Available
online
at
ScienceDirect
www.sciencedirect.com
Original
article
Computed
tomography
evaluation
of
hip
geometry
restoration
after
total
hip
resurfacing
A.
Arnould
a,b,∗,
F.
Boureau
a,b,
K.
Benad
a,b,
G.
Pasquier
a,b,
H.
Migaud
b,c,
J.
Girard
b,c,daServiceorthopédieD,centrehospitalierrégionaluniversitairedeLille,2,avenueOscar-Lambret,59037Lillecedex,France
bUniversitéLille-NorddeFrance,59000Lille,France
cServiceorthopédieC,centrehospitalierrégionaluniversitairedeLille,2,avenueOscar-Lambret,59037Lillecedex,France
dDomainemédecineetsport,facultédemédecinedeLille2,59037Lillecedex,France
a
r
t
i
c
l
e
i
n
f
o
Articlehistory: Received22September2014 Accepted16April2015 Keywords: Biomechanicalreconstruction Hipresurfacing CTreconstruction Femoraloffseta
b
s
t
r
a
c
t
Background: Anatomicreconstructionofthehipisamongthemainrequirementsforhiparthroplasty tobesuccessful.Resurfacingarthroplastymayimprovereplicationofthenativejointgeometrybuthas beenevaluatedonlyusingstandardradiographs.Wethereforeperformedacomputedtomography(CT) studytoassessrestorationofhipgeometryaftertotalhipresurfacing(HR),comparativelywiththe non-operatedside.
Hypothesis:HRdoesnotchangenativeextra-medullaryhipgeometrybymorethan5mmand/or5◦.
Patientsandmethods:CTwasusedtoevaluateunilateralHRin75patientswithameanageof52.2years (range,22–67years).Thenormalnon-operatedsideservedasthecontrolineachpatient.Meanfollow-up was2.5years(range,1.9–3.1years).Theprimaryevaluationcriteriawerefemoraloffset(FO)andfemoral neckanteversion(FNA)andthesecondarycriteriawerecupinclinationangle,cupanteversionangle,and lower-limblength.
Results:FOshowedanon-significantdecrease(mean,−2.2mm;range,−4.5to+3.7mm).FNAwas pre-served,withadifferenceoflessthan2◦atlastfollow-upversusthepreoperativevalue.Cupmeasurements
showedameananteversionangleof24.8◦(0.9–48.6)andmeaninclinationangleof44.1◦(32.1–56.3);
correspondingvaluesforthenativeacetabulumwere38.9◦(20.5–54.8)and24.8◦(4.8–33.6).Theresidual lower-limblengthdiscrepancywaslessthan1mm(mean,−0.04mm[−1.2to+1.6mm]).Themeanangle betweenthefemoralimplantandthefemoralneckaxiswas5.4◦ofvalgus.
Discussion:OurresultsshowthatHRaccuratelyrestoredthenativeextra-medullaryhipgeometry.
Levelofevidence:III,prospectivediagnosticcase-controlstudy.
©2015ElsevierMassonSAS.Allrightsreserved.
1. Introduction
Accuratebiomechanicalreconstructionofthehipisessential fortotalhiparthroplasty(THA)tobesuccessful[1],asgeometric parameterscorrelatewithjointandmusclefunction[2].Thus, fail-uretoreplicatethenativegeometrycancausealimporinstability ofthehip[3].Themanymethodssuggestedtorestorehip geome-tryincludepreoperativeplanning(usingtracingpaperordedicated computersoftware),navigation,andtheuseofmodularprostheses (e.g.,lateralizedfemoralstemsand/ormodularnecks)[4–6].In the-ory,hipresurfacing(HR)almostautomaticallyrestoresthenative hipanatomy[6]:incontrasttoTHA,HRpreservesthefemoralneck
∗ Correspondingauthorat:ServiceorthopédieD,centrehospitalierrégional
uni-versitairedeLille,2,avenueOscar-Lambret,59037Lillecedex,France. Tel.:+33320446828;fax:+33320446607.
E-mailaddress:[email protected](A.Arnould).
andthereforedoesnotinducelateralizationofthefemur, lengthen-ingofthelimb,orchangesinthecentreofrotationofthefemoral head[7].AlthoughrestorationofhipanatomyafterHRhasbeen evaluated,onlystandardradiographshavebeenusedtomeasure thegeometricparameters[6].Standardradiographslackprecision andcannotprovideinformationaboutanteversion[2].
Wethereforeconductedaprospectivestudyusingcomputed tomography(CT)toassesshipgeometryafterHRcomparativelyto thenon-operatednormalside.WehypothesisedthatHRrestored nativeextra-medullaryhipgeometrywithlessthan5mmand/or 5◦differenceversusthenon-operatedside.
2. Patientsandmethods 2.1. Patients
Aprospectivenon-randomisedstudyofpatientsmanagedbya singlesurgeonwasperformed.Inclusioncriteriawereasfollows:
http://dx.doi.org/10.1016/j.otsr.2015.04.004
Fig.1.Measurementoncomputedtomographyimagesofcupinclinationinthecoronalplane,astheanglesubtendedbythetransversecupaxisandtheinter-teardropline.
adultwhounderwentunilateralHRduringthefirsthalfof2010, normal contralateral hip (no degenerative disease or surgery), preoperative limblength discrepancy absent orless than1cm, absenceofpost-traumaticlesionsofthespineandpelvis,absence ofnegative-anglehipdysplasia, andnormal kidneyfunction. In all patients, the prosthesis used was the Conserve® Plus Total ResurfacingHipSystem(WrightMedicalTechnology,Arlington,TN, USA),withanacetabularcomponentshapedasatruncated hemi-sphereandacoverageangleof170◦.Allpatientssignedaninformed consentdocumentbeforestudyinclusion.
There were 75patients –46males (61.3%) and 29females (38.7%)–with a mean age of 52.2years (range, 22–67), a mean body weight of 80.1±17kg (95% confidence interval [95%CI], 67–92),andameanbodymassindexof26.2±4.6kg/m2(95%CI,
23–28). Among them, 32 (42.7%) had HR on the left side. The reasons for HR were primary hip osteoarthritis, n=34 (45%); osteoarthritis complicating femoro-acetabular impinge-ment,n=23(31%),osteoarthritiscomplicatinghipdysplasia,n=10 (13%);osteoarthritiscomplicatingacetabularprotrusion,n=3(4%), avascularnecrosisofthefemoralhead,n=2(3%);polyepiphyseal dysplasia,n=1,osteoarthritiscomplicatingosteochondromatosis,
n=1;andresidualabnormalitiesafterfemoralepiphysiolysis,n=1. Meanpostoperativefollow-upwas2.5years(range,1.9–3.1).
2.2. Operativetechnique
AllHRprocedureswereperformedunderlaminarflow,bya sin-glesurgeon(JG),afterpreoperativeplanningaimedatreplicating thenativegeometryofthehip(centreofrotation,femoraloffset [FO],andlower-limblength[LLL][8]).Apostero-lateralapproach wasusedandthefemurwastreatedfirst[8].Theacetabularcup waspress-fitintheanatomicposition,i.e.,paralleltothe trans-verseligamentandtotheacetabulumclearedofanyosteophytes. ThispositionwassoughtregardlessofthereasonforHR.An ante-rioroverhangof1mmwasmaintainedtoeliminateallriskofcup impingementonthepsoasmuscle.Thefemoralinstrumentation allowedchangesinfemoralcomponentpositioninallthreeplanes. Thefemoralcomponent wasimplantedin theneutral position, inthesagittalplaneof theneckand inslightvalgusrelativeto thecoronalplane.Theabsenceofnotchingof thefemoralneck waschecked.Meansizeswere57.3mm(52–66)forthecupand 51.4mm(46–60)forthefemoralcomponent.
2.3. Assessmentmethods
FOandFNAweretheprimaryevaluationcriteria.Thesecondary evaluationcriteriaweretheinclinationanglesofthecupandbony acetabulum,theanteversionanglesofthecupandbony acetabu-lum,andLLL.
Helical CT with metal artefact suppression was performed. Imageswereacquiredfromtheuppermostpointoftheiliaccrests tothelessertrochanters;slicesthroughthefemoralcondyleswere obtainedalso.CTfindingswereanalysedrelative totheanterior pelvic plane (APP, Lewinnek reference plane) [9]. Cup inclina-tioninthecoronalplanewasmeasuredbetweenthetransverse axisof the cup and theinter-teardrop line(Fig.1). The native neck-shaftangle(CC’D)wasmeasuredonthenormal contralat-eralhipbetweentheaxisof theneck(CC’)and theaxisofthe shaft(C’D).Thestem-shaftangleoftheimplant(C’C”D)was mea-sured between the axis of the shaft (C’D) and the axis of the femoralstem(C’C”)(Fig.2aandb).FOontheoperatedand non-operatedsideswasdeterminedasdescribedbyMcGroryetal.[10]. Onthenormalnon-operatedside,versionandinclinationofthe acetabulum wererecorded (Fig. 3a).Prostheticcup anteversion wasassessedastheanglebetweenthetransversecupaxisand thesagittalplane,intheaxialplane(Fig.3b).Femoralneck ver-sionwasmeasuredrelativetotheposterior bicondylarplaneof thefemur.Thehead-neckratiowascomputedasthefemoral-head diameterdividedbythefemoral-neckdiameter,measuredinthe planethroughthemiddleofthefemoralhead.LLLdiscrepancywas assessedbydeterminingthelengthoftheperpendicularsegments joiningthelinethroughthemiddlesofthelessertrochantersto theCTteardroponeachside,inthecoronalplane.Thisdistance wasmeasuredonbothsidestoassessanyLLLchangesinducedby HR.
All measurements were taken by an independent observer, who had no role in the surgical procedures and who used image-processing software (OSIRIX, OsiriXFoundation, Geneva, Switzerland) to obtain three-dimensional multi-planar recon-struction(MPR).Thisimagereconstructionsoftwarehas0.3-mm precisionandgoodreproducibilitywithaninterclasscorrelation coefficient>0.9[11].
Atlast follow-up, thefollowing clinicaldatawere recorded: Oxfordhipscore[12],Merle d’Aubigné-Postel(MAP)score[13], HarrisHipScore(HHS)[14],Devaneactivityscore[15],andUCLA activityscore[16].
Fig.2. aandbMeasurementofthenativeneck-shaftangle(CC’D,subtendedbytheneckaxisCC’andtheshaftaxisC’D)andprostheticneck-shaftangle(C’C”D,subtended bythestemaxisC’C”andtheshaftaxisC’D).
2.4. Statisticalmethods
Thestatisticalanalysiswasperformedbythemethodological supportplatformofthepublichealthdepartmentattheLille teach-ing hospital,Lille, France.Sample size wasestimated based on theprimaryevaluation criterion(FO within5mmof thenative value±2.5mm).Assumingastandarddeviationof4.7forFO,with thealpharisksetat0.05andpowerat80%,weneededtoinclude 56patients.
Descriptivestatisticswerecomputedforthequantitative meas-uresofimplant positionand ofnon-operated hipgeometry.To comparethetwosides,Student’sttestwaschosenfor paramet-ricdata (C’C”Dangle,FO, neck version relative to theposterior bicondylarplane,andoverallversion)andWilcoxon’stestfor non-parametricdata(LLLdiscrepancy, head-neckratio,neckversion relativetothecoronalplane,prostheticheaddiametervs.native headdiameter,cupversionvs.acetabularversion,andinclination ofthecupvs.inclinationoftheacetabulum).Hipgeometry parame-tersontheoperatedandnon-operatedsideswerecomparedusing Student’sttestforparametricdataandthepairedWilcoxontest fornon-parametricdata.ValuesofP<0.05wereconsidered signif-icant.
3. Results
Alltheclinicalparametersevaluatedinourstudyimproved sig-nificantly(Table1).Table2reportstheCTfindings.ThemeanFO decreasewas−2.2mm(−4.5to+3.7mm),whichwasnot statis-ticallysignificant(P=0.2).Thedifferenceinfemoralanteversion relativetothenormalsidewaslessthan2◦(P=0.3).Cupinclination
wasnotsignificantlydifferentfrominclinationoftheacetabulum onthenormalside.Incontrast,meancupanteversionwas signif-icantlygreaterthanmeananteversionoftheacetabulumonthe normalside(P=0.001):24.8◦ (15◦ to30◦ for92.7%ofcups)and 19◦,respectively.Meancupinclinationwas44.1◦(<50◦for94.9% ofcups),comparedto38.9◦forthecontralateralnativeacetabula. ThedifferenceinLLLontheoperatedsideversusthenon-operated sidewassmall(mean,−0.04mm[−1.2to+1.6mm]).Theposition ofthefemoralimplantrelativetothefemoralneckaxiswasin5.6◦ ofvalgus.Noneofthefemoralimplantswasinvarusrelativetothe nativefemoralneck.
4. Discussion
HR restores native hip geometry, as shown by our CT studydemonstratingaccuratereplicationofgeometricparameters within5mmand5◦(FO,LLL,andfemoralneckanteversion).
Nevertheless,severallimitationsofourstudymustbe acknowl-edged.WedidnotincludeconsecutivepatientsundergoingHR,as weconfinedourstudytopatientswhosecontralateralhipwas nor-mal.However,anormalcontralateralhipwasneededtoobtaindata onthenativehipgeometryofeachpatient.Thesamplewas rela-tivelysmall(n=75)butwaslargerthanthesampleneededtoobtain 80%statisticalpowerand,therefore,allowedareliableanalysisof hipreconstruction.Wechosetheanteriorpelvicplaneasthe refer-enceforanalysingourdata.Thisplanewasusedinmostpublished studiesofhipimplantposition,afactthatallowscomparisonsofour findingstothoseobtainedbyothers.Nevertheless,theorientation ofthisplanevarieswidely[17].
Thedataaremean±SD(range).
aStatisticallysignificantdifferences.
Table2
Descriptiveanalysisofhipgeometryparametersdeterminedbycomputedtomographyonthesidetreatedwithtotalhipresurfacingandonthenon-operatednormalside. Operatedside Non-operatedside Pvalue C’C”Dangle(◦) 137.3±5.7(127/151) 131.9±5.2(120/143) 0.4 LLLdiscrepancy(mm) −0.05±0.5(−1.2/1.6) −0.05±0.5(−1.2/1.6) 0.6
Femoraloffset(mm) 39.4±4.7(29/52) 41.6±5(32/54) 0.2
Head-neckratio(mm) 2±0.2(1.6/2.5) 1.8±0.1(1.5/2.1) 0.5 Versionneck/coronalplane(◦) 7.2±11.2(−17.8/30.4) 10.2±12.2(−19.4/38.7) 0.1 Diameterofprostheticheadvs.nativehead 51.4±3.7(46/60) 49.7±3.8(38/54) 0.3 Versionneck/posteriorbicondylarplane(◦) 16.7±7.1(0/41.4) 15.1±7.7(−1.1/33.2) 0.3 Versioncupvs.acetabulum(◦) 24.8±11.1(0.9/48.6) 19.1±5.8(4.8/33.6) 0.001a Overallversion(cuporacetabulumandfemur) 28.2±15.7(−7.5/62.9) 35.8±9.1(11.8/59.3) 0.07 Cupinclinationvs.acetabulum(◦) 44.1±5.5(32.1/56.3) 38.9±4.4(20.5/54.8) 0.3
Thedataaremean±SD(range).CC’D:neck-shaftangle.
aStatisticallysignificantdifferences.
Fig.3.aandbMeasurementoncomputedtomographyimagesofanteversionof thenativeacetabulum(a)andoftheresurfacingcup(b),astheanglessubtended bythetransverseaxisofthecup(oroftheacetabulumidentifiedbasedonitstwo horns)andthesagittalplane.
ArandomisedtrialcomparingTHAandHR[6]basedon radio-graphsofthereconstructedhipsshowedthatHRwasassociated withbetterrestorationofLLLandFOandwithdecreased variabil-ityofthecorrection.FOwasrestoredwithin±4mmin57%ofcases
afterHRcomparedtoonly25%afterTHA(P<0.001); correspond-ingproportionsforLLLrestorationwithin4mmwere86%and60%, respectively(P<0.001).Similarly,inaretrospectivecomparisonof THAandHR[18],restorationofFOandLLLwassignificantlybetter afterHR(meanvariationinFO,1.3mmafterHRvs.3.5mmafter THA;correspondingvalues forLL,4.9mmand 11.9mm, respec-tively).InanothercomparisonofTHA(n=50)andHR(n=40)[19]
performedduringthesameperiodbythesamesurgeon, restora-tionofFOandLLwasmorereliablewithHR.Ourresultssupport thesefindingsbyshowingthatHRaccuratelyreplicatesthenative hipgeometry.However,inallpreviousstudies,hipgeometrywas assessedusingstandardradiographs,whichareknowntoperform lesswellthanCTinthisindication,particularlyformeasuringFO andanteversion[20,21].Thus,anoriginalfeatureofourstudyis thatanatomichipreconstructionwasassessedbasedonCT,i.e., onmoreaccuratedatathanthoseobtainedbystandard radiogra-phy.
Thefemoralcomponent wasin 5.4◦ of valgusrelative tothe femoralneckinourstudy.Thispositioninslightvalgusconverts shearforces(whichcanpromoteimplantloosening)into compres-sionforces[22].Beauléetal.[23]recommended7.8◦ofvalgusto decreasetheriskoffailedfemoralcomponentfixation.Thefemoral implantshouldnotbeinvarus,asthispositionincreasesnotonly theriskofvarustiltingoftheimplant,butalsotheriskofthinning ofthecervicalcortex,whichcanresultinafemoralneckfracture
[24].
Meancupinclinationinourstudywaslessthan45◦(44.1◦).This positionmaybeoptimalinHR withametal-on-metalimplant, inordertopromote lubricationandgoodfunction[23].Marked inclinationcanresultinedgeloading,whichbreaksupthe lubri-cationfilm,therebypromotingasepticcuplooseningbyincreasing thelooseningmoment[25,26].Inaddition,cupinclinationgreater than 55◦ is nearly always associated with the risk of adverse reactionstometaldebris.Thesameappliestothesagittalplane. Retroversion of the cup can result in anterior cam impinge-ment of the cup on the anterior femoral neck, which would be expected to result in subluxation and increased metal ion shedding [27].Great care shouldtherefore be taken toremain withinthesafetyzone(inclination<45◦ andanteversion15–30◦)
5. Conclusion
Ourresultsshowthat,withtheexceptionofacetabular ante-version,HRmorecloselyreplicatesthenativehipgeometrythan doesTHA.Anteversionof thecupwasgreaterthan anteversion ofthenativebonyacetabulum, butthis positioncomplieswith currentrecommendationsforacetabularcomponentpositioning duringHR.
Disclosureofinterest
AlexandreArnould,FlorianBoureau,andKevinBenaddeclare thattheyhavenoconflictsofinterestconcerningthisarticle.
Henri Migaud declares that he has no conflicts of interest concerningthis articlebutreportsworkingasaneducationand researchconsultantforZimmerandTornier.
Gilles Pasquier declares that he hasno conflicts of interest concerningthis articlebutreportsworkingasaneducationand researchconsultantforZimmer.
JulienGirarddeclaresthathehasnoconflictsofinterest con-cerning this article but reports working as an education and researchconsultantforSmith&NephewandforMicroport. References
[1]AsayamaI,ChamnongkichS,SimpsonKJ,KinseyTL,MahoneyOM. Recon-structed hipjoint position andabductor musclestrengthafter total hip arthroplasty.JArthroplasty2005;20:414–20.
[2]LecerfG,FessyMH,PhilippotR,etal.Femoraloffset:anatomicalconcept, definition, assessment,implications for preoperative templating and hip arthroplasty.OrthopTraumatolSurgRes2009;95:210–9.
[3]CharlesMN,BourneRB,DaveyJR,GreenwaldAS,MorreyBF,RorabeckCH. Soft-tissuebalancingofthehip:theroleoffemoraloffsetrestoration.InstrCourse Lect2005;54:131–41.
[4]MatsushitaA,NakashimaY,JingushiS,YamamotoT,KuraokaA,IwamotoY. Effectsofthefemoraloffsetandtheheadsizeonthesaferangeofmotionin totalhiparthroplasty.JArthroplasty2009;24:646–51.
[5]BachourF,MarchettiE,BocquetD,VasseurL,MigaudH,GirardJ.Radiographic preoperativetemplatingofextra-offsetcementedTHAimplants:howreliable isitandhowdoesitaffectsurvival?OrthopTraumatolSurgRes2010;96:760–8. [6]GirardJ,LavigneM,VendittoliPA,RoyAG.Biomechanicalreconstructionof thehip:arandomisedstudycomparingtotalhipresurfacingandtotalhip arthroplasty.JBoneJointSurgBr2006;88:721–6.
[7]FessyMH,N’DiayeA,CarretJP,BéjuiJ,FischerLP.Locatingthecenterofrotation ofthehip.SurgRadiolAnat1999;21:247–50.
[8]GirardJ.Resurfac¸agedehanche.EMC-Techniques chirurgicales-Orthopédie-Traumatologie2013;8(2):1–9[Article44-660].
[9]LewinnekGE,LewisJL,TarrR,CompereCL,ZimmermanJR.Dislocationsafter totalhipreplacementarthroplasties.JBoneJointSurgAm1978;60:217–20. [10]McGroryBJ,MorreyBF,CahalanTD,AnKN,CabanelaME.Effectoffemoraloffset
onrangeofmotionandabductormusclestrengthaftertotalhiparthroplasty. JBoneJointSurgBr1995;77:865–96.
[11]KimG,Jung HJ,LeeHJ,LeeJS,KooS,ChangSH.Accuracyandreliability oflengthmeasurementsonthree-dimensionalcomputedtomographyusing open-sourceOsiriXsoftware.JDigitImaging2012;25:486–91.
[12]DelaunayC,EpinetteJA,DawsonJ,MurrayD,JollesBM.Cross-cultural adap-tationsoftheOxford-12HIPscoretotheFrenchspeakingpopulation.Orthop TraumatolSurgRes2009;95:89–99.
[13]MerleD’AubignéR.Numericalclassificationofthefunctionofthehip.RevChir Orthop1990;76:371–4.
[14]HarrisWH.Traumaticarthritisofthehipafterdislocationandacetabular frac-tures:treatmentbymoldarthroplasty.Anend-resultstudyusinganewmethod ofresultevaluation.JBoneJointSurgAm1969;51:737–55.
[15]DevanePA,HorneJG, MartinK,ColdhamG,KrauseB. Three-dimensional polyethylenewearofapress-fittitaniumprosthesis.Factorsinfluencing gen-erationofpolyethylenedebris.JArthroplasty1997;12:256–66.
[16]ZahiriCA,SchmalzriedTP,SzuszczewiczES,AmstutzHC.Assessingactivityin jointreplacementpatients.JArthroplasty1998;13:890–5.
[17]PinoitY,MayO,GirardJ,LaffargueP,AlaEddineT,MigaudH.Lowaccuracy oftheanteriorpelvicplanetoguidethepositionofthecupwithimageless computerassistance:variationofpositionin106patients.RevChirOrthop 2007;93:455–60.
[18]AhmadR,GillespieG,AnnamalaiS,etal.Leglengthandoffsetfollowinghip resurfacingandhipreplacement.HipInt2009;19:136–40.
[19]SilvaM,LeeKH,HeiselC,DelaRosaMA,SchmalzriedTP.Thebiomechanical resultsoftotalhipresurfacingarthroplasty.JBoneJointSurgAm2004;86: 40–6.
[20]SarialiE,MouttetA,PasquierG,DuranteE,CatoneY.Accuracyofreconstruction ofthehipusingcomputerisedthree-dimensionalpreoperativeplanninganda cementlessmodularneck.JBoneJointSurgBr2009;91:333–40.
[21]PasquierG,DucharneG,AliES,GiraudF,MouttetA,DuranteE.Totalhip arthroplastyoffsetmeasurement:isCTscanthemostaccurateoption?Orthop TraumatolSurgRes2010;96:367–75.
[22]WatanabeY,ShibaN,MatsuoS,HiguchiF,TagawaY,InoueA.Biomechanical studyoftheresurfacinghiparthroplasty:finiteelementanalysisofthefemoral component.JArthroplasty2000;15:505–11.
[23]BeaulePE,DoreyFJ,LeDuffM,GruenT,AmstutzHC.Riskfactorsaffecting out-comeofmetal-on-metalsurfacearthroplastyofthehip.ClinOrthopRelatRes 2004;418:87–93.
[24]GirardJ.Isittimeforcementlesshipresurfacing?HSSJ2012;8:245–50. [25]KomistekRD,DennisDA,OchoaJA,HaasBD,HammillC.Invivo
compari-sonofhipseparationaftermetal-on-metalormetal-on-polyethylenetotalhip arthroplasty.JBoneJointSurgAm2002;84:1836–41.
[26]JarrettCA,RanawatAS,BruzzoneM,BlumYC,RodriguezJA,RanawatCS.The squeakinghip:aphenomenonofceramic-on-ceramictotalhiparthroplasty.J BoneJointSurgAm2009;91:1344–9.
[27]BeaulePE,HarveyN,ZaragozaE,LeDuffMJ,DoreyFJ.Thefemoralhead/neck offsetandhipresurfacing.JBoneJointSurgBr2007;89:9–15.
[28]VailTP,Mina CA,YerglerJD,PietrobonR.Metal-on-metalhipresurfacing comparesfavorablywithTHAat2yearsfollow-up.ClinOrthopRelat Res 2006;453:123–31.