Ebook Pediatric and adolescent knee surgery: Part 2

SECTION

3

OsteochondritisDissecans




CHAPTER
EricW.Edmonds
HenryG.Chambers



21

OsteochondritisDissecans:Overview,
Epidemiology,Etiology,
Classification,Assessment
INTRODUCTION
AlthoughloosebodieswithinajointwerefirstdescribedbyPaget,1Königlater
suggestedthreemethodsbywhichloosebodiescouldbecreated:(1)direct
traumawithacutefracture,(2)minimaltraumathatdevelopsintoosteonecrosis
andsubsequentfragmentation,or(3)notraumawithspontaneousfragmentation.
Thelattervarietyhecoinedosteochondritisdissecans(OCD).2,3Althoughit
shouldbepointedoutthattheexactpathophysiologyremainsunknown,itis
agreedthatOCDislikelyanacquiredlesionofsubchondralbone.
Beyondthischaracterization,itislessclear.Therearedegreesofosseous
resorption,collapse,andsequestrumformationwithpossibleinvolvementofthe
articularcartilagethroughdelaminationunrelatedtoanacuteosteochondral

fractureofnormalcartilage(Fig.21.1).4,5Thisunderstandingoftheendpointof
thediseaseprocesshasledtomanyetiologiesofOCDbeingpostulated
(particularlyconcerningtheknee)includingtrauma,6,7inflammation,2,8
genetics,9vascularabnormalities,10,11andconstitutionalfactors.12However,the
etiologyremainsunknown,eventhoughourveterinarymedicinecolleagueshave
madesomeleapsinunderstandingoverrecentyears.13,14


Historically,therehasbeenadistinctionbetweenjuvenile-onsetOCDand
adult-onsetOCD.Manysurgeonshavesuggestedthatskeletallyimmature
patients(juvenileonset)haveabetterprognosisthathasbeeninconsistently
definedintheliteratureaseitherradiographichealingormerelyresolutionof
pain.4,5,9,11,12,15Despitethelackofanopenphysisatthetimeofdiagnosisinan
adultOCD,however,manyauthorssuggestthattheonlytruedifferencebetween
juvenile-andadult-onsetOCDispurelyareflectionofpatientageatthetimeof
diagnosis.
ArecentdefinitionofhumanOCDlesions,proposedbytheResearchin
OsteochondritisoftheKnee(ROCK)studygroup,highlightsthefactthatthese
are(1)focal,(2)idiopathic,(3)involvesubchondralbone,and(4)riskinstability
anddisruptionofarticularcartilagewithpotentiallong-termconsequences,such


asprematureosteoarthritis.16Thisdefinitionisthesummaryofepidemiology,
etiology,classification,andassessmentofOCD.

EPIDEMIOLOGY
ThereareonlythreetrueepidemiologypapersregardingkneeOCD.10,17,18The
firstwasperformedbyMarsdenandWiernik18inareviewof18,405radiographs
atamilitaryhospital.TheyfoundanincidenceofsymptomaticOCDof2.3%of
theradiographsandanoverallincidence(includingincidentaldiscovery)of4%

OCDintheircohort.
AclassicstudybyLinden10in1977fromMalmo,Swedendemonstratedan
incidenceof29per100,000boysand19per100,000girls.Moreimportantly,
afterhereviewedradiographsandobtainedfollow-upwithmanyofthese
patients33yearslater,hediscoveredthattherewasan“accumulatedriskof
roentgenographicgonarthrosisinpatientswhohavehadosteochondritis
dissecans.”Ifheassumedtheriskofarthritistobe0%atthestartoflife,this
wasunchangedatage40yearswithanOCD;buttheriskincreasedto70%at
age48yearsandcontinuedtoincreaseto95%at70yearsofage.However,in
thoseinitiallydiscoveredwithjuvenile-onsetOCD,hecouldnotdirectly
correlateanypathologysuchasgonarthritiswiththeirOCDdirectly.
Theonlyotherstudywaspublishedin2014andincludedareviewofjustover
1millionchildrenaged2to19yearswithinaclosedhealthsystem.17These
authorsfound192childrenwith206OCDlesionsoftheknee.Themajority
(64%)ofthelesionsinvolvedthemedialfemoralcondyle,andtheoverall
incidenceseeninthecohortwas18.1per100,000boysand3.9per100,000
girls.TheincidenceofkneeOCDvariedbyethnicity:non-Hispanicwhitewas
10.3per100,000overall(17.3and3.0per100,000forboysandgirls,
respectively),non-Hispanicblackwas10.3per100,000overall(17.3and3.0per
100,000forboysandgirls,respectively),Hispanicwas8.6per100,000overall
(14.3and2.8per100,000forboysandgirls,respectively),andAsianwas4.7
per100,000overall(9.1and0.0per100,000forboysandgirls,respectively).
Theseauthorsperformedmultivariablelogisticregressionanalysisthatrevealed
a3.3-foldincreasedriskofOCDofthekneeinchildrenaged12to19years
comparedwiththoseaged6to11years.Moreover,boyshad3.8timesgreater
riskofOCDthangirls.


Finally,thereisthediscretepossibilitythatkneeOCDmayoccurinboth
knees.Intheliterature,thereisarelativelywiderangeofbilateralitynotedwith

abouta3%to30%chanceofdiscoveringitinbothkneesbyx-ray.5,18–23

ETIOLOGY
NodefinitiveetiologyhasyetbeendeterminedfortheoriginofkneeOCD.
Thereare,ofcourse,manyhypothesesthathavebeenpresentedandtested
primarilyviaexvivohistology.Thepotentialetiologiesincludeinflammation,
spontaneousosteonecrosisandvasculardeficiency,geneticpredisposition,and
repetitivetrauma.Eachofthesewillbediscussed.
Assuggestedbythename“osteochondritis,”thefirstdescriptionin1888by
König,2bydefinitionwasatraumaticanditwaspostulatedtobearesultof
inflammation.However,histologicstudieshavenotsupportedthisetiology7,24;
instead,theseworksappeartohighlightfindingsofnecrosiswithintheOCD
lesionsratherthaninflammation.
Basedontheirhistologyfindings,GreenandBanks12,19proposedthat
ischemiawastheprimaryetiology,andMilgram,22identifyingrevascularization
inpartiallydetachedOCDlesion,furtherpromotedthisconceptofpoor
vascularity.Yet,Yonetaniandcolleagues25foundnoevidenceofnecrosison
biopsies.Uozumiandcolleagues26discoveredadiscreteabsenceofsubchondral
boneinmanyoftheirbiopsysamplesand,inthosewhohadanosseous
componentpresent,onlytwodemonstratednoviableosteocytes.Thedifference
betweenthefirsttwostudiesandthesecondtwoisstateoftheOCDlesionprior
tobiopsy.Whenthelesionwasunstable(orevenaloosebody)atthetimeof
biopsy,thennecrosiswasfoundbymicroscopicevaluation.However,ifthe
lesionwasfullyintactandinsitu,thentherewaslessdefinitiveevidencefor
avascularnecrosis.
AnotherhypothesisoftheetiologyisfamilialinheritanceofOCDlesions.A
mildformofskeletaldysplasiawithassociatedshortstaturehasevenbeen
proposed.9,23,27–29Twodifferentauthors,viadifferentfamilytreeassessments,
haveidentifiedwhattheybelievetobeanautosomaldominantinheritance
pattern.9,28Incontrast,Petrie30reportedonaradiographicexaminationoffirstdegreerelativesofthosepatientswithknownOCDanddiscoveredonly1.2%

withOCDthemselves.ThissuggeststhatthemajorityofkneeOCDdoesnot
followapredictableinheritancepattern.However,itdoesnotexcludethe
possibilitythatgeneticsmaystillplayaroleinOCDetiology.
Whethertheaforementionedetiologiesareprimaryorsecondaryremainsa
questionuntoitself,andtheconceptofrepetitivemicrotraumaisanotherdistinct
possibility.Forexample,despitetheoriginaldescriptionbyKönig,therecould
beasignificanttraumaticeventleadingtoavascularnecrosisortherecouldbe
repetitivetraumaticeventsthatprogressivelydevelopvasculardisruption.In
1933,Fairbanks31proposedtraumaticcontactbetweenthelateralaspectofthe
medialfemoralcondyleandthetibialspine;yet,thistheorywouldonlyexplain
OCDlesionsonthelateralaspectofthemedialfemoralcondyle.However,the
ideaofrepetitivetraumastillholdsourattentionfortheotherlocationsofOCD
aswell.Aichroth6demonstratedthat60%ofthepatientswithkneeOCDinhis
cohortwereinvolvedinhigh-level,competitivesports.Moreover,Linden11
showedanassociationbetweentheincidenceofOCDandtheincreased
involvementinorganizedsportsinSwedenbetween1965and1974.Alarge
multicenterstudyconductedbytheEuropeanPediatricOrthopaedicSociety
demonstratedthatnearly55%ofthosewithOCDwereregularlyactiveinsports
orperformed“strenuousathleticactivity.”20Evidencetosupportrepetitive
traumaasanetiologyforOCDismostlyconjectureandcircumstantialatbest.
Perhapsabnormalpressureonthecartilageanlageisabetterwayof
describing“microtrauma,”asseveralstudieshavenotedassociationsbetween
lateralfemoralcondyleOCDlesionsanddiscoidmenisci.32–34Moreover,there
isanapparentassociationwithpoormechanicalaxisalignmentandthepresence
ofkneeOCD.35Infact,theassociationbetweenalignmentshiftsandlocationof
OCDdevelopmentwaspredictive,withvarusmalalignmentpredictingmedial
OCDlesionsandvalgusmalalignmentpredictinglateralOCDlesions.35These
findingssuggestthataberrantmechanicalpressureonthecondylesmaybean

etiologytotheformationofOCD(atleastintheknee).
Perhapstheunifyingtheorycanbefoundintheworkbyourveterinary
colleaguesonhorseandpigcomparativemodels.TheybelievethattheOCD
lesionsintheseanimalsoriginatewithinthecartilageanlageofthesubchondral
bone.13,14TheyhaveactuallybeenabletocreateananimalmodelforOCDby
descriptionofthecartilagecanals(whichisthebloodsupplytotheepiphyseal


anlagecartilage).Thisfurtherhintstowardapossibletraumaticinjurythat
disruptsthebloodsupply.Theyalsosuggestmodifyingthetermosteochondritis
toosteochondrosisgiventhecompletelackofinflammationbeingseenasa
possibleetiology.Furthermore,theywouldincludethemodifier“dissecans”
whenthearticularcartilagebecomesinvolvedbutrecommendusingtheterms
latenswhenthereisonlynecrosisoftheanlagecartilageormanifestainthe
presenceoffocalfailureofenchondralossificationtodifferentiatethetimingof
discovery.Inotherwords,theseauthorsbelievethatbythetimethispathologic
processbecomesclinicallyevidentintheknee,ithasalreadyevolvedthroughall
threestagesandisinthefinalnecroticformofOCD(Fig.21.2).

Thisconceptisnotnewtohumanphysicians,buttheconceptofdisrupted
cartilagecanalsisnoveltotheprevioushypothesisofRibbing36in1937.His
thesiswaspresentedina107-pagesupplementtothejournalActaRadiologica
discussingabnormalitiesofendochondralossificationoftheepiphysis.His
hypothesiswasthenfurthermodifiedbyBarrie7,24tofurtherdefinepossible
etiologiesofOCDformationviaepiphysealsecondarygrowth.Basically,atan
indexevent,thereisaninsult(singleorrepetitive)totheendochondral
epiphysealgrowthplateorperhapstothecartilagecanalsjustdeepto
subchondralbone.Withskeletaldevelopment,theuninjuredregionof
endochondralepiphysealossificationcontinuestoossifyunhindered,creatingan

everenlargingOCDcentraltothenormalgrowth.
Althoughtheconceptsofdisruptedcartilagecanalsorendochondral
epiphysealgrowthplatesareappealingetiologies,thereremainsagapinthe
explanationregardinghowandwhythesestructuresbecomeinjuredinthefirst
place.Allchildrenarequiteactive,jumping,playing,running,andfalling.So
whydosomedevelopanOCDandsomedonotdevelopanOCD?Thisquestion
isyettobeanswered.

CLASSIFICATION
ThesimplestformofkneeOCDclassificationisthatsetforthbyHeftiand
colleagues20thatmerelydescribeslocationoftheOCD.Themostcommonsite
ofOCDisthelateralaspectofthemedialfemoralcondyle(51%involvement),
withothersitesbeinginvolvedwithlessfrequency:19%centralmedialfemoral
condyle,17%lateralfemoralcondyle,7%medialsideofthemedialfemoral
condyle,and7%patella.Itmaybeimportanttonotethatirregularossification
centersofthedistalfemoralcondyletendtobefoundmoreposteriorlyonthe
condyle(althoughtheycanbeanywhereonthecondyle)andareassociatedwith
youngerage.Cahill12wroteareviewofOCDandidentifiedmultipleother
authorswhohavesimilarlyclassifiedkneeOCDbylocation,buthemakesno
mentionofclassifyingthelesionsinanyothermethod(stability,outcomes,etc.).
DeSmetandcolleagues37tookthingstothenextlevelandclassifiedtheOCD
eitherstableorunstableandcorrelatedtheirmagneticresonanceimaging(MRI)
findingswitharthroscopicfindings.Theyreportedthatahighsignalinterface
betweentheOCDlesionandthenormalfemoralbonewasindicativeof
instabilitybyarthroscopy.Thissameauthorthenlaterimprovedhisassessment
byMRIandaddedthreeothersignsofinstabilitythatarediscussedinthe
followingtext(Fig.21.3).38

ThearthroscopicclassificationofOCDlesionsisthegoldstandardin
understandingthelesion,yetevenhere,wedonothaveassociatedoutcomes.
Guhl39wasthefirsttosetanarthroscopyclassificationandithasbeenusedasa
templateforotherauthors:(1)intactlesions,(2)lesionsshowingsignsofearly
separation,(3)partiallydetachedlesions,and(4)craterwithloosebodies
(salvageableorunsalvageable)(Fig.21.4).Tothispoint,nocorrelationwith
outcomeshasbeenmade.

ASSESSMENT
Mostauthors20,40suggestthattreatmentforkneeOCDshouldbebasedon


skeletalmaturityandlesionstability.Anumberofclassificationsystems
mentionedearlierincludesthesetwoprinciples.12,37–39Therefore,itisimportant
toidentifyOCD,identifyskeletalmaturity,andidentifyOCDstability.
ChildrenwithOCDmayhavetwodifferentpresentations,eithertheOCDis
anincidentalradiographicfindingsoritmaybethoughttobethecauseoftheir
symptoms.41,42CahillandAhten43reportedthat80%ofthechildrenhadmild
painandlimpforameanof14monthspriortopresentation—suggestingthatthe
symptomsassociatedwithkneeOCDmaynotdiffergreatlyfromidiopathic
adolescentkneepain.However,theremaybesubstantialsymptomsassociated
withtheOCDlesion,suchasmechanicalsymptomsoflockingandpopping.
Effusionispresentinlessthan20%atpresentation.20Wilson53describedatest
onphysicalexaminationtodiagnoseOCD:Thekneeisflexedto90degrees,the
tibiaisexternallyrotated,andthekneeisgraduallyextendedto30degreesof
flexion.Apositivetestischaracterizedbypainovertheanteromedialaspectof
thekneeasthekneeisextendedto30degreeswithreliefofpainwithinternal
rotationofthetibia.Anatomically,thismaneuverisbelievedtocause
impingementofthetibialspineonthelateralaspectofthemedialfemoral
condyle(andthereforewouldonlybepositiveonthoseparticularOCDlesions).

TheWilsontestisoflimiteddiagnosticvaluewithareportedpositivetestin
only16%ofkneeswithradiographicallyprovenOCDlesions.20
ThediagnosisofOCDistrulydependentonimagingratherthanhistoryor
physicalexamination.EspeciallyconsideringthatmanyOCDlesionsmaybe
asymptomaticuntilseparationordislodgmentoccursresultinginsignificant
effusionandpain—thisisconsideredalatefindinginthetreatmentofthese
lesions.Mostlesionscanbeidentifiedbyplainradiographsandwillusuallynot
bemissedaslongasfourviewsareobtained,includinganteroposterior(AP),
lateral,tunnel(notch),andMerchant(sunrise)views.Todate,thereisno
historicalliteratureassessingthesensitivityorspecificityofradiographsto
identifyOCDlesions,andtheAmericanAcademyofOrthopaedicSurgeons
(AAOS)clinicalpathwayguidelinetaskforcerecommendedthattherewasonly
limitedevidencetosupportobtaininginitialplainradiographsintheassessment
ofkneeOCD(andinconclusiveevidencetoobtaincontralateralfilmsatinitial
evaluation).40IntheassessmentofchildrenwithanOCDlesion,itappearstobe
clinicallyacceptabletoobtaincontralateralfilmstoevaluateforbilateral


involvement,giventhehighrateofbilateralinvolvementseeninpast
studies.45,46
Often,MRIisthenextstepintheassessmentoftheOCD,lookingfor
instabilityofthelesionaswellasbetterdefiningthesizeandextentofthe
pathologicchanges.MRIcanalsobehelpfulindistinguishingirregularcenters
ofossificationfromOCD,andsomeauthorshavesuggestedthattheimproved
outcomesseeninjuvenileOCDmayberelatedtoanincorrectdiagnosisof
OCD.However,onceagain,theAAOSclinicalguidelinestaskforcefound
limitedevidenceforobtaininginitialMRIforevaluationandinconclusive
evidenceforusingMRIinsubsequentfollow-upevaluation.40MRI
characteristicsconsistentwithirregularossificationincludeslowsignalintensity
ofthelesionbonebeingsimilartothesurroundingsubchondralboneandhigh

signaldemarcationbeingequaltothesurroundingarticularcartilage.48The
characteristicsofOCDarevaried,butthereareafewfindingstonote:Thereis
oftensurroundingbonemarrowedema,thereisascleroticrimofbone
demarcatingthelesion±highsignalintensitylayerhighlightingthedemarcation,
theremayormaynotbebonewithintheprogenyfragment,theremaybecystlikeformationsalongthedemarcation,andtherecouldbearticularcartilage
disruption.
Radiographicassessmentshouldbeobtainedatintervalsdependingonthe
treatmentbeingrenderedbutshouldbecontinueduntilthereisaclear
demonstrationoflesionresolutionbecausesymptomsarenottrustworthyguides
inthestabilityofthelesion.Ifthelesionexhibitsanyradiographicsignsof
worsening(enlargementorincreasedscleroticborder),orthelesionfailsto
resolveasthechildgrows,thennotonlyisadiagnosisofOCDappropriateover
irregularossificationbutalsoachangeininterventionmaybewarranted.The
definitionofhealingbyplainfilmorMRIisnotwelldefined.Thevarious
definitionsusuallytrytotakeintoaccounttheconversionofcartilaginouslesion
toosseousdensity,orsomepercentageofthatprocess,whereasothersdefine
healingbyresolution,orresorption,ofthescleroticmargin.49Unfortunately,
someauthorsstilldefinehealingastheresolutionofsymptomsregardlessofany
radiographicchanges.44,47Noone,asyet,correlatedradiographicfindingswith
healingbecausehealinghasnotyetbeendefined.
ThereissomeagreementregardingthedefinitionofanunstableOCDlesion:


eitherfracturedcartilageorseparationoftheunderlyingsubchondralbone.50
Thisdistinction,stableversusunstable,isimportantbecauseitcanoftendictate
thetreatmentandprognosis.Ingeneral,stablelesionsarebelievedtohavebetter
outcomesandstableOCDlesionsmaybeconsideredfornonoperative
management,whereasunstablelesionsarebettertreatedwithsurgical
intervention.20,40Thereare37,38fourcriteria(onT2weightedmagnetic
resonance[MR]images)thatcorrelatewithinstabilitywhencomparedwith

arthroscopyfindings:(1)ahighsignallinebeneaththelesion,(2)afocaldefect
inthearticularcartilage,(3)afractureofthearticularcartilage,and(4)the
presenceofsubchondralcysts.ThepresenceofahighT2signallinebetweenthe
lesionandtheremainderofthenormalbonewasthemostpredictivefactorfor
instabilityinamixedagecohortofpatients,butYoshidaetal.34evaluateda
skeletallyimmaturecohortwiththissamehighT2signallineseparatingthe
lesionandhefoundnocorrelation.However,hereportedthatthislinecouldnot
predictstabilitybecausethemajorityofthejuvenileOCDlesionshealed.
Anotherstudy51suggestedthatifthecriteriaforinstabilityincludedboththe
highT2signalseparatingthelesionandabreachofthearticularcartilageonT1
weightedimages,thentheaccuracyinpredictinginstabilityshiftedfrom45%to
85%.Kijowskiandcolleagues52examinedtheaforementionedcriteriaofMRI
stabilityandappliedittopredictingarthroscopicinstability(definedascartilage
disruptionormotionoftheoverlayingcartilageattheOCD)andfoundthatall
fourcriteria(together)had100%sensitivityand100%specificityforinstability
inadultOCDand100%sensitivitybutonly11%specificityforjuvenileOCD
lesions.

REFERENCES
1.PagetJ.Ontheproductionofsomeoftheloosebodiesinjoints.St
Bartholomew’sHospRep.1870;6:1–4.
2.KönigF.Ueberfreiekorperindengelenken.DtschZChir.
1887;27:90–109.
3.FreifbergAH,WoolleyPG.Osteochondritisdissecans:concerningits
natureandrelationtoformationofjointmice.JBoneJointSurgAm.
1910;8:477–494.
4.CrawfordDC,SafranMR.Osteochondritisdissecansoftheknee.JAm


AcadOrthopSurg.2006;14:90–100.

5.KocherMS,TuckerR,GanleyTJ,etal.Managementof
osteochondritisdissecansoftheknee:currentconceptsreview.AmJSports
Med.2006;34:1181–1191.
6.AichrothP.Osteochondritisdissecansoftheknee.Aclinicalsurvey.J
BoneJointSurgBr.1971;53:440–447.
7.BarrieHJ.Hypothesis—adiagramoftheformandoriginofloose
bodiesinosteochondritisdissecans.JRheumatol.1984;11:512–513.
8.SchenckRCJr,GoodnightJM.Osteochondritisdissecans.JBone
JointSurgAm.1996;78:439–456.
9.MubarakSJ,CarrollNC.Familialosteochondritisdissecansofthe
knee.ClinOrthopRelatRes.1979;(140):131–136.
10.LindenB.Osteochondritisdissecansofthefemoralcondyles:along-term
follow-upstudy.JBoneJointSurgAm.1977;59:769–776.
11.LindenB.Theincidenceofosteochondritisdissecansinthecondylesofthe
femur.ActaOrthopScand.1976;47:664–667.
12.CahillBR.Osteochondritisdissecansoftheknee:treatmentofjuvenileand
adultforms.JAmAcadOrthopSurg.1995;3:237–247.
13.OlstadK,HendricksonEH,CarlsonCS,etal.Transectionofvesselsin
epiphysealcartilagecanalsleadstoosteochondrosisandosteochondrosis
dissecansinthefemoro-patellarjointoffoals;apotentialmodelofjuvenile
osteochondritisdissecans.OsteoarthritisCartilage.2013;21(5):730–738.
14.YtrehusB,CarlsonCS,EkmanS.Etiologyandpathogenesisof
osteochondrosis.VetPathol.2007;44(4):429–448.
15.BradleyJ,DandyDJ.Osteochondritisdissecansandotherlesionsofthe
femoralcondyles.JBoneJointSurgBr.1989;71:518–522.
16.EdmondsEW,SheaKG.Osteochondritisdissecans:editorialcomment.Clin
OrthopRelatRes.2013;471(4):1105–1106.
17.KesslerJI,NikizadH,SheaKG,etal.Thedemographicsandepidemiology
ofosteochondritisdissecansofthekneeinchildrenandadolescents.AmJ
SportsMed.2014;42(2):320–326.

18.MarsdenC,WiernikG.Theincidenceofosteochondritisdissecans.JRArmy
MedCorps.1956;102(2):124–130.
19.GreenWT,BanksHH.Osteochondritisdissecansinchildren.JBoneJoint


SurgAm.1953;35:26–47;passim.
20.HeftiF,BeguiristainJ,KrauspeR,etal.Osteochondritisdissecans:a
multicenterstudyoftheEuropeanPediatricOrthopedicSociety.JPediatr
OrthopB.1999;8:231–245.
21.HughstonJC,HergenroederPT,CourtenayBG.Osteochondritisdissecansof
thefemoralcondyles.JBoneJointSurgAm.1984;66:1340–1348.
22.MilgramJW.Radiologicalandpathologicalmanifestationsofosteochondritis
dissecansofthedistalfemur.Astudyof50cases.Radiology.
1978;126:305–311.
23.MubarakSJ,CarrollNC.Juvenileosteochondritisdissecansoftheknee:
etiology.ClinOrthopRelatRes.1981;(157):200–211.
24.BarrieHJ.Hypertrophyandlaminarcalcificationofcartilageinloosebodies
asprobableevidenceofanossificationabnormality.JPathol.
1980;132:161–168.
25.YonetaniY,NakamuraN,NatsuumeT,etal.Histologicalevaluationof
juvenileosteochondritisdissecansoftheknee:acaseseries.KneeSurg
SportsTraumatolArthrosc.2010;18:723–730.
26.UozumiH,SugitaT,AizawaT,etal.Histologicfindingsandpossiblecauses
ofosteochondritisdissecansoftheknee.AmJSportsMed.2009;37:2003–
2008.
27.KozlowskiK,MiddletonR.Familialosteochondritisdissecans:adysplasia
ofarticularcartilage?SkeletalRadiol.1985;13:207–210.
28.PhillipsHO,GrubbSA.Familialmultipleosteochondritisdissecans.Report
ofakindred.JBoneJointSurgAm.1985;67:155–156.
29.WrightRW,McLeanM,MatavaMJ,etal.Osteochondritisdissecansofthe

knee:long-termresultsofexcisionofthefragment.ClinOrthopRelatRes.
2004;(424):239–243.
30.PetriePW.Aetiologyofosteochondritisdissecans.Failuretoestablisha
familialbackground.JBoneJointSurgBr.1977;59:366–367.
31.FairbanksH.Osteochondritisdissecans.BrJSurg.1933;21:67–82.
32.MitsuokaT,ShinoK,HamadaM,etal.Osteochondritisdissecansofthe
lateralfemoralcondyleofthekneejoint.Arthroscopy.1999;15:20–26.
33.StanitskiCL,BeeJ.Juvenileosteochondritisdissecansofthelateralfemoral
condyleafterlateraldiscoidmeniscalsurgery.AmJSportsMed.


2004;32:797–801.
34.YoshidaS,IkataT,TakaiH,etal.Osteochondritisdissecansofthefemoral
condyleinthegrowthstage.ClinOrthopRelatRes.1998;(346):162–170.
35.JacobiM,WahlP,BouaichaS,etal.Associationbetweenmechanicalaxisof
thelegandosteochondritisdissecansoftheknee:radiographicstudyon103
knees.AmJSportsMed.2010;38:1425–1428.
36.RibbingS.Studiesonhereditary,multipleepiphysealdisorder.ActaRadiol.
1937;34:1–107.
37.DeSmetAA,FisherDR,GrafBK,etal.Osteochondritisdissecansofthe
knee:valueofMRimagingindetermininglesionstabilityandthepresence
ofarticularcartilagedefects.AJRAmJRoentgenol.1990;155:549–553.
38.DeSmetAA,IlahiOA,GrafBK.ReassessmentoftheMRcriteriafor
stabilityofosteochondritisdissecansinthekneeandankle.SkeletalRadiol.
1996;25:159–163.
39.GuhlJF.Arthroscopictreatmentofosteochondritisdissecans.ClinOrthop
RelatRes.1982;(167):65–74.
40.ChambersHG,SheaKG,AndersonAF,etal.AmericanAcademyof
OrthopaedicSurgeonsclinicalpracticeguidelineon:thediagnosisand
treatmentofosteochondritisdissecans.JBoneJointSurgAm.

2012;94(14):1322–1324.
41.GlancyGL.Juvenileosteochondritisdissecans.AmJKneeSurg.
1999;12:120–124.
42.KocherMS,MicheliLJ,YanivM,etal.Functionalandradiographic
outcomeofjuvenileosteochondritisdissecansofthekneetreatedwith
transarticulararthroscopicdrilling.AmJSportsMed.2001;29:562–566.
43.CahillBR,AhtenSM.Thethreecriticalcomponentsintheconservative
treatmentofjuvenileosteochondritisdissecans(JOCD).Physician,parent,
andchild.ClinSportsMed.2001;20:287–298,vi.
44.WallEJ,VourazerisJ,MyerGD,etal.Thehealingpotentialofstable
juvenileosteochondritisdissecanskneelesions.JBoneJointSurgAm.
2008;90:2655–2664.
45.CahillBR,PhillipsMR,NavarroR.Theresultsofconservativemanagement
ofjuvenileosteochondritisdissecansusingjointscintigraphy.Aprospective
study.AmJSportsMed.1989;17:601–605;discussion605–606.


46.GarrettJC.Freshosteochondralallograftsfortreatmentofarticulardefectsin
osteochondritisdissecansofthelateralfemoralcondyleinadults.Clin
OrthopRelatRes.1994;(303):33–37.
47.GebarskiK,HernandezRJ.Stage-Iosteochondritisdissecansversusnormal
variantsofossificationinthekneeinchildren.PediatrRadiol.
2005;35:880–886.
48.NawataK,TeshimaR,MorioY,etal.Anomaliesofossificationinthe
posterolateralfemoralcondyle:assessmentbyMRI.PediatrRadiol.
1999;29:781–784.
49.EdmondsEW,AlbrightJ,BastromT,etal.Outcomesofextra-articular,
intra-epiphysealdrillingforosteochondritisdissecansoftheknee.JPediatr
Orthop.2010;30:870–878.
50.WallE,VonSteinD.Juvenileosteochondritisdissecans.OrthopClinNorth

Am.2003;34:341–353.
51.O’ConnorMA,PalaniappanM,KhanN,etal.Osteochondritisdissecansof
thekneeinchildren.AcomparisonofMRIandarthroscopicfindings.J
BoneJointSurgBr.2002;84:258–262.
52.KijowskiR,BlankenbakerDG,ShinkiK,etal.Juvenileversusadult
osteochondritisdissecansoftheknee:appropriateMRimagingcriteriafor
instability.Radiology.2008;248:571–578.
53.WilsonJN.Adiagnosticsigninosteochondritisdissecansoftheknee.JBone
JointSurgAm.1967;49:477–480.


CHAPTER

22

SamirK.Trehan
BentonE.Heyworth



TransarticularDrillingof
OsteochondritisDissecans
INTRODUCTION
Osteochondritisdissecans(OCD)isapathologicprocessaffectingsubchondral
boneandtheadjacentarticularcartilage,whichcanadvancetoosteochondral
fragmentinstabilityandloosebodyformation.1Thepathophysiologicbasisfor
theconditionhasnotbeendefinitivelyestablishedbut,inmostcasesisfeltto
involveimpairedvascularsupplyand/orhealingresponseinafocalareaof
subchondralbone.Thekneeisthemostcommonlyaffectedjointinthebody,
withthemedialfemoralcondylebeingmostfrequentlyaffectedlocationinthe

knee.Althoughdescribedinadults,OCDmostcommonlyaffectsskeletally
immaturepatients(whichhastraditionallybeenreferredtoasjuvenileOCD,a
termmostauthorsnowdonotfavor),whichgenerallyhaveabetterprognosis
thanOCDinadults.
PresentingsymptomsofOCDarevariablebutmayincludelimp,kneepain,
mechanicalsymptomssuchasclickingorlocking,and/oreffusion.Obtaining
standardanteroposterior(AP)(Fig.22.1)andlateralkneeradiographs,withthe
additionofatunnelornotchview,toallowforbetterassessmentofthe
commonlyaffectedposterioraspectofthecondylerepresentstheprimaryinitial
diagnosticstep.However,magneticresonanceimaging(MRI)hasemergedas
thegoldstandardforclassifyingOCDlesionsaseitherstable(Fig.22.2A,B)or


unstable,adistinctionthatdictatestheapproachtotreatment.Stablelesionsare
generallycharacterizedbyanintactarticularcartilagesurfaceoverlyingthe
subchondralbonelesion.Thischapterfocusesonthemostcommonlyused
techniqueofsurgicaltreatmentforstablejuvenileOCDlesionstransarticular
drilling.


TREATMENT
Indications
PatientswithstableOCDlesionsofthekneeandopenphysesshouldbeinitially
managednonoperatively.Conservativemodalitiesmayrangefromsimple
activitymodification(i.e.,avoidingoffendingactivitiessuchasrunning,
competitivesports,orimpactactivities)toprotectedweightbearingwith
crutches,bracing,and/orimmobilization.Althoughthetraditionalapproachof
immobilizationthroughcastingisrarelypursuedtoday,useofalockedhinged
kneebraceorkneeimmobilizermaybeareasonablesubstitute.Alternatively,
valgusunloaderbracinghastheadvantageofallowingkneemotionwhilestill

decreasingloadsontheaffectedmedialfemoralcondyle.Incomplete,transient,
orlackofradiographicorMRI-basedhealingafterapproximately3to6months
ofnonoperativetreatmentisgenerallyanindicationforsurgicalintervention.2

OPERATIVE
DrillingoftheOCDlesionisconsideredthegoldstandardofsurgicaltreatment
ofstableOCDinskeletallyimmatureknees.Drillingisthoughttodisruptthe
scleroticmarginofthelesionandconsequentlypromotehealingviagrowth
factorsreleasedfromhealthyunderlyingcancellousbone.Therearetwoprimary
methodsofdrillingkneeOCDlesions:transarticular(orretrograde)and
retroarticular(alsoreferredtoasanterograde).
Transarticulardrillinginvolvesperforationofthearticularcartilageunder
directvisualizationoftheOCDlesion,whereasretroarticulardrillingthroughthe
affectedcondylesparesthearticularcartilageandusedintraoperative
fluoroscopytoperforatetheOCDfromthebacksideofthelesion.Theoretical
advantagesofthetransarticulardrillingtechniqueincluderelativetechnical
simplicityanddirectvisualizationandassessmentoftheOCDlesionandK-wire
entrysitesduringdrilling(Fig.22.3).Consequently,drillingpassescanbe
optimallyspacedrelativetothelesion’sbordersandrelativetoeachother(Fig.
22.4).Finally,intraoperativefluoroscopyisrarelyrequiredtolocalizelesion(in
contrasttoretroarticulardrilling).Thetheoreticaldisadvantageoftransarticular
drillingisthedirectarticularsurfaceperforationanduncertainlong-term
implicationsoftheseperforations.Also,fromapracticalstandpoint,farposterior


condylarlesionsmaybedifficulttoaccessfromtransarticularapproachdueto
requiredkneehyperflexion.

Althoughhigh-qualitystudiescomparingtransarticularandretroarticular
drillingarelacking,arecentsystematicreviewconcludednodifferencein

patientoutcomesorradiographichealingforskeletallyimmaturepatientswith
stablekneeOCDlesionstreatedwiththesetechniques.Intotal,12studieswere
includedwhichdemonstrateda91%radiographichealingrateinaverageof4.5
monthsfortransarticulardrillingversus86%inaverageof5.6monthsfor


retroarticulardrillingwithnodifferencesinclinicalresponse.3
Asummaryofthecurrentlypublishedliteratureontransarticulardrillingin
isolationforthetreatmentofjuvenilekneeOCDlesionsissummarizedinTable
22.1.4–9Acrossallstudies,thereisagreementontheindicationsfordrilling,in
thatalmosteverypatienthadastableOCDlesionwhichhadfailedatleast3
monthsofnonoperativemanagement.Theclinicalandradiographicoutcomes
reportedareexcellentoverall.Significantvariabilityexistsintermsof
intraoperativedrillingtechniqueandpostoperativerehabilitation.


AUTHORS’PREFERREDTREATMENT


SurgicalTechnique
Thepatientispositionedsupineonoperatingroomtablewithanonsterilethigh
tourniquet.Standardanterolateralandanteromedialkneearthroscopicportalsare
establishedandadiagnosticarthroscopyisperformed.TheOCDlesionisthen
assessedforsizeandstability.Thelatterisespeciallyimportant,giventhatin
rareinstances,a“stable”lesionmayhaveprogressedsincethetimeoflast
imagingthatmayalteroperativeplan.Drillingisperformedwitheithera0.062or0.045-inK-wire.WeprefertousethesmallersizedK-wiretominimize
disruptionofthearticularcartilage.TheK-wireisdirectedperpendiculartothe
chondralsurfaceandadvancedwithapowerwiredrivertoanappropriatedepth
throughtheaffectedsubchondralbone.Withdrawalofthewirefromthearticular
surfacemayoccasionallybefollowedbyvisualizationoffatbubblesand/orbone

marrowbleeding,indicatingadequateperforationintothedeeperhealthy
cancellousbone.Althoughinjurytothedistalfemoralphysisisaconcernwith
overpenetrationduringdrilling,thiscanbeavoidedbymeasuringthedistance
fromthearticularsurfacetothephysisonpreoperativeMRIandensuring
penetrationnodeeperthantheshorteddistance.However,wearenotawareof
anypublishedcasesdetailingthiscomplication.Inordertosuccessfullyaccess
thelesionandappropriatelyorienttheK-wireperpendiculartothechondral
surface,thewiremayneedtobeintroducedpercutaneouslyatseveralsitesand
kneepositionmayneedtobeadjustedinvaryingdegreesofflexion.Thenumber
anddistributionofdrillingpassesvariesconsiderablyintheliterature.
Dependingonlesionsize,wewillgenerallyuse6to10drillingpasseswith
articularcartilageperforationsbeingevenlydistributedandseparatedbyatleast
severalmillimeterssoasnottopromotelesioninstability.10

Rehabilitation
TheoptimalapproachtopostoperativerehabilitationafterOCDdrillingisnot
wellestablishedintheliterature.Theauthors’preferenceistoprotectweight
bearingwithcrutchesfor6weekswhileperformingnon–weight-bearingknee
rangeofmotionexercisestooptimizethehealingenvironmentforthe
subchondralbonewhileminimizingchanceoflesionprogressionorinstability.
Rangeofmotionispursuedasearlyaspossiblepostoperatively.Postoperative


radiographsareobtainedatfollow-upvisitsevery6to12weekstoassess
healing,whichischaracterizedbyprogressivedecreaseinthescleroticmarginof
thelesionandincreaseintheopacificationofthepriorlucency(Fig.22.5).
Returntosportisdictatedbybothclinicalandradiographicimprovement.
Specifically,significantsubchondralbonehealingonradiographsorMRI,as
wellasresolutionofpain,effusion,andanyquadricepsorhamstringweakness,
shouldbeobservedpriortoallowingreturntoathletics.Thesecriteriaare

especiallyimportantforpatientsplanningtoreturntosportsrequiringcutting,
pivoting,andimpactactivities.Finally,patientsarefollowedforatleast1year
afterreturntoplaytoensurecontinuedsymptomaticimprovementandno
radiographicevidenceoflesionprogression.

Complications/PearlsonAvoiding
Complicationratesforthisprocedureareverylowwithnoreported
complicationsinthecaseseriesreferencedinTable22.1.Theoretical
intraoperativeconsiderationsincludekeepingthedrillperpendiculartothe
chondralsurfacewitheachdrillpasstoavoidiatrogenicinjuryaswellashaving


fluoroscopyavailableintraoperativelyincaselesionmarginscannotbe
ascertainedfromprobe-basedpalpation.Postoperatively,patientscandevelop
stiffness,whichiswhywefavorimmediatepostoperativerangeofmotionas
opposedtoimmobilizationinacast,forexample.

REFERENCES
1.HughstonJC,HergenroederPT,CourtenayBG.Osteochondritis
dissecansofthefemoralcondyles.JBoneJointSurgAm.1984;66(9):1340–
1348.
2.KocherMS,TuckerR,GanleyTJ,etal.Managementof
osteochondritisdissecansoftheknee:currentconceptsreview.AmJSports
Med.2006;34(7):1181–1191.
3.GuntonMJ,CareyJL,ShawCR,etal.Drillingjuvenile
osteochondritisdissecans:retro-ortransarticular?ClinOrthopRelatRes.
2013;471(4):1144–1151.
4.AgliettiP,BuzziR,BassiPB,etal.Arthroscopicdrillinginjuvenile
osteochondritisdissecansofthemedialfemoralcondyle.Arthroscopy.
1994;10(3):286–291.

5.AndersonAF,RichardsDB,PagnaniMJ,etal.Antegradedrillingfor
osteochondritisdissecansoftheknee.Arthroscopy.1997;13(3):319–324.
6.CeperoS,UllotR,SastreS.Osteochondritisofthefemoralcondylesin
childrenandadolescents:ourexperienceoverthelast28years.JPediatr
OrthopB.2005;14(1):24–29.
7.KocherMS,MicheliLJ,YanivM,etal.Functionalandradiographic
outcomeofjuvenileosteochondritisdissecansofthekneetreatedwith
transarticulararthroscopicdrilling.AmJSportsMed.2001;29(5):562–566.
8.LouisiaS,BeaufilsP,KatabiM,etal.Transchondraldrillingfor
osteochondritisdissecansofthemedialcondyleoftheknee.KneeSurg
SportsTraumatolArthrosc.2003;11(1):33–39.
9.BradleyJ,DandyDJ.Resultsofdrillingosteochondritisdissecans
beforeskeletalmaturity.JBoneJointSurgBr.1989;71(4):642–644.
10.HeyworthBE,EdmondsEW,MurnaghanML,etal.Drillingtechniquesfor
osteochondritisdissecans.ClinSportsMed.2014;33(2):305–312.