenresultsinaconsiderablelossofimagequality,thusrequiringseparateacquisitionsforeachplane[11,41]. T1weightingT1weightatfastandtubers.Pulsesequencestypicallyuserepetitiontimesoflessthan600 msandminimizeechotimes,typicallybelow10 ms.T1-weightedmayalsobeachievedwithgradingpulsesequences;however,inourpractice,wereserveT1-weightedgradingechoMRItovisualizecontrastenhancementafterintravenousandintra-articulargadoliniumconversationinjections.Theconspicuityofcontrastenhancementmaybeincreasedbysubtractingnon-fat-suppressedpre-andpost-contrastimagesorapplyingfatsuppressiontopost-contrastimages.Dixontechniquesadditionallyderivefat-onlyimages,whicharemostfat-specificduetotheinherentfat–waterseparation,butprovidelessstructuraldetailthanconventionalT1-weightedMRimages[43,44].AsT1-weightedMRimagesaregenerallyfatbutnotfluid-sensitive,theyfunctionprimarilyasananatomicalpulsesequenceforsubchondralbone,osteophytes,andjointbodies[45].Non-fat-suppressedT1-weightedMRimagesareusefulfordetectingsubtlesubchondralfracturesandsubchondralsclerosis[28].ThenormalappearanceofthearticularcartilageonT1-weightedMRimagesishomogenousisointensity[34,46];however,T1-weightedMRimagesoftenmissarticularcartilagedefectsandmeniscustearsduetoinsufficientcontrastresolutionbetweenarticularcartilageandsynovialfluid[11].Forthisreason,T1-weightedMRimagesoftenfailtodepictotherabnormalities,suchasligamentousandmeniscustears(Fig. 3)[11].  Fig.3CoronalT1-weighted(A)andfat-suppressedT2-weighted(B)turbospin-echoMRimagesoftheknee,illustratingthelimitationsofT1-weightedMRimagesforextra-osseouslesiondetection.Anearverticallongitudinaltearofthelateralmeniscustear(arrows)iswellseenonthefat-suppressedT2-weightedturbospin-echoMRimage(B)butomittedontheT1-weightedturbospin-echoMRimage(A)  Inourpractice,weutilizeT1-weightedMRimagesforprotocolsdesignedtodetectmarrowinfiltratingandmarrow-replacingprocesses,suchasosteomyelitisandbonemarrowneoplasms,respectively,butlessoftenforfracturesandosteoarthritis.Inourexperience,thecombinationofnon-fat-suppressedintermediate-weightedpulsesequencesandfat-suppressedfluid-sensitiveT2-weightedandSTIR2DFSE/TSEpulsesequencesarehighlyaccurateforthecharacterizationofsubchondralbone,detectionofjointbodies,andgradingofcentralandmarginalosteophytes,thuseffectivelyobviatingtheneedforadditionalT1-weightedpulsesequences. Protondensityand intermediateweightingProtondensity-andintermediate-weighted2DFSE/TSEpulsesequencesproducefluid-sensitiveMRimagesthatarehighlyaccuratefordetectingandcharacterizingarticularcartilagelesions.Eachpulsesequenceischaracterizedbylongrepetitiontimestypicallygreaterthan3500to4000 msanddiffersbytherangeofechotimes.Protondensityweightingtypicallyincreaseswhenechotimesarekeptaround20 ms,whereasintermediateweightingistypicallyachievedwithechotimesaround35–45 ms(Fig. 4).Nativeprotondensity-andintermediate-weighted2DFSE/TSEpulsesequenceshavethemostdiversifiedcontrastresolutionofarticularcartilageandmusculoskeletalsofttissues.Eitherpulsesequencemaybeobtainedwithoutorwithspectralfatsuppression.Sincethereistypicallynofatwithinarticularcartilageandforbetterdifferentiationoftheosteochondraljunction,weobtaineitherwithandwithoutfatsuppression(Fig. 5).  Fig.4Axialturbospin-echoMRimages(repetitiontime,4500 ms)ofthepatellademonstratethespectrumofprotondensity-weighted(PD),intermediate-weighted(IW),andT2-weighted(T2W)articularcartilagecontrast.Withincreasingechotimefrom17to58 ms,theintrinsicsignalintensityofthearticularcartilagedecreases,whereasthecharacteristicgradientlayersignatureincreases.Anoptimalbalancebetweenintrinsicarticularcartilagesignal,characteristicgradientlayering,andfluidbrightnessmaybeachievedaroundanechotimeof33 ms,whichconstitutesintermediate-weightedimagecontrast  Fig.5CoronalandsagittalturbospinechoMRimagesoftheknee,demonstrating differencesinarticularcartilagevisibilitywithoutandwiththatdepression.Coronalnative(A)andfat-suppressed (B)intermediate-weightedMRimagesdemonstratepartialthicknessarticularcartilagelossofthelateralfemoral condyle(arrowsinAandB),whichmaybebetterseenwithoutfatsuppression(A)andoverestimatedwithfat suppression(B).Sagittalnative(C)intermediate-weightedandfat-suppressedT2-weighted(D)MRimages demonstratefocalfull-thicknessthicknessarticularcartilagelossoftheposteriorfemoralcondyle(arrowsinCand 10.1007/s00256-023-04309-4 D),whichmaybebetterseenwithoutfatsuppression(C)andoverestimatedwithfatsuppression(D)  Protondensity-weightedMRimageshavehighercontrastresolutionatthetidemark,nearthecalcifiedtononcalcifiedarticularcartilagejunction,whereasintermediate-weightedMRimageshavehigherfluidspecificity(Fig. 4).Wepreferintermediate-weightedMRimageswithanechotimeofaround30–40 msasthecharacteristicgradientlayeringaidsinidentifyinghealthyarticularcartilage,andthehigherfluidconspicuityaidsindetectingarticularcartilagedefects[11,47].Longerechotimesreducemagicangleeffectsofproton density-weightedpulsesequencewithlowerechotime[48].Theshiftfromprotondensitytointermediateweightingisfurtheraccentuatedbyusinglongerechotrains,suchasabove10to15,asmoreT2weightingisintroduced.Tominimizeblurring,echotraincompactionshouldbeappliedthroughfastradiofrequencypulsesandapplyingthe. T2weightingT2-weightedMRimagesmayonlybereliablyobtainedwithFSE/TSEpulsesequences.ThegradientinversionofgradientechopulsesequencesforspinreversalcreatesT2contrastratherthantrueT2-weightedcontrast,whereastheradiofrequency-inducedspinreversalofspin-echo-basedpulsesequencescreatestrueT2contrast.InmusculoskeletalMRI,substantiallyT2-weightedMRimagesmaybeobtainedwithrepetitiontimesgreaterthan4000 ms,echotrainlengthof15,andechotimesgreaterthan60 ms[26].Comparedtoprotondensity-andintermediate-weightedMRimages,T2-weightedMRimagesaremoreheavilyfluid-weightedand,therefore,mostspecificfordepictingfluid-filledarticularsurfacedefects;however,theoverallcontrastresolutionofmusculoskeletaltissuesandarticularcartilageisdiminished.T2-weightedMRimagesprovideexcellentcontrastbetweencartilagesurfaceandfluidwithinthejoint,increasingthedetectionofdelaminatingareasandfocaldefects[11,49].Assuch,intermediate-weightedMRimagingmaybeconsideredbetweenprotondensityandT2-weightedMRimagingregardingthebestbalanceofcontrastresolutionandfluid(Fig. 4). T2mappingT2relaxationtimeisaquantitativeimagingapproachthattargetsthecartilagecompositionoftheextracellularmatrixcomponentsandwatercontent,allowingthedetectionandquantificationofearlybiochemicalandmicrostructuralalterations(Fig. 6)[12,14,23,28,31].Therefore,theimagingtechniquecandetectearlyalterations,suchaslossofproteoglycansanddegradationofthecartilagecollagennetwork,beforemorphologicalarticularcartilagedefectsoccur[23,28,50–52].Furthermore,thisimagingtechniqueallowsforevaluatingarticularcartilagerepairwithoutneedingGadoliniumcontrast[11,23,52,53].  Fig.6T2mappingofarticularcartilageofthepatella.AAxialspinechopulsesequenceandlevelofthepatellawithrepetition10of4000 msandfivedifferentechotimes.BScatterplotofmeasuredsignalintensities(y-axis)ofthepatellararticularcartilageagainstechotime(x-axis).CTheT2value(32 ms)iscalculatedbyfittingaT2decaycurveintothemeasuredsignalintensities  Aspartofarticularcartilagedeterioration,theinternalelasticitydecreases,allowingforincreasedamountsandmovementofwatercontentwithinarticularcartilage[50–52].ChangesinhydrationandthetypicallyanisotropicorientationofcollagenfibrilswithintheextracellularmatrixresultinT2prolongationcomparedtointactcartilage[11].ThequantitativemapscanbecreatedbyassigningrelaxationvaluesofcartilageduringT2sequenceacquisition[14,50,52].UnlikemorphologicalMRimages,whichpermitasubjectiveevaluationofcartilagealterations,quantitativeT2mappinggeneratesobjectivedatabycreatingagrey-scaleorcolormapofmeasurablerelaxationtimeswithincartilage[11,14,31](Fig. 6).ThemappedT2valuesaretypicallyprolongedindegeneratedarticularcartilageandlow-gradechondrallesions(GradeIandII)[4,11,14].Furthermore,T2valuesseemtobeinfluencedbytheorientationofthecollagen(e.g.,increasedinthelateralfacetofthepatelladuetothehorizontalcourseofthefibers)andphysicalactivity[4,11].ItmaybeprudenttonotethatT2mappingismostusefulintheabsenceofarticularcartilagedefects,whereasthepresenceofarticularcartilagedefectsmayinterferewithaccurateT2measurements[54].T2mapsmayshowmagicangleeffects(Fig. 7).  Fig.7T2mappingofarticularcartilageoftheknee.ASagittalfusionofintermediate-weightedmorphologicalMRimageandT2mapofarticularcartilagedemonstratesnormalT2valuesofcentralfemoralandtibialarticularcartilage(grayarrows)andartifactuallyprolongedT2valuesofthefarposteriorfemoralcondyle(whitearrow)duetomagicangleeffects.BAxialintermediate-weightedMRimagedemonstratingthemedialandlateralfemoralcondylesliceorientations.CMRIprotocolforT2mappingofarticularcartilage  FurtherdrawbacksofT2mappingarelongeracquisitiontimesforspin-echo/multi-echosequences,theincapacitytoevaluatecalcifiedcartilageattheosteochondraljunction,andlimitedknowledgeaboutreproducibilitybetweenvendors[23,53].Chemicalshiftartifacts,whichoccuratthecartilage-intraarticularfattissue-andbone-cartilageinterface,maycauseerrorsinT2mapping,whichmaybereducedbyswitchingthefrequencyencodingdirection[14].SyntheticMRItechniquesprovideanadditionalavenueforefficientlyacquiringcombinedmorphologicalandquantitativeevaluationofarticularcartilage,includingT1,protondensity,andT2maps[55–58].Othercompositionalimagingtechniquesareavailabletoquantifycartilagecompositionforearly-onsetosteoarthritis,suchasT1rhomapping,sodiumMRI,anddelayedgadolinium-enhancedMRIofcartilage(dGEMRIC).However,T2mappingismostwidelyavailableandusedinclinicalpractice[27,28,53,59,60]. 3DMRIIncontrasttotheanisotropicvoxelsin2Dtechniques,most3Dtechniquesallowtheacquisitionofvolumetricdatawithisotropicvoxels,permittingmultiplanarreformattingofimagedatainuser-definedplaneswithonlyasingleacquisition[11,32,35,42].Thisisparticularlybeneficialforassessinganatomicalstructures,suchasligaments,tendons,andnerves,withnonorthogonalreformattingalongorperpendiculartothesestructures[61].Isotropicdatasetsalsoallowforcurvedplanerreformation,permittingtheunfoldingofnonplanarstructures,suchasarticularsurfacesandtendons[62,63].Furthermore,slicesarethinandcontiguouswithoutsectiongaps,reducingpartialvolumeartifacts[11,33,35,42].3DMRItechniques,initiallyprimarilydevelopedforarticularcartilageresearch,areincreasinglyusedinclinicalpracticeforquantitativeMRImethods[61].Duetoconstantlyevolvingaccelerationtechniquesandautomationtechnologies,long3DMRIacquisitiontimeshavebecomemoreapplicableinclinicalroutines[32,61]. 3DgradientechotechniquesThree-dimensionalgradientecho-basedpulsesequenceshavebeenappliedinmusculoskeletalradiologyformanyyears.TheprimarycontrastscanbecategorizedintoT2-weightedbrightandT1-weighteddarkfluidsignalsequences[11,15,42,61].Theadvantagesof3Dgradientechopulsesequencesaretheinherentlyshorterrepetitiontimesandfasterechogeneration,resultinginshorteracquisitiontimesthan3DFSE/TSEspinechopulsesequences.Thosecharacteristicsof3Dgradientechopulsesequencesaretypicallytranslatedintofastacquisitions,highspatialdetail,orcombinationsthereof.Combinedwithspectralandwaterexcitationfatsuppression,3Dgradientechopulsesequencescangeneratehighcontrastresolutionspecifictocartilageandsurroundingbone,tissues,andfluid,whichisnear-idealformanualandautomatedcartilagesegmentationandvolumetricanalyses(Fig. 8).Theinherentlimitationof3DgradientechopulsesequencesisthegenerationofT2ratherthantrueT2contrast,resultinginlowerfluidspecificity.Whilearticularcartilagedefectscanbedetectedwithgradedechopulsesequences,theaccuraciesareoveralllowercomparedtointermediate-weightedandT2-weightedFSE/TSE-based3Dpulsesequences.TechniquestoimprovefluidbrightnessandmodelT2decayshaveimprovedthefluidspecificityof3Dgradientechopulsesequences;however,toachievethefulldiagnosticperformanceofMRI,pairingwithFSE/TSEpulsesequencesisrequired[64].Inclinicalpractice,3Dgradientechopulsesequencesaretypicallypartofstandard2DkneeMRIprotocolsforextendedcartilageassessmentsbuthavenotreplaced2DFSE/TSEpulsesequencesforstructuralandcontrastassessmentsofotheranatomicalstructures[11,15,41,61].3Dgradientechotechniquesaresensitivetosusceptibilityartifactpile-upsduetothemissingradiofrequencyrefocusingpulses[35,61].Lastly,thesequencesarevulnerabletomagicangleartifactsduetothecharacteristicallyshortechotimes[61].  Fig.8Fat-suppressedthree-dimensional(3D)gradientechopulsesequencesgeneratehighcontrastresolutionspecifictoarticularcartilageandsurroundingbone,tissues,andfluid,whichisnear-idealformanualandautomatedcartilagesegmentationandvolumetricanalyses.GRE,gradientecho;FISP,fastimagingwithsteady-stateprecession;VIBE,volumetricinterpolatedbreath-holdexamination;DESS,dualechosteadystate  3Dspoiledgradientrecalledecho(SPGR)sequenceThree-dimensionalspoiledgradient-recalledechopulsesequencesusesemi-randomphasevariationsoftheradiofrequencypulsetospoilthetransversesteady-statemagnetization[11,42,61].Combinedwithfatsuppression,thesignalintensityofarticularcartilageincreasescomparedtoadjacenttissuesandjointcontents,providinghighcontrastdepictionofthearticularcartilagelayers.Limitationsincludemotionartifacts,lowcontrastbetweencartilageandfluidwithappliedT2weightinghinderingthedetectionofcartilagelesions,andthevulnerabilitytowardssusceptibilityartifactsthatcanobscureabnormalitiesinbonemarrow[11,32].Furthermore,non-cartilaginousstructuresarenotreliablyassessableduetogradient-echo-basedcontrastresolutionsandmagicangleeffectswiththepotentialforhigherratesoffalse-positiveandfalse-negativefindings[11,61]. 3Ddualechosteadystate(DESS)sequence3DDESSpulsesequencesaretypicallyfluidbrightthathaveadvancedintothemostoftenused3DMRItechniquefordetectingcartilagelesions.Atleasttwopairedgradientechoesaresampledduringtheacquisitionwitharefocusingpulse.Improvementsincludetheincreasedsignalintensityofarticularcartilageandfluid,overallhighsignal-to-noiseratios,andreducedacquisitiontimeswithlessvulnerabilitytoimageartifactscomparedtoother3Dtechniques,allowingforhigherdetectionaccuracyofcartilagedefects[11,32,35,42].OptimizedDESSsequencesdepictallcartilagelayers,includingthedeepzone,resultinginthehighestarticularvolumedisplayandmaybeconsideredtheprimarypulsesequenceforvolumetricarticularcartilageanalysesinresearchsettings[28,35].While3DDESSMRimageshavehighcontrastresolutionofarticularcartilageandthesurroundingboneandjointfluid,thecontrastresolutionbetweenfluidandarticularcartilage,aswellasanatomicalstructureswithnativelylowprotondensity,suchasligaments,tendons,andmenisciaresubstantiallylower,contributingtoloweraccuraciesfordetectingsmallcartilagelesionsandtendon,ligament,andmeniscusdefects(Fig. 9)[11].Flipangleoptimizationcanmaximizefluidconspicuityandthedetectabilityofsmallcartilagedefects.  Fig.9Sagittal3DMRimagesofthekneedemonstratecontrastresolutiondifferencesbetweenfat-suppressedT2-weighted3Dturbospinecho(A)and3DDESS(B)pulsesequences.The3DDESSMRimagedemonstratesahighercontrast-to-noiseratiobetweenarticularcartilageandfat-suppressedsubchondralbone,whereasthe3Dturbospin-echoMRimagehasahighercontrast-to-noiseratiobetweenfluidandmeniscaltissue,increasingtheconspicuityofthenearverticalmeniscustear(arrows).SPACE,samplingperfectionwithapplication-orientedcontrastusingdifferentflipangleevolutions;DESS,dualechosteadystate;CAIP-IRINHA,controlledaliasinginparallelimagingresultsinhigheracceleration  Inrecentresearch,anovelaccelerationmethodbasedondeeplearningwasintroducedthatmayreducethescantimeofhighin-planeresolution3DDESSsequences.Themethodusesasuper-resolutiontechniquewithconventionalneuralnetworkstocreatedatasetswith0.7-mmslicethicknessfromdatasetswith3-mmslicethickness,maintainingimagesharpnessandhighSNR,whichisusuallyreducedinotheraccelerationtechniques[65]. Steadystatefreeprecession(SSFP)sequencesThree-dimensionalSSFPsequencesaretypicallyfluidbrightgradientechosequenceswithT1/T2-weightedcontrastresolutions[42].Theyarecharacterizedbyhighcartilagesignalandhighercontrastbetweencartilageandadjacentjointstructuresthan3Dspoiledgradient-recalledechosequences[11,32,35,42].Forthisreason,theyaremoreusefulthanother3Dgradientechopulsesequencesforcombinedassessmentsofarticularcartilage,ligaments,tendons,andarticularcartilage[11,32,35,42].However,studiessuggestedthatsusceptibilityartifactsweremoresevereandimagequalitylowerthanother3Dgradientechopulsesequences[35].Furthermore,shadingartifactscanbeamajorissuedespitebeingabalancedsteady-statesequence[66].Commerciallyavailable3DSSFPsequencesincludeFIESTA(fastimagingemployingsteady-stateacquisition;G.E.),TruFISP(truefastimagingwithsteady-stateprocession;Siemens),andbalanced-FFE(fastfieldecho;Philips)[11,32,42]. 3DFSE/TSEpulsesequences3DFSE/TSEspinechosequenceshavebecomeavailablemorerecentlyfortheassessmentofcartilage[32,42].ThesesequencesuselargeturbofactorswithvariableflipanglemodulationstoconstrainthedecayofT2overanextendedechotrain,allowingforprimarilyintermediate-weightedimageswithbrightfluidtobeacquiredwithminimizedblur-ring[11,32,35].Themajoradvantageof3DFSE/TSEover3Dgradient-recalledecho(GRE)pulsesequencesistheabilitytogeneratetrueT2contrast.Thisenablessimilarcontrastpropertiesofconventional2DFSE/TSEpulsesequences,permittingcomparativeassessmentofcartilage(Fig. 10),ligaments,menisci,nerves,andbonemarrowabnormalities[11,15,24,35,41,61].  Fig.10Isotropicintermediate-weighted3DMRIofarticularcartilage.Sagittal(A),coronal(B),andaxial(C)3DCAIPIRINHASPACEturbospin-echoMRimageswithaspatialvoxelresolutionof0.5 mm×0.5 mm×0.5 mmdemonstrateexquisitethree-dimensionalvisualizationofapartial-thicknessarticularcartilagedefectofthecentralfemoralcondylealongtheinnermargin(arrows)withfocalconvexremodelingofthesubchondralplate.SPACE,samplingperfectionwithapplication-orientedcontrastusingdifferentflipangleevolutions;CAIPIRINHA,controlledaliasinginparallelimagingresultsinhigheracceleration  Comparedtoother2Dand3Dacquisitions,3DFSE/TSEsequencesrequirethelongestacquisitiontimes[36,37].However,advancedaccelerationschemeshavemarkedlyshortenedacquisitiontimes,includingbidirectionalparallelimagingwithashiftedCAIPIRINHApatternandcompressedsensing-basedundersampling[62,63,67–69].Modern3DFSE/TSEsequencescoveringentirejointscanbeacquiredinunder5 minwithspatialresolutionsof0.5×0.5×0.5mm3forvariousnon-fatsuppressedandfat-suppressedcontrasts[67–69].3DFSE/TSEpulsesequencesarenotlimitedtobutworkbestat3 Tfieldstrength[25,26].Commerciallyavailable3DFSE/TSEpulsesequencesincludeCUBE(G.E.),SPACE(samplingperfectionwithapplication-orientedcontrastusingdifferentflipangleevolutions;Siemens),andVISTA(volumeisotropicturbospin-echoacquisition;Philips)[15,36,37]. DiscussionMRIofarticularcartilageisanintegralpartoftheworkupinpatientswithjointpain,osteoarthritis,trauma,andprevioussurgery[11,32,70,71].NumerousMRItechniquesforassessingarticularcartilagehavebeeninvestigatedandimplementedinclinicalroutine,rangingfromconventional2Dand3Dpulsesequencesformorphologicalassessmenttoquantitativetechniques,suchasT2mapping,T1rho,anddGEMRIC[72].Clinicalmusculoskeletalimagingisstillprimarilyperformedwithtri-planar2DFSE/TSEprotocols[32,72].Yet,usingnewer3Dtechniques,especiallymodern3DFSE/TSEpulsesequences,canimprovethedepictionofabnormalities,reduceexaminationtime,andincreaseclinicalefficiency[35,42,72,73].Meta-analysesfor2Dand3DFSE/TSEpulsesequencesfoundthesamediagnosticperformanceoftimeforevaluatingarticularcartilage,anteriorcruciateligamenttears,andmeniscustears.Therefore,theclinicalfeasibilityanddiagnosticperformanceof2Dand3Dsequenceshavebeenevaluatedandcomparedinnumerousstudies,whichmostlyfocusedonthekneejointandoftenincludedsurgicalcorrelation[33,74,75].Three-dimensionalgradientechopulsesequenceshavebeenusedsuccessfullyforvolumetricandstructuralarticularcartilageanalysesformanyyears.However,3Dgradientechopulsesequencesprimarilyapplytoarticularcartilageanalyses,whereastheiraccuracyfordetectingandcharacterizingabnormalitiesofotherarticularstructuresislimitedandlesssuitedtoreplaceFSE/TSEpulsesequencesincomprehensiveMRIprotocolofjoints[11,15,41,61].However,MRIprotocolsthatincludetri-planar2DFSE/TSEand3Dgradientechopulsesequencescanimprovetheoverallperformanceofarticularcartilageanalysis[32,76].Fieldstrengthmayinfluencethediagnosticperformanceof3D-SPGR,3D-DESS,and3DSSFPforcartilagelesionsdetection[32,33,77].At1.5 Tfieldstrength,3DbSSFPand3D-SPGRdemonstrated45%and85%sensitivities,respectively,withspecificityvaluesrangingfrom70to99%todetectcartilagedefects[32,42].Incomparison,3 Tfieldstrengthdemonstrated3DDESSsensitivitiesof42–97%fordetectingcartilagelesions,whilethespecificitiesrangedbetween80and99%[32,33,77–80].However,limitationsincludethelackofanindependentstandardofreference,asarthroscopymaybeunabletodifferentiategrade2andgrade3articularcartilagelesionsreliablyandwithasinglearthroscopist,theinter-observeragreementisunknown[81,82].Modern3DFSE/TSEpulsesequenceshaveshownincreasedsensitivityandspecificityfordetectingandcharacterizingarticularcartilagecomparedto3D-GREtechniques[32,42,61].Incontrast,theliteratureismoreheterogeneousconcerningdiagnosticperformancebetween2and3DFSE/TSEtechniques.Althoughindividualstudiesoftenshowednosignificantdifferencesinsensitivityorspecificityforidentifyingsurgicallycorrelatedarticularcartilagedefects,ameta-analysisindicatedthesuperiorityof3Dover2DMRI(sensitivity:77.3%vs.88.3%;specificity:91.5%vs.93.6%).Thissuperioritywasattributedto3DFSE/TSEadvantages,includingintermediate-weightedtissuecontrast,increasedsignal-to-noiseratio,reducedeffectsofpartialvolumeaveraging,andmultiplanarreformationcapabilities[32,33].Continuedimprovementsinaccelerationenabledthepossibilityofimplementing5-minhigh-qualityisotropic3DFSE/TSEsequencesasanalternativeinclinicalMRI[15,62,63].HighdiagnosticperformanceclinicalMRIprotocolsforcomprehensivearticularcartilageassessmentmayconsistofaxial,sagittal,andcoronal2DFSE/TSEsequencesandahighspatialresolution3DFSE/TSEsequenceformultiplanarandcurveplantarreformations(Fig. 10)[15,61].Thethicknessandcurvatureofthearticularcartilageandthedepthofstructuralcartilagelesionsinfluencethediagnosticperformanceof2Dand3Dpulsesequences[32].Intheknee,theperformanceoflesiondetectionishighestinthickretropatellararticularcartilageandlowestinthecurvedarticularsurfaceofthelateraltibialplateauduetopartialvolumeeffectandimagingartifacts[32,83].Articularcartilageassessmentswith2Dand3DFSE/TSEsequencesinthehipjointmaybemostchallengingwithoveralllowerdiagnosticperformancesduetothethinnessandsphericityofthearticularcartilage,deeperjointlocationwithinthebodycreatingalargerdistancetothesurfacecoil,andoftenmoreabundantsurroundingsofttissues[32,84,85].Thesensitivityforlow-gradepartial-thicknessarticularcartilagelesionsrangesbetween30and75%,whiledeeperpartial-thicknessandfull-thicknessdefectsdemonstratehigheraccuracies[32,77,86,87].Thediagnosticperformanceof3DMRIpulsesequenceshassteadilyimprovedoverthepastdecadestocomparablesensitivity(74.8%)andspecificity(93.3%)to2DMRI[32,33].Noveldeeplearningalgorithmsmayimproveimagequality,acquisitiontimes,anddiagnosticperformanceof2Dand3DMRI[39].Pioneeringresearchhasshownthefeasibilityofartificialintelligenceanddeeplearningmethodforarticularcartilagesegmentation(Fig. 11),detectionandcharacterizationofarticularcartilagelesions,andidentificationofstructuraljointalterations[88–90].  Fig.11Three-dimensionalcartilagesegmentation.Utilizinganisotropicfat-suppressedT2-weighted3DCAIPIRINHASPACEturbospinechopulsesequenceinsagittal,axial,andcoronalplanesfacilitates3Dcolor-codedrenderingofpatellar(whitearrows),femoral(grayarrows),andtibial(blackarrows)articularcartilage.CAIPIR-INHA,controlledaliasinginparallelimagingresultsinhigheracceleration  Inconclusion,various2Dand3DMRItechniquesexistforassessingandcharacterizingosteoarthritisinresearchsettingsandclinicalpractice.Differenttechniquesandpulsesequenceshaveadvantagesandlimitationsregardingimagequality,acquisitiontime,post-processingcapabilities,anddetectabilityofcartilagelesions.Accuratedetectionandcharacterizationofshallowlow-gradepartialandsmallarticularcartilagedefectsarethemostchallengingforanytechnique,butwherehighspatialresolution3DMRItechniquesperformbest.

膝内翻畸形分类与分级系统：基于冠状畸形程度、关节外或关节内位置和骨性关节炎分级(2021)VarusKneeDeformityClassificationBasedonDegreeandExtra-orIntra-Articular LocationofCoronalDeformityandOsteoarthritisGrade BagariaV,KulkarniRV,SadigaleOS,SahuD,ParviziJ,ThienpontE.VarusKneeDeformityClassificationBasedonDegreeandExtra-orIntra-Articular LocationofCoronalDeformityandOsteoarthritisGrade[J].JBJSRev,2021,9(10).. 转载文章的原链接1：https://pubmed.ncbi.nlm.nih.gov/34695035/ 转载文章的原链接2：https://journals.lww.com/jbjsreviews/abstract/2021/10000/varus_knee_deformity_classification_based_on.5.aspx AbstractBackground:Medialcoronalplanemalalignment,alsoknownasvarusalignment,iscommonlyreportedinosteoarthriticknees.Althoughthedegreeofdeformityprovidessomeinsightregardingtheseverityofthedisease,itdoesnotalwaysreflectthepotentialcomplexityofthesurgicaltreatment. Methods:Thisprospectiveobservationalstudywasconductedbyanalyzingtheradiographsof100consecutivekneesinpatientsundergoingtotalkneearthroplasty.Foreachknee,coronalalignment,expressedasthehip-knee-ankleangle,wasmeasuredonafull-legstandingradiographandclassifiedin3stages.Theprimarylocationofthevarusdeformitywasidentifiedasintra-articularand/orextra-articular.Additionally,kneeswereevaluatedtoassessfor10radiographicfeaturesofvarusdeformityandthenclassifiedin3gradesofosteoarthritisseverity. Results:Themean(andstandarddeviation)preoperativevarusdeformitywas11°±6°ofvarus(hip-knee-ankle,169°),asmeasuredonstandardizedfull-legradiographs.Extra-articularvarusdeformitywasobservedin14%ofpatients.Ahighernumberofradiographicfeaturesofvarusseveritycorrespondedwithhigherdegreesofdeformity.Varusgradecorrelatedstronglywithstageofvarusdeformity.Twenty-three(100%)of23stage-IIIdeformitieshadgrade-Cfeatures;however,13(48%)of27stage-Ipatientsalsohadgrade-Cdisease. Conclusions:Oneofevery7osteoarthritispatientswithvarusdeformityhadanextra-articulardeformity,and1of2ofthesepatientshadsevereintra-articulardisease(gradeC)despitelimitedcoronaldeformity(stageI).Thesefindingsreconfirmtheneedforindividualdeformityanalysisthataccountsforthedegree,location,andseverityofthevarusdeformity.Thisinsightmayhelptoformulateanalgorithmictreatmentapproachspecifictotheepiphysealkneeanatomyofthepatientandaccordingtothesurgicalpreferencesofthesurgeon. ClinicalRelevance:Kneesurgeonstendtoconsiderkneeswithhigherdegreesofcoronaldeformityasmoretechnicallydifficult,butthepresentstudyshowsthatkneeswithlessdeformitycanstillpresentwithseveregradesofosteoarthritisinsidetheknee,leadingtomorechallengingjointreconstruction. Medialcoronalplanemalalignment,alsoknownasvarusalignment,iscommonlyreportedinosteoarthriticknees1.Varusdeformityismorefrequentbecause,forthegeneralpopulation,nativealignmentismoreoften1°to2°ofvarusratherthanneutral2.Medialload-bearingandloadtransmissionpatternsofthenativekneeleadtomoremedialwear3.Mechanicalchangesinthejoint,afterthelossofanatomicalstructuressuchasthemedialmeniscusoranteriorcruciateligament,canalsoleadtomedialcompartmentdisease4.Duringthedevelopmentofvarusosteoarthritis,asidefromobviousjointspacenarrowing,theremightbeseveralotherradiographicchanges,suchasosteophyteformation,metaphysealbeaking(i.e.,osseousprojectioninthemedialmetaphysealarea),andcoronal-planesubluxation3,5,6.Insomecases,theremayevenbestressfracturesanddiaphysealremodelingofthetibiaandfibula,whicharethemarkersofseveredisease.Theseradiographicfeaturesareprogressiveandaretheresultsofbothmechanicalandbiologicalprocessesatthemolecularandcellularlevel7.Itiscurrentlynotknownhowtheseradiographicfeaturesofdiseaseseveritycorrelatewiththedegreeofvaruskneedeformity.Additionaluncertaintiesincludewhethermoresevereradioraphicfeaturesarealwaysfoundinmoreseverelyvaruskneesandwhethermilderradiographicfeaturesarealwayspresentinkneeswithlower-stagedvarusdeformity.Thereareseverallongstandingquestionsregardingvarusdeformityanditstreatment,suchaswhyseverevarusdeformityissometimesobservedintheabsenceofimpressivesignsofintra-articularwear,orwhyadvancedmedialosteoarthritissometimespresentswithalowdegreeofvarusalignment.Inarecentstudy,ThienpontandParviziproposedanimprovedclassificationsystemforvaruskneedeformitythatcouldhelpaddresssomeofthosequestions8.Byutilizingdataonthestage,location,andgradeofvarusdeformity,thiscurrentlyproposedclassificationsystemcouldbeutilizedtoidentifyclinicallyrelevantsignsofdiseaseandaidsurgeondecision-makingbeforeandduringkneearthroplasty.Treatingextra-articularvarusalignmentwithuseofanintra-articularosteotomy,suchastotalkneearthroplasty,mayprovemuchmorecomplicatedthanpreviouslyimagined.Theprimaryaimofthepresentstudywastoidentifytheradiographicfeaturesofvarus-deformedosteoarthritickneesandcorrelatethesefeatureswiththeseverityofthevarusdeformityandthedegreeofcoronaldeformity,andtodeterminethelocationofvarusdisease. MaterialsandMethodsFromJanuary2018untilDecember2018,theradiographsof100consecutivekneesundergoingtotalkneearthroplastywereprospectivelyassessedforvarusdeformityasaresultofprimaryosteoarthritis.Foreachpatient,weidentifiedtheseverityofcoronalplanedeformity(expressedasthehip-knee-ankleangle[HKA],withanHKAof＜178°consideredvarusdisease),theprimarylocationoftheapexofdeformity(i.e.,intra-orextra-articular),andthepresenceorabsenceof10differentradiographicfeaturesofosteoarthritis.Kneeswerethengraded(A,B,orC)accordingtotheseradiographicfeaturesandassessedastage(I,II,orIII)accordingtotheseverityofthedeformityinthecoronalplane(Figs.1through4).Kneesweresubsequentlyclassifiedaccordingtotheirrespectivegradesandstages,aswellasthelocationofthedeformity,asdescribedbyThienpontandParvizi(Fig.1).  Fig.1Illustrationshowinggrade-A,grade-B,andgrade-Cvarusdeformity.  Fig.2Radiographsshowinggrade-A,grade-B,andgrade-Cdeformity.  Fig.3Bargraphshowingthedistributionofradiographicfeaturesofvarusdeformityinthestudypopulation.  Fig.4Scatterplotshowingthedegreeofvarusandthenumberofradiographicfeaturesofvarusosteoarthritisforindividualknees.  StagesofVarusDeformityTheseverityofthevarusdeformity,reportedastheHKA,wasmeasuredonfull-lengthscanogramfilmsbydrawinglinesfromthecenterofthefemoralheadtothecenterofthekneeabovethetibialspineandthentothecenterofthetalusboneintheankle.Theanglesubtendedbetweenthe2linesonthemedialsideofthekneewasrecordedandsubtractedfrom180°.Theresultinganglewasrecorded,withHKAsof＜178°consideredvarus.Kneeswerethenclassifiedinto1of3stagesofvarusdeformity:stageI,0°to10°ofvarus(HKA,180°through170°);stageII,11°to20°ofvarus(HKA,169through160°);andstageIII,＞20°ofvarus(HKA,＜160°). LocationofVarusDeformityThelocationofthevarusdeformitywasalsoevaluated.Deformitiesassessedasextra-articularwerethosethathadanapex≥7cmbeyondthetibialjointlineand≥3cmbeyondthefemoraljointline.Theseanatomicalreferencepointswerechosenbecausethemedialcollateralligamentreportedlyinsertsdistallyat6.2cm±5.5mmfromthetibialjointline9andproximallyatthemedialepicondyle,whichis3cmfromthefemoraljointline(Fig.5).  Fig.5Illustrationdemonstratingthecalculationofextra-articularvarusinthelowerlimb.Jt=joint.  Theextra-articularapexlocationwasdeterminedbyidentifyingthecenterofrotationofangulation.Todoso,theepiphysealaxisanddiaphysealaxisofthetibiaweredrawn,withtheepiphysealaxisperpendicularinitsmiddletothetangentofthetibialplateau.Thediaphysealaxiscorrespondswiththeanatomicalaxisofthepatient,whichwasdrawnfromthemid-diaphysealpointsofthetibiaat2differentlocations.Thecenterofrotationofangulationisthepointatwhichtheproximalanddistalanatomicalsegmentsintersect,anddefinesthelocationofthedeformity10(Fig.5).Iftheapexofthedeformitywasextra-articular,thekneewasconsideredtohaveconstitutionalvarusbecauseallcasesinvolvedprimarykneeosteoarthritis.Deformitiesassessedasintraarticularwerethosearisingfromtheepiphysealanatomyandwithanoriginin1of3angles.(1)Themechanicallateraldistalfemoralangle(LDFA),whichisformedbetweenthemechanicalaxisofthefemurandthedistalaspectofthefemurinthefrontalplane.Ifthisanglewas＞90°,thefemurwasconsideredasourceofthevarusdeformity.(2)Themedialproximaltibialangle(MPTA),whichissubtendedbetweenthemedialslopeofthetibialjointlineandalineperpendiculartothemechanicalaxisofthetibia.Whenthisanglewas＜87°,theproximalaspectofthetibiawasconsideredasourceofthevarusdeformity.(3)Thejointlineconvergenceangle,whichwastheanglemadebyadrawingalinebetweenthefemoralcondylesandalinealongthetibialplateau.Aconvergenceangleof＞2°signifiedmedialcartilagelosswithtighteningofthemedialstructuresand/oropeningofthelateralsideasacontributortothevarusdeformity. RadiographicFeaturesofVarusKneeDeformity(Figs.1and2)Thepresenceorabsenceof10radiographicfeaturesofvaruskneedeformitywasnotedoneachradiograph11-13.Thesewere:1.Medialjointspacenarrowing(uptocompletelossofthemedialjointspace).2.Osteophytesattheepimetaphysealaspectofthetibia.3.Osteophytesattheepimetaphysealaspectofthefemur.4.Tibialbonedefectoralterationintheshapeoftheproximalaspectofthetibia.5.Femoralbonedefectoralterationintheshapeofthedistalaspectofthefemur.6.Metaphysealbeaking.7.Subluxationof＞3mmatthemidpointofthetibiofemoralkneejointorientationline.8.Stressfractureinthesubchondral-metaphysealregion,whichwasidentifiedonthebasisofabonedensification.9.Tibialplateauremodeling,whichwasidentifiedbyadifferentialincreaseincorticaldensityintheconcaveandconvexsidesofthetibialplateauorevidenceofahealedinsufficiencyfracture.10.Fibularremodeling,whichwasidentifiedbyadifferentialincreaseincorticaldensityintheconcaveandconvexsidesoftheproximalaspectofthefibulaorevidenceofahealedinsufficiencyfracture. Theangularmeasurementsweremadeby2differentorthopaedicsurgeons,including1consultantand1final-yeartrainee,andrepeatedafteranintervalof1week.Theinterobserverreliabilitywascalculatedwithuseofthekappastatistic.Thesemeasurementswereperformedmanuallyonthescanogramsby2independentobservers.Theimagingsystem(PACSystem;AgfaHealthcare,Belgium)hadameasurementaccuracyof0.5°.Intraobserveraccuracywastestedbyrepeatingthesameanglemeasurement10times,andwas1°forthedescribedmeasurements. GradesofVarusDeformityAccordingtowhichofthe10radiographicfeatureswereidentified,kneeswerethenclassifiedas1of3grades,gradeA(milddeformity),gradeB(moderatedeformity),andgradeC(severedeformity),witheachsubsequentgradedenotingatemporalprogressionofseverity.Thesegradeswereassessedbythefirstauthor(V.B.).Grade-Akneeswerethosethathad≥1ofthefollowing:medialjointspacenarrowing,tibialosteophytes,orfemoralosteophytes.Grade-Bkneeswerethosethathadatibialbonedefect,femoralbonedefect,and/ormetaphysealbeaking.Grade-Ckneeswerethosethathadsubluxationof＞3mm,stressfracture,tibialplateauremodeling,and/orfibularremodeling(TableI).  TABLEIDistributionofStagesandGradesofVarusDeformity  CorrelationStatisticsCorrelationbetweenthestageandgradeofthevarusdeformitywasassessedwithuseofascatterplot(Fig.4). RadiographicProtocolFull-lengthscanogramsofthelowerlimbswereassessedwithuseofInstaRadversion1.9(Meddiff).Thesestandingfull-legdigitalanteroposteriorradiographsweremadewithbothpatellaefacingforwardandwiththekneesinfullextensioninordertoeliminateerrorscausedbyrotationalmalalignment.Bothpatellaewerefacingthex-raybeam,whichwaspositionedata10°caudaltiltparalleltothejointline,asdescribedbyDixonetal.andPaleyandTetsworth14,15.Inadditiontotheseradiographs,single-legweightbearingradiographsweremadeinpatientswhohadligamentlaxityorcartilagelossin1ofthecompartments,astheeffectsofweightbearingareimportanttodocumentinsuchcases.Theseradiographsweremadebyaskingthepatienttoputmaximumweightontheoperativelegandnotmorethan10to15kgonthecontralateralleg.Theseradiographswerethenrepeated,switchingtheweightbearingoftheinjuredandcontralaterallegs. SourceofFundingTherewasnoexternalfundingforthisstudy. ResultsThemeanage(andstandarddeviation)ofthepatientswas67±8.5years.Sixty-twokneeswereinfemalepatientsand38wereinmalepatients.Themeanvarusdeformitywas11°±6°(correspondingtoanHKAof169°). StagesandLocationsofVarusDeformityOfthe100knees,27hadstage-Ivarus,50hadstage-IIvarus,and23hadstage-IIIvarus.Therewere86casesofintra-articulardeformityand14casesofextra-articulardeformity(Fig.6).  Fig.6Venndiagramshowingthelocationofdeformityforthe100kneesincludedinthestudy.Fourteenkneeshadanextra-articulardeformity,orconstitutionalvarus.  RadiographicFeaturesandGradesofVarusDeformityThemostcommonradiographicfindingsweremedialjointspacenarrowing(96%),femoralosteophytes(96%),andtibialosteophytes(94%).Otherfindingsincludedtibialremodeling(70%),subluxation(42%),andmetaphysealbeaking(13%).Therewerenocasesofstressfracture(Fig.3).Atotalof16kneeshadgrade-Adeformity,including10withstage-Ivarusand6withstage-IIvarus.Atotalof10kneeshadgrade-Bdeformity,including4withstage-Ivarusand6withstage-IIvarus.Finally,atotalof74patientshadgrade-Cdeformity,including13withstage-Ivarus,38withstage-IIvarus,and23withstage-IIIvarus. CorrelationBetweenStagesandGradesofVarusDeformityWithacorrelationcoefficientof0.668,therewasastrongrelationshipbetweenthestageandgradeofvarusdeformity(p＜0.001)(TableI;Fig.4).Allpatientsinthiscohortwereindependentlyclassifiedby2differentobservers.ThePearsoncorrelationcoefficientwascalculatedfortheresultsproducedbybothobservers.Therewasastrongpositivecorrelation,withanRvalueof0.887,andthepvaluecalculatedfromthePearsoncoefficientwas0.00001. DiscussionThepresentstudyhad3importantfindings.First,formostknees,higherdegreesofvarus(i.e.,stageIII)hadworseradiographicfeaturesofdeformity(i.e.,gradeC).Second,somekneeswithlesserdegreesofvarusalsohadworseradiographicfeaturesofdeformity,with13stage-Ikneesand38stage-IIkneeshavinggrade-Cdeformity.Third,asmallsubsetofpatientswithstage-IIIvarushadmilderradiographicfeaturesofdeformity(gradeAorB)asaresultofthedeformitybeingextra-articular.Despitedifferingfromtheconventionaltechniquesforestimatingcoronaldeformitiesandvarusdisease,thepresentgradingsystemmightbemorerepresentativeoftheunderlyingpathology.Osteoarthritisgradesshouldreflecttheneteffectonthenativeepiphysealanatomyofosseousdiseaseandsoft-tissuechanges,suchasligamentousalterationssecondarytothemechanicalandbiologicaleffectofdiseaseprogression.Increasedmechanicalstressleadstothelocalreleaseoftransforminggrowthfactorbeta,aswellasothergrowthfactorssuchasvascularendothelialgrowthfactorandfibroblastgrowthfactor,whichareresponsibleforosteophyteformation16.Usually,kneeswithahigherstageofdeformitywillalsohaveahighergradeofosteoarthritis.Thiscorrelationwasshowninthepresentstudy,withgrade-Cdeformityobservedmorefrequentlyamongkneeswithvarusof.20°(i.e.,stageIII).Otsukietal.5demonstratedthattheincreasedmechanicalloadresultingfromvarusdeformityledtotheproliferationofmesenchymalprogenitorcellsandtheiraggregationasclusters.Otherstudieshavealsodescribedthiseffectatdifferentstagesofdiseasepro-gression3,17,18.Jointspacereduction,subchondralsclerosis,andosteophyteformationaresucceededbysubluxationandtheprogressivedevelopmentoftibialandfemoraldefects19.Themostseverecasesofvarusdiseasewillshowfeaturesofmetaphysealanddiaphysealremodelingofthebonefurtherawayfromthejointline20.Inthepresentstudy,somekneeswithlowdegreesofvarus(i.e.,stageI)showedadvancedradiographicfeaturesofdeformity(i.e.,gradeC).Thesecasesofteninvolvedkneesinwhichtheoverallmorphotype(i.e.,varusorvalgus)hadbeenmodifiedbyaniatrogenicinterventionortrauma.Patientswithanativevalgusalignment(i.e.,avalgusmorphotype,withavalguship,avalgusfemurandtibia,andaplanovalgusfoot)maydevelopmedialbone-on-boneosteoarthritisfollowingacompletetearoftheanteriorcruciateligamentandmedialmeniscectomy;thisisbecausetheoverallvalgusalignmentreducestheintra-articularvaruscomponentasaresultofchangestotheepiphysealanatomy.Theresultsofthepresentstudyshowthatacursoryreviewoffulllegstandingradiographsmaynotrevealthefullseverityofvarusdeformityintheknee.Withuseofthegradingandstagingsystemoutlinedinthepresentstudy,surgeonscanbeforewarnedabouttheneedforamorecustomizedsurgicalapproach.Despitesignsofmedialbone-on-bonedisease,apatientmayhaveavalgus-typefemurwithdysplasiaofthetrochleaandthelateralposteriorcondyle.Byconsideringbothgradingandstaging,thesurgeonmightbealertedearliertoaconditionthatrequiresanalternativesurgicalapproach,suchasanextra-articulardeformityoranepiphysealanatomythatnecessitatesamorespecificreconstructiontechnique.Patientsmightalsohavemoresoft-tissuelaxityasaresultofacompensatoryunloadingmechanismormoreexternalrotation,leadingtoanunderestimationofthevarusmeasurement.Asmallsubsetofkneesinthepresentstudyhadmilderradiographicfeaturesofdeformity(i.e.,gradeAorB)despitehavingahigherdegreeofvarus(i.e.,stageIIorIII).Subgroupanalysisrevealedthatkneeswithstage-IIorIIIvarusgrade-Adeformitytypicallyhadanextra-articulardeformitylocation.Inprimarykneeosteoarthritis,anextra-articulardeformitycanbeconsideredasignofconstitutionalvarus.Correctionofonlytheintra-articulardeformitywillretaintheconstitutionalvarusforthispatientgroup.Althoughthephilosophyofthisconceptcontinuestobedebated,theresultsofthepresentstudyprovidedataforsurgeonstomakeinformeddecisionsregardingthetreatmentofextra-articulardefects,deviatingfromthelong-standingbeliefthatthesurgicaltreatmentofkneeosteoarthritisshouldresultinaneutralmechanicalalignment20.Forsuchcases,theuseof3-dimensionalimagingwillbeusefultofindtheoriginofthedeformityandtoallowthesurgeontodecidewhichcomponentpositioningandwhattypeofintra-articularcorrectionshouldbeutilizedfortheindividualpatient.Suchpreoperativeplanningwillmakeclearerwhetherthecorrectioncanbeperformedwithinthesoft-tissueenvelopeofthevariouscollateralligamentstructures.Thepresentstudyisthefirst,toourknowledge,todemonstratethatthedegreeofvarus(i.e.,stage)iscorrelatedwiththeseverityofradiographicdeformity(i.e.,grade)inkneeswithvarus osteoarthritis.Further,theinteractionbetweenvarusstageandosteoarthritisgradeseemedtodifferbetweenkneeswithintra-versusextra-articulardeformities,leadingtoamagnificationofthestage(varusmorphotype)orareductionofthestage(valgusmorphotype).Thepresentstudyhadseverallimitations.First,arelativelysmallnumberofkneeswereanalyzed,andonlyradiographicassessmentswereperformed.Furthermore,coronaldeformityisoftentheresultofdynamicinterplaybetweenthesofttissuesandtheosseousanatomy21,afactorthatisoftenmissedonstaticradiographicstudies.Computedtomographywasnotperformedtodifferentiatetheindividualrotationalalignmentsofthefemurandtibia.Furthermore,theutilityandpotentialpositiveimpactoftheproposedclassificationsystemhavenotbeentested.However,thisstudydiddemonstratetheimportanceofassessing themorphotype,epiphysealanatomy,anddiseasestagingandgradingofeachpatient,alongwithradiographicor,ifavailable,3-dimensionalimaging.Simulationsperformedwithuseofthesepreoperativefindingswillhelpthesurgeontomakebetter-informeddecisionsregardingthesurgicalprocedure.Forexample,notableangulardeformityintheabsenceofgrade-Cradiographicfeaturesshouldalertthesurgeontoapotentialextra-articulardeformityorconstitutionalvarus;asaresult,thesurgeonisabletoadjustthesurgicalplanpreoperativelyratherthanduringtheprocedure. ConclusionsCoronaldeformity,oftenreportedastheHKA,hasalwaysbeenconsideredasanexpressionofdiseaseseverity.Osteoarthritiscanbestaged,butasshowninthepresentstudy,angulardeformityshouldbeconsideredalongsidethelocation,intra-orextra-articular,ofthedeformity.Gradingtheosteoarthritiscanhelpthesurgeontobetterunderstandtheextentofintra-articulardiseaseandthepotentialcomplexityoftheepiphysealreconstructionduringkneearthroplasty.Constitutionalvarusandstage-gradedisparitymaybeconsideredsignsofmorphologicaloutliers.Advancedpreoperativeplanningandapatient-specificpreoperativeapproachandimplantselectionmaymakeadifferenceinpatientswithvarusdeformityoftheknee.

3.0-T磁共振人类骨性关节炎关节软骨的T2魔角效应的局部效应研究(2015)Investigationofregionalinfluenceofmagic-angleeffectont2inhuman articularcartilagewithosteoarthritisat3T WangL,RegatteRR.Investigationofregionalinfluenceofmagic-angleeffectont2inhuman articularcartilagewithosteoarthritisat3T[J].AcadRadiol,2015,22(1):87-92. 转载文章的原链接1：https://pubmed.ncbi.nlm.nih.gov/25481517/ 转载文章的原链接2：https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4258616/ AbstractRationaleandObjectivesTheobjectivesofthisresearchstudyweretodeterminethemagic-angleeffectondifferentsubregionsofinvivohumanfemoralcartilagethroughthequantitativeassessmentoftheeffectofstaticmagneticfieldorientation(B0)ontransverse(T2)relaxationtimeat3.0T. MaterialsandMethodsHealthyvolunteers(n=5l;meanage,36.4years)andclinicalpatients(n=5;meanage,64years)withearlyOAwerescannedat3.0-Tmagneticresonanceusingan8-channelphased-arraykneecoil(transmit-receive). ResultsTheT2mapsrevealedsignificantlygreatervaluesinventralthanindorsalregions.Whenthecartilageregionswereorientedat55°toB0(magicangle),thelongestT2valuesweredetectedincomparisonwiththeneighboringregionsoriented90°and180°(0°)toB0.Thesubregionsoriented180°(0°)toB0showedthelowestT2values. ConclusionsThedifferencesinT2valuesofdifferentsubregionssuggestthatmagic-angleeffectneedstobeconsideredwheninterpretingcartilageabnormalitiesinOApatients. Keywords:Osteoarthritis,T2mapping,cartilageimaging,magic-angleeffect,transverserelaxationtime TheprimarymacromoleculesinhumancartilagearecollagentypeIIandproteoglycans.Proteoglycanisresponsibleformuchofthecompressivestiffnessthroughelectrostaticrepulsion,whereascollagenprovidesthetensileandshearstrength(1).Theearliestbiochemicalchangesinosteoarthritis(OA)arethemodificationsatthemolecularlevelofcartilagematrix,whichoccurwithoutobviousmorphologicchanges.ThelossofglycosaminoglycanandthecollagenbreakdownarethetypicalcharacteristicsofearlyOA(2–4).Humanarticularcartilageisahighlyorderedanddepth-dependentultrastructureandisessentiallydefinedbytheorganizationofthecollagenfibrils(5).Collagenfibersinhistologyhavethreepredominantorganizationalzonesacrossthedepthofthecartilagetissue.Inthesuperficial(tangential)zone,thecollagenfabrilsareparalleltothecartilagesurface,whereasintheradial(deep)zone,thefabrilsareorientedperpendiculartothesurface.However,inthetransitional(middle/intermediate)zone,thearrangementofcollagenfibersisalmostrandom.Thecharacteristicarrangementofcollagenfibersresultsinthe“magic-angleeffect”andexhibitsanisotropicpropertieswhenmeasuredatdifferenttissuedepthsandfromdifferentphysicalorientationsintheprotonmagneticresonance(MR)images(5–7).T2mappingisapromisingapproachforassessingtheunderlyingcollagenmicrostructureintheextracellularmatrixofarticularcartilage.DamagestotheextracellularmatrixofarticularcartilageandtheincreaseofwatercontentindegeneratedcartilagegiverisetoincreasesintheT2relaxationtimes(8–10).Previousstudies(5–7,11–14)haverevealedthattheT2valueswillbeelevatedwhenthearticularcartilagesurfaceisplacedatapproximately55°withrespecttotheexternalstaticmagneticfieldB0.Thesignalchangesthatoccuratanglesapproximating54.7°areknownasthemagic-anglephenomenon魔角现象becausethedipolarinteractionbetweentwonucleiscalesis3cos2θ−1,whereθistheanglebetweentheinternuclearvectorsjoiningthenucleiandB0.Thedipolarinteractionthattendstoreducesignalintensitywillvanishwhen(3cos2θ−1)=0,whichissatisfiedwhenθequals54.7°.Onemanifestationofthismagic-anglephenomenonisthattheT2decayofcartilagetissueisgreatlyretarded,andthesignalintensityismaximalwhenthecollagenfiberisorientedatthisangleinrelationtoB0(5,11).Otherinvestigators(7,10,13,15–19)reportedthatT2relaxationreflectstheabilityoffreewaterprotonmoleculestomoveandtoexchangeenergyinsidethecartilaginousmatrix.TheT2variationmaybeduetotheregionaldifferencesincartilagecompression.Theweight-bearingportionofthefemoral–tibialjointissubjecttocompressiveforcethatmaylowerthewatercontentofthecartilage.TheaimofthisworkwastoperformaquantitativeassessmentoftheeffectofB0orientationonT2relaxationtimestodeterminethemagic-angleeffectondifferentsubregionsofinvivohumanfemoralcartilageat3.0T.WemeasuredandcomparedtheglobalandregionalchangesoffemoralcartilagesinT2relaxationtimesforhealthycontrolsandOApatientsusingquantitativeT2relaxationmethodat3.0T. MATERIALSANDMETHODSHumanSubjectsFivehealthyvolunteers(n=4menandn=1woman,ranginginagefrom24to45years,withanaverageageof35.6years)andfivepatients(n=3menandn=2women,ranginginagefrom53to82years,withanaverageageof65years)withclinicallydocumentedearlykneeOAbyradiography(Kellgren–Lawrence[K–L]gradingscale1and2)(20)wererecruited.AllhealthyvolunteersandOApatientswerescannedforT2mapping.Tolimitthepatientmotionbetweenacquisitions,thekneewasfixedwithfoampadding,whichisveryimportantbecausethesubject’spositioningisextremelycriticalforsubsequentorientationstudiesofregionalcartilageT2values.Allthehumansubjectsprovidedinformedconsenttoparticipateintheresearch,whichwasapprovedbyourinstitutionalreviewboard. ImagingHardwareAllMRIexperimentswereperformedona3.0-TclinicalMRscanner(MAGNETOMTimTrio;SiemensMedicalSolutions,Erlangen,Germany).An18-cmdiameter,8-channel,transmit–receive,phased-array(PA)kneecoilwasusedforalltheimagingmeasurements. ImagingProtocolTheprotocolincludedthefollowingsequence:2DsagittalT2-weightedspin-echo(SE)imagingwiththefollowingimagingparameters:timeofrepetition(TR)/timeofecho(TE)=4000/16.5,33,49.5,66,82.5milliseconds;fieldofview=15cm;matrix=256×256;bandwidth=130Hz;slicethickness=1.5mm. MRImagesAnalysisandProcessingAlltheMRimageswereanalyzedbasedonglobalandregionalcompartments.Threesubregions(55°,90°,and180°[0°]withrespecttoB0)weredefinedinthefemoralcartilagesofeachsubject.Thein-housedevelopedroutinesinMATLAB(version7.1;TheMathWorks,Natick,MA)andC++wereusedforofflineprocessingoftheacquiredMRimages.T2-weightedimageswiththeshortestTE(16.5milliseconds)wereusedforthesegmentationoffemoralcartilages.Regionsofinterest(ROIs)weresegmentedmanuallyforeachsliceforallthesubjects.ThesesegmentationswereusedtodrawROIsforeachMRimagewithdifferentTEvalues.T2mapswerecomputedwithcustom-builtMATLABroutinesusingthecorrespondingexpression(10,12,13,15).TheintersubjectvariabilityoftheT2mapswasquantifiedusingrootmeansquarecoefficientsofvariationpercentage(RMS-CV%),andtheStudentttestwasusedtodeterminewhethertherewereanystatisticallysignificantdifferencesintheT2valuesamongthelocal/globalregionsoffemoralcartilagesforasymptomaticandOAsubjects. RESULTSFigure1adisplaysthethreesubregionsoriented55°,90°,and180°(0°)relatedtotheexternalstaticmagneticfieldB0intheventralanddorsalregionswiththecorrespondingmagnifieddetailsshowingthehighlyorganizedcollagenstructureofthehumancartilage,respectively.Figure1bistheplotshowingthe(3cos2θ−1)factorasafunctionofanglewithrespecttoB0fornucleardipolarinteraction.Inthetwopositionswithtwoarrowsidentifyingthediscretesamplingpointswhere(3cos2θ−1)=0,theθequalsapproximately55°and125°,respectively,andthemagic-angleeffectmayemergeinthesetwosamplingpositions.Otherarrowsshowthesamplingpointswherethe(3cos2θ−1)factorhasthemaximalandminimalvalues,respectively.  Figure1(a)Theschematicdiagramofthearrangementofcollagenfibersacrossdifferentlayersofhumancartilage.Threemagnifieddetailsofthelocalregionsofhumanfemoralcartilagedisplaythecorrespondingsubregionsoriented90°,55°(magicangle),and180°withrespecttotheexternalstaticmagneticfieldB0,respectively.(b)Theplotshowingthe(3cos2θ−1)factorasafunctionofanglewithrespecttoB0fornucleardipolarinteraction.Twoarrowsidentifythediscretesamplingpointswhere(3cos2θ−1),theθequalsapproximately55°and125°,respectively,andthemagic-angleeffectmayemergeinthesetwosamplingpositions.Otherarrowsshowthesamplingpositionswherethe(3cos2θ−1)factorhasthemaximalandminimalvalues,respectively.  TworepresentativeT2(toprow)slicesobtainedfromanOApatientoverlaidontotheshortestTE(16.5milliseconds)weredisplayedinFigure2aandb,respectively.Figure2canddcorrespondinglyshowedaseriesofsubregionsonthefemoralcartilagesegmentedatevery20°withrespecttoB0.Figure2eistheT2profilesofthecorrespondingsubregionssegmentedinFigure2cord.AsshowninFigure2e,theT2valuesofOAgenerallyweregreaterthanthoseofhealthycontrols.Inthesectionsoriented55°relativetoB0(magicangle,120°–140°and240°–260°asshowninFig2cord),thelongestT2valuesweredetectedincomparisonwiththeneighboringsectionsoriented90°(100°and280°asshowninFig2cord)and180°withrespecttoB0.Thesubregionsoriented180°relativetoB0showedthelowestT2values.Furthermore,theT2valuesdisplayedobviouslygreatervaluesintheventralthaninthedorsalregions.  Figure2TworepresentativeT2(toprow)slicesobtainedfromanosteoarthritis(OA)patientweredisplayed(aandb).Aseriesofsubregionsonthefemoralcartilagesegmentedatevery20°withrespecttoB0(candd).TheT2profiles(e)ofthecorrespondingsubregionssegmentedincandd. Figure3displayedthebarchartsoftheaverageventralanddorsalT2valuesinthesubregionsoriented55°,90°,and180°relativetoB0forhealthycontrols,respectively.TheT2valuesshowedobviouslygreatervaluesintheventralthaninthedorsalregions.Thesubregionsoriented55°,90°,and180°withrespecttoB0showedthegreatest,theintermediate,andthelowestT2valuesforventralanddorsalregions,respectively(forventral:48±6milliseconds[mean±standarddeviation]vs.43±4millisecondsvs.37±3milliseconds;fordorsal:47±4millisecondsvs.41±2millisecondsvs.37±3milliseconds).ThePvaluesofsubregionsoriented55°versus90°,90°versus180°,and55°versus180°relativetoB0forventralwere.0048,.0285,and.0112,respectively,whereasthePvaluesofsubregionsoriented55°versus90°,90°versus180°,and55°versus180°relativetoB0fordorsalwere.0019,.1128,and.0153,respectively.Therootmeansquarecoefficientsofvariationpercentage(RMS-CV%)were11.96,10.25,and7.22forventralsubregionsoriented55°,90°,and180°relativetoB0amonghealthysubjects,respectively.TheRMS-CV%were8.33,5.46,and7.22fordorsalsubregionsoriented55°,90°,and180°relativetoB0acrosshealthysubjects,respectively.  Figure3ThebarchartsoftheaverageventralanddorsalT2valuesinsubregionsoriented55°,90°,and180°relativetoB0forhealthycontrols,respectively.  Figure4wasthebarchartsoftheaverageventralanddorsalT2valuesinsubregionsoriented55°,90°,and180°relativetoB0forOAsubjects,respectively.Similarly,theT2valueswereobviouslygreaterintheventralthaninthedorsalregions.Thesubregionsoriented55°,90°,and180°withrespecttoB0showedthegreatest,theintermediate,andthelowestT2valuesforventralanddorsalregions(forventral:56±5millisecondsvs.51±2millisecondsvs.47±3milliseconds;fordorsal:54±5millisecondsvs.48±3millisecondsvs.47±3milliseconds).ThePvaluesofsubregionsoriented55°versus90°,90°versus180°,and55°versus180°withrespecttoB0forventralwere.0478,.0327,and.0234,respectively.And,thePvaluesofsubregionsoriented55°versus90°,90°versus180°,and55°versus180°relativetoB0fordorsalwere.0031,.1778,and.0018,respectively.TheRMS-CV%were9.7,4.26,and5.92forventralsubregionsoriented55°,90°,and180°relativetoB0inOAsubjects,respectively.TheRMS-CV%were8.46,7.06,and5.92fordorsalsubregionsoriented55°,90°,and180°relativetoB0inOAsubjects,respectively.  Figure4ThebarchartsoftheaverageventralanddorsalT2valuesinsubregionsoriented55°,90°,and180°relativetoB0forosteoarthritissubjects,respectively.  Figure5displayedthecombinedbarchartsoftheaverageventralanddorsalT2valuesinsubregionsoriented55°,90°,and180°relativetoB0forhealthyandOAsubjects,respectively.Generally,theT2valueswereobviouslygreaterinOAsubjectsthaninhealthycontrolsexceptthattheT2valueswereobviouslygreaterintheventralthaninthedorsalregions.Thesubregionsoriented55°,90°,and180°withrespecttoB0showedthegreatest,theintermediate,andthelowestT2valuesforventralanddorsalregions.AlltheseT2variationtrendswereconsistentwiththoseshowninFigure2e.Insubregionoriented55°relativetoB0,thePvaluesofcontrolversusOAsubjectsforventralanddorsalwere.0217and.0011,respectively.Insubregionoriented90°relativetoB0,thePvaluesofcontrolversusOAsubjectsforventralanddorsalwere.0019and.003,respectively.Similarly,insubregionoriented180°relativetoB0,thePvaluesofcontrolversusOAsubjectsforventralanddorsalwere.0039and.0039,respectively.  Figure5ThecombinedbarchartsoftheaverageventralanddorsalT2valuesinsubregionsoriented55°,90°,and180°relativetoB0forhealthyandosteoarthritis(OA)subjects,respectively.  DISCUSSIONInthiswork,wecomparedtheeffectofstaticmagneticfieldorientationontheglobalandlocalT2relaxationtimeofinvivohumanfemoralcartilagetodeterminethemagic-angleeffectondifferentsubregionsofinvivohumanfemoralcartilageat3.0T.ItisgenerallyacceptedthatthefreewaterisresponsiblefortheelevationofT2valuesincartilage(8,10).OAisrelatedtothedamagetoorthelossofthecollagencomponentofthecartilagematrix,whichresultsinthedecreaseofthecartilagetensilestrength(18).T2relaxationtimemappingisoneofthemorepromisingmethodsforevaluatingtheunderlyingcollagenstructureintheextracellularmatrixofarticularcartilage.T2mappingstudiesofarticularcartilagehavedemonstratedastrongrelationshipbetweenwaterT2incartilageandtheunderlyingcollagenstructure(10,12).Itislikelythatearlydegenerativechangeinthecartilagematrixwillalterthemobility,andthus,theT2ofcartilagewater(7).AsshowninFigure2e,ourT2relaxationtimemeasurementsinbothventralanddorsalregionsofhumancartilageforOAsubjectsweregenerallygreaterthanthoseforthehealthycontrols,whichisinagreementwithpriorwork(7–10).TheangularanisotropyofT2inhumancartilageimpliesthatT2isstronglyinfluencedbythestructureofthecollagenextracellularmatrix.AsshowninFigure1a,thehighlyorganizedstructureincartilagecollageniscapableofrestrictingthemotionofwatermoleculessuchthatthespin–spincouplingbecomestosomedegreeorientationdependent.Ontheotherhand,becausehumanfemoralcartilagehasahighercurvaturethantibialorpatellacartilage,theanglebetweenitscollagenfibersandtheexternalstaticmagneticfieldorientationthroughoutthefemoralcartilagetissueattainsawiderangeofvalues(14).Ourresultsfurthervalidatedthismagic-angleeffectofmaximalT2valuesatapproximately55°withrespecttoB0orientation.AsshowninFigure1b,Figure2e,Figure3,andFigure4,theT2valuesofsubregionsorientedapproximately55°relativetoB0orientationwereobviouslygreaterthantheneighboringsubregionsorientedapproximately90°and180°relativetoB0orientationforhealthyandOAsubjectswiththeminimalincreaseofabout4.5%inT2valuesacrossallthesamplingsubregionsalongthefemoralcartilage.Thefemoral/tibialjointisaweight-bearingjointandissubjecttodifferentbiomechanicalstressesthanpatellacartilage(13).Generally,thefemoralandtibialcartilagesbearmoreweightthanpatellacartilage.ThisisanotherreasonthatwechosehumanfemoralcartilageofhealthyandOAsubjectsforT2quantitativeassessmentexceptthatthefemoralcartilagehasahigherdegreeofflexion,whichismoresensitivetothemagic-angleeffect.TheoverallT2valuesinventralfemoralcartilagearegreaterthanthoseindorsalfemoralcartilageforhealthyandOAsubjects,whichcorrespondstotheweight-bearingandnon–weight-bearingfemoralcartilageregions(Fig2e,Fig3,andFig4).Thisismostlikelyduetotheknownexudationofwaterfromcartilageundercompression.BecauseT2relaxationtimevarieslinearlywithwatercontent,cartilageunderbiomechanicalloadshouldshowlowervalues(7).Ontheotherhand,themagic-angleeffectseemstoshowthesametrendalongthemeasuredsubregionsofhumanfemoralcartilagewiththemaximalincreaseof~17%inventralregionand~14%indorsalregionofT2valuesbetweenOAandhealthysubjects(Fig2e).ThepositionsofthemaximalincreaseinT2valuesbetweenOAandhealthysubjectsinbothventralanddorsalregionsareallclosetothemagicangle(120°–140°and240°–260°).Therefore,itishypothesizedthatthemagic-angleeffectmayneedtobeconsideredintheinterpretationofT2data.AswasshowninFig2e,theaverageT2valueswere48and39millisecondsnearthesubregionsoriented55°and180°relativetoB0orientationinventralregionsforhealthycontrols,respectively.However,theaverageT2valueswere56and45millisecondsnearthesubregionsoriented55°and180°relativetoB0orientationinventralregionsforOAsubjects,respectively.ThepercentageincreaseinT2valuesforthesubregionsoriented55°and180°relativetoB0orientationinventralregionsbetweenhealthycontrolsandOAsubjectswere~17%and~15%,respectively.Ontheotherhand,thepercentageincreaseinT2valuesbetweenthesubregionsoriented55°and180°relativetoB0orientationinventralregionsforhealthycontrolswas~23%,whereasthepercentageincreaseinT2valuesbetweenthesubregionsoriented55°and180°relativetoB0orientationinventralregionsforOAsubjectswas~24%.Itseemsthatthemagic-angleeffectisrelativelymoreobviousbasedonthepercentagechangebetweenthesubregionsoriented55°and180°relativetoB0orientationinventralregionsforOAsubjects.AswereshowninFigure3,Figure4,andFigure5,thereexistssignificantdifferenceinT2valuesbetweensubregionsoriented55°relativetoB0andthoseoftheneighboringsubregionsoriented90°and180°relativetoB0inventralanddorsalfemoralcartilagesforhealthyandOAsubjects(P<.05).AlthoughthelongestT2valuesweredetectedinsectionsoriented55°toB0(magicangle)incomparisonwiththeneighboringsectionsoriented90°and180°toB0,thesubregionsoriented90°toB0showedtheintermediateT2values,andthesubregionsoriented180°toB0showedthelowestT2values,thereisnosignificantdifferenceinT2valuesbetweensubregionsoriented90°relativetoB0andthoseofsubregionsoriented180°relativetoB0inventralanddorsalfemoralcartilagesforhealthyandOAsubjects(forhealthycontrols:ventral,P=.0285;dorsal,P=.1128,asshowninFig3;forOAsubjects:ventral,P=.0327;dorsal,P=.1778,asshowninFig4).Thismaybebecauseoftherelativelylowerinsensitivityofmagic-angleeffectonsubregionsoriented90°and180°relativetoB0incomparisonwiththesectionsoriented55°toB0andthemorebiologicloadinsubregionsoriented180°thansubregionsoriented90°withrespecttoB0mayleadtothelowestT2valuesfortheformerwhichmayexudemorewaterfromthecartilagecollagenmatrixandthusgiverisetothedecreaseofT2values(2,7–10).TheT2relaxationtimeissensitivetoslowmolecularmotionsofwaterprotonsandanisotropyofthehumancartilageandhasbeenshowntodemonstratemodificationsinanatomicallyintacttissueinvivo(10).TheinherentT2relaxationtimeofcartilageisderivedfromtheinteractionsofcartilagewaterwiththecartilagemacromolecules(predominantlyfromcollagenintegrity),whicharemodulatedbytheirchemicalandstructuralstates.Previousinvitrostudies(5,6)havereportedthatT2relaxationtimevariationsshowedtheobviousorientationdependenceonthestaticmagneticfield.ThepresentpreliminarystudyisthefirsttoeverreportthataquantitativeassessmentoftheeffectofstaticmagneticfieldorientationonT2relaxationtimewasperformedtodeterminethemagic-angleeffectondifferentsubregionsofinvivohumanfemoralcartilageat3.0TforhealthyandOAsubjects.Althoughitwasunlikelythatthemagic-angleeffectforT2mapaccountedforregionaldifferencesincartilagesignalintensityinthecaseofthesmallorientationeffect(12)andtheangulardependenceontheexternalB0magneticfieldforT2valueshasmadeitdifficultinacquiringacorrectappearanceofT2maps,previousinvivo(10)andourownresultsshowedobviousincreaseinT2valuesinOApatientswhencomparedtothehealthysubjects(P<.5,Fig5);therefore,itisofgreatsignificancetomeasureandcomparetheT2mappingofdifferentsubregionsofhumancartilagebecauseofthemagic-angleeffect,whichmayprovideaconsistentinformationfortheinvivoevaluationofcartilagephysiologyandpathology.Althoughpreviousstudies(5,6,10)haveshowntheT2dependenceoforientedwaterintheorderedcollagenmatrixincartilage,thepresentworkextendstheseexperimentstobothinvivoandregionalcartilageexperiments.TheresultsofthisstudyarelikelytoimpactfutureresearchstudiesonT2relaxationincartilage.Thereareseverallimitationstothiswork.First,weevaluatedthisphenomenonexclusivelyat3T,extrapolationto1.5T,or7Tshouldbedonewithcare.Secondwastherelativelysmallnumberofbothvolunteersandpatientsstudied.FuturestudymayincludetheinvestigationoftheeffectofstaticmagneticfieldorientationonT2relaxationtimeandthedeterminationofthemagic-angleeffectoninvivohumantibialandpatellacartilagesatdifferentfieldstrengthswithlargernumbersofindividualsevaluated. CONCLUSIONSThepreliminaryresultsindicatethatmagic-angleeffectmayplayanimportantroleincartilageT2mappingandneedstobeconsideredwheninterpretingthecartilageabnormalitiesinOApatients.