The Effect of Site and Type of Nerve Injury on Spinal Glial Activation and Neuropathic Pain Behavior

Apreclinicalpostlaminectomyratmodelmimicsthe

humanpostlaminectomysyndrome

JenniferB.Massiea,BillHuanga,ShelleyMalkmusb,TonyL.Yakshb,

ChollW.Kima,StevenR.Garfina,WayneH.Akesona,∗

ab

DepartmentofOrthopaedics,VeteransAdministrationSanDiegoHealthCareSystemandUniversityofCalifornia,

3350LaJollaVillageDrive,SanDiego,CA92161-112D,USA

DepartmentofAnesthesiology,VeteransAdministrationSanDiegoHealthCareSystemandUniversityofCalifornia,

3350LaJollaVillageDrive,SanDiego,CA92161-112D,USA

Received12November2003;receivedinrevisedform26February2004;accepted26February2004

Abstract

Chroniclowbackpainwithsciaticacomplicatingpostlaminectomysurgeryispoorlyunderstood.Itislikelythatsomeaspectsofpersistentpainofthesyndromeresultsfromspinalfacilitationinwhichthereisloweringofpainexcitationlevels.Asmallanimalpreclinicalmodelisneededthatmimicstheclinicalconditiontopermitdetailedstudiesoftheunderlyingalteredneurochemistryofthesensorypathways.Weproposehereinaratlaminectomymodelcontainingtheelementsrequiredforstudyoftheneurobiologyofthecondition.ThemodelconsistsofasurgicallaminectomythatincludesL5spinalnervemanipulationanddiscinjury,elementsnecessarilyemployedinhumandischerniationsurgery.At8weekspostlaminectomytheproposedmodeldemonstratesparaspinousmusclespasm,tailcontracture,behavioralpainbehavior,tactileallodynia,epiduralandnerverootscarring,andnerverootadherencebyscartotheunderlyingdiscandadjacentpedicle.Twounderlyingpainfacilitationstatesareinvokedintheclinicalcondition:(1)aninflammatorystaterequiredtoachievewoundhealing;and(2)anerveinjurystateresultingfromnervemanipulationandsubsequentepiduralscarring,spinalnervescarring,andspinalnervetetheringtotheadjacentdiscandpedicle.Bothpainfacilitationstatesareactiveinthemodel.©2004ElsevierB.V.Allrightsreserved.

Keywords:Preclinicalmodel;Spine;Laminectomy;Discinjury;Nerveroot;Tactileallodynia

1.Introduction

Theunderlyingneurobiologyoflowbackpainispoorlyunderstood,andasaresultmostclinicalapproachesinthepastseveraldecadeshavefocusedontheanatomicalpathologyofbackdisorders.Suchtreatmentshaveevolvedwithoutaclearunderstandingoftheneurobiologyofbackpain,thepresentingcomplaint.Themajorityofdiagnosticandtherapeuticmeasuresdealingwithbackpainandsci-aticahavefocusedonstructuralelementsofthecondition.Theyhaveembracedabreadthofconservativeapproachesincludingoralmedication,injections,physicaltherapyincludingmanipulationandbracing.Patientsfailingconser-vativetreatmentnotuncommonlyundergoattemptsatsur-gicalcorrectionemployingavarietyofgoalsandavariety

Correspondingauthor.Tel.:+1-858-552-8585x3841;fax:+1-858-552-4350.

E-mailaddress:wakeson@ucsd.edu(W.H.Akeson).

0165-0270/$–seefrontmatter©2004ElsevierB.V.Allrightsreserved.doi:10.1016/j.jneumeth.2004.02.036

∗

oftechniques.WiththeevolutionofCTandMRimagingtheconsiderationofsurgicalcorrectionofobservedpatho-logicalanatomyhasbeenaccelerated.However,surgicalcorrectionofanatomicaldisordersisnotwithoutsideeffectsandcomplications.Outcomemeasureshavedemonstratedthefailureofpresentapproachesinsignificantnumbersofcasesinrespecttopainreliefandwithconsiderableaddedriskinsomeofthem(Nachemson,1993,1996;Nachemsonetal.,2004).Repeatlaminectomyfacesaddedriskssec-ondarytoanatomicalchangesincaseswhereproliferativefibroplasishasdevelopedwithinthespinalcanal.Suchrisksincludedissectioninjurytospinalnerves,inadvertenttear-ingofthedura,andadditionalproliferativescarringwithinthecanalpostoperatively.

Sincethereislimitedknowledgeofbasicneurobiologicalmechanismsthatgeneratebackpainandsciatica,thebasisforpharmacologicaltreatmentofbackpainisalsoalmostentirelyempirical.Theuseofnarcoticshasseriousdeleteri-oussideeffects.Theyrapidlyloseeffectivenesswithtime,

284J.B.Massieetal./JournalofNeuroscienceMethods137(2004)283–289

makingtheiruseinchronicconditionssuchasdisablingbackpainlargelycontraindicated.Variousinjectionsfortheconditionarecommonlyemployed,usuallywithonlytem-poraryrelief,andwiththeirownsetofsideeffects,thatoccasionallymayaddtothelevelofdisability.

Preclinicalanimalmodelsarerequiredthatmirrorchangesseeninthepostlaminectomysyndromeinordertoclarifyitsneurochemistryandunderlyingneuropathology.There-quiredmodelsmustincludepainevokedbehavioralchanges,andassociatedgrossandhistologicalfeaturesconsistentwiththeclinicalsyndrome.Suchmodelsshouldbeconducivetocharacterizingneurochemicalalterationprofilesinthespinalnerve,dorsalrootganglia(DRG),nerverootanddorsalhornofthespinalcord.ArecenteditorialinthejournalPainpointsoutthatifsuchpre-clinicalmodelscanbedeveloped,theywillbeusefulin:(1)developinganunderstandingoftheneurobiologicalmechanismsunderlyingthedisorder;(2)canbeusefulinstudyingtherisk/benefitratiosofexistingtherapies;(3)canprovideinsightsusefulindevelopingnoveltherapies(Mantyh,2002).

Wedescribeinthispaperaratmodelforstudyingtheneu-robiologyofthepostlaminectomysyndrome.Theratspeciesisreadilyavailable,relativelyinexpensiveandiscommonlyusedforresearchontheneurobiologyandneurochemistryofpain(DeLeoandWinkelstein,2002;Huntetal.,2001;KimandChung,1992;Yakshetal.,1999).Ourmodelcon-sistsoflaminectomiesattheL5andL6levelwhichaL5–6discinjuryisperformedononesideonly.GentleretractionoftheL5spinalnerveisrequiredtoexposethedisc.Evalua-tionofthepostlaminectomyoutcomeincludesserialtactileallodyniatesting,dailyobservationofpainrelatedauton-omybehavior,recordingofpostsurgicalpainmedicationrequirement,evidenceofparaspinousmusclespasm,grossobservationsofdissectedspecimensrelatingtoproliferativescarformationwithinthespinalcanalalongwithassessmentofnerverootscarringandtetheringtounderlyingdiscandadjacentpedicle.Additionalassessmentmeasuresincludequantitativeestimationofcollagencontentwithinthespinalcanalandcorroborativehistopathologicalobservationsonasimilarlytreatedgroupofratspostlaminectomy.Themodeldescribedshowspromiseasamodelusefulforstudiesoftheneurobiologyofthepostlaminectomypainsyndromeasitmirrorsmanyoftheclinicalfeaturesofthedisorderinthehumancondition.

GeneralizationsfrompriorpainmodelsincludingthoseofChung(DeLeoandWinkelstein,2002;DeLeoandYezierski,2001;Huntetal.,2001;KimandChung,1992),acancerpainmodel(Hableretal.,2000;Honoreetal.,2000c;Mantyh,2002)andtheinflammatoryCFA(CompleteFreund’sAd-juvant)pawinjectionmodel,cannotbemade,becauseeachisreportedtohaveuniqueneurochemicalandcytokinepro-filecharacteristics(DeLeoandWinkelstein,2002;DeLeoandYezierski,2001;Honoreetal.,2000b).Itisevidentfromthesestudiesthatparticularpainevokedsyndromesre-quireprecisecharacterizationoftheiruniqueneurobiologi-cal/neurochemicalsignaturestopermitfullunderstandingof

underlyingprocessesandrationaldesignofpotentialphar-macologicalcorrectionoftheneurochemicalorcytokineim-balance.

Theopportunityforprogressinthisfieldappearstobehighlypromising.Ifcharacteristicneurochemicalandcy-tokinechangescanbeclearlyidentifiedinamodelmirroringtheclinicalpostlaminectomysyndrome,newpharmaco-logicalstrategiesaimedatneutralizingkeyelementsoftheparticularpainrelatedcascadesincitedbecomeclosertorealization.Thedevelopmentofsuchpreclinicalmodelsiscrucialtoresolutionofcertainpersistentclinicalquestionsthatcannotbeansweredbyclinicaltestprotocolsinhumans.

2.Methods

Forty-fourmaturemaleHarlanSprague–Dawleyrats,weighing>400gwereusedinthisstudy.Allanimalsweremaintainedinastandardcontrolledenvironment.Bilaterallaminectomies(L5andL6)andarightunilateraldiscinjury(L5–6)wereperformedon25oftherats,untreatedgroup(protocolapprovedbytheVAInstitutionalAnimalCareUseCommittee).

Additionalfourmalerats(weightmatched)receivedashamoperation.Fifteenweightmatchedmaleratsservedasnon-operativecontrols,normalgroup.2.1.Surgicalprocedure

Undergeneralanesthesia(usingisoflurane)administeredbymask,theratsweremonitoredforpulserateandoxygensaturation.Understerileconditionsaposteriormidlineinci-sionwasmadeexposingthebonyposteriorelementsfromL4toL7.Theshamoperativeanimalsincludedthispartofthesurgicalprocedureonly.Usinganoperativemicroscope(Super-Lux400,CarlZeissInc.),bilaterallaminectomieswereperformedatL5andL6.Duetothesmallsize(2mm)ofthelaminae,thelaminectomywasperformedbilaterallytoalloweasyandsafeaccesstotherightdisc.Additionally,bi-laterallaminectomypermittedasidetosidecomparisonbe-tweeneffectoflaminectomyaloneversuslaminectomyplusdiscinjuryandspinalnerve/duralsleeveretraction.Anoph-thalmologybovieandsmallamountofbonewaxwasappliedtotheexposedcancelloussurfacesoftherongueredlaminaeinordertopreventspontaneousclosureofthelaminotomy.Theduraandneuralelementswerethenretractedmediallyundermoderatetensionexposingthepostero-lateralaspectoftherightL5–6disc.A27gaugeneedlewastheninsertedintotheexposeddisccreatingarightunilateraldiscinjury.Duetotheextremesmallsize(approximately1mm)ofthediscthe27gaugeneedlewaswedgecreatingalargeholeinrelationshiptothesizeofthedisc.Theneedlewasmanipu-latedina“joystick”mannertoenlargetheopeningbeforewithdrawalcreatingalargesurfacedefect.Thewoundswereclosedinaroutinemanner.Theanimalsweresacrificedat3and8weeksforanalysis.

J.B.Massieetal./JournalofNeuroscienceMethods137(2004)283–289285

2.2.Postoperativemonitoring

Theanimalswererecoveredwithoxygenbymaskonawaterfilledthermalpad.Followingrecovery,thepostlaminectomyanimalsrequiredBuprenex(buprenor-phine)painmedicationtwiceadayoverthefirstthreepost-operativedays.Theratsbegantodevelopananoma-loustailpostureconcavetothediscinjurysidearoundthethirdandfourthweek.Thistailposturepersistedthroughtheendofthe8weeksofthestudy.Weattributethisfindingtoasymmetryofparaspinousmusclespasminthelumbarregion,moreforcefulonthediscinjuryside.2.3.Behavioralpainobservations

Behavioralchangeswererecordedforeachanimalfollow-ingdailyinspectionduringthepost-operativeperiod.Someoperated(withlaminectomyanddiscinjury)animalsbegantodemonstrateautotomy(Dixon,1980)betweenthefifthandsixthweekpostoperativelywithoccasionalchewingbe-haviordirectedagainsttheirtail.Theyalldemonstratedex-cessivegroomingandalteredsteppingandplacing.Theratswithself-inducedlesionsrequiredadditionaloralandtopi-calpainmedicationtomoderatetheselfmutilation.Almostalltheself-inducedlesionswereeliminatedbylayeredthebeddingthickerintheircageboxpreventingthetailorpawsfromtouchingahardcoldsurface.Painmedicationwasin-terrupted12hpriortothetactileallodyniatestingprocedure.2.4.Tactileallodyniaprocedure

Behavioraltestingfortactileallodyniawasperformedonfivenon-operative,fourshamoperativeandfivelaminec-tomyratsatweeklyintervalsbeginningat3weeks.Abe-havioraltestcagewithawiremeshfloorallowedforthetactileallodyniatesting.GraduatedvonFreyhairs(Stoelt-ing,Il.)whosestiffnessincreasedlogarithmicallyfrom0.41to15gwereusedforthesetests.Thetechniqueusedwastheso-calledupanddownmethoddescribedbyDixonandmod-ifiedbyYaksh(Chaplanetal.,1994;Dixon,1980;Yaksh,1985).Apositiveresponsewasnotedifasharpwithdrawalwaselicitedwhentheplantaraspectofthepawwastouched

byavonFreyhairontheL5sensorydermatomearea.Test-ingcontinueduntilthemaximumstimuluswasreachedoruntilthevonFreyhairstrengthwasreachedcausing100%withdrawalresponse.

2.5.DissectionprocedureforhydroxyprolineassessmentEightratsinthenon-operativegroupandeightratsintheoperativegroupweresacrificedat3weeksandeightat8weeksandfourratsintheshamgroupweresacrificedat8weeksforhydroxyprolineanalysisofthespinalcanalprolif-erativescar.Aftertheratsweresacrificedtheoperativearea(sameareaforthenon-operativerats)wasdissectedthroughananteriorapproach.Thisapproachallowedclearvisualdiscriminationbetweenposteriorincisionalwoundhealingandepiduralandperineuralscarring.Epiduralscarringwasevidentbelowthelevelofthelaminectomyspaceandwasadherenttothedorsalaspectoftheduraandcouldbereadilydissectedwiththedurafromtheoverlyingwound.ThedurawasremovedwithattachedscarfromthecaudalaspectofthebodyofL4tothecephaladaspectofthebodyofL7(1.5cminlength)includingtheattachedscar(Fig.1).TheL5spinalnervewasenvelopedbytheproliferativescarringprocessandadherenttotheunderlyingdiscandadjacentpedicle.Ithadtobesharplydissectedfreefromthosestructures.Thenon-operativeandshamoperativeratsweredissectedinasimilarfashiontoyieldaspecimenofsimilardimensions.2.6.Biochemicalanalysis

Thesampleswereanalyzedbiochemicallyforhydrox-yprolineafterfatextractionwithacetoneanddeterminationoffatfreedryweights.Thetissueswerehydrolyzedwithhy-drochloricacidtodeterminethehydroxyprolinecontentbyamodifiedWoessnermethod(Woessneretal.,1967).Theamountoftotalcollagenwasderivedfromhydroxyprolinecontentandexpressedinmilligramsoffatfreedrytissue.2.7.Histologyanalysis

Twoanimalsineachgroupwereusedforhistologicalstudies.Thelumbarspecimenswereremovedenblocand

Fig.1.Schematicandhistologicalsectionillustratingspecimensobtainedforbiochemicalanalysis.Whitelineoutlinesspinalcanalspecimenforhydroxyprolineanalysis.

286J.B.Massieetal./JournalofNeuroscienceMethods137(2004)283–289

fixedin10%bufferedformalinfollowedbydehydrationinethanol,theninfiltratedandembeddedinmethylmethacry-late.Afterthemethylmethacrylatehardened,thespecimenswereaxialcutandmilledintoapproximately100␮msec-tions.ThesectionswerestainedwithMasson–GoldnerTrichromeforcollagen.2.8.Statisticalanalysis

StatisticalanalyseswereperformedusingANOVAandaFishercomparisont-test.

3.Results

Theoperative(untreatedlaminectomygroup)ratsdevel-opedatailcontractureconsistentwithasymmetricallumbarmusclespasm.Itpresentedasaconcaveposturetowardtheipsilateraldiscinjuryside(Fig.2).Theanimalsshowedpainrelatedbehaviorbeginningatthethirdweekpostoper-atively.Inthemodeldescribedtheratsdemonstratedexces-sivegrooming,alteredsteppingandplacing,andoccasionalself-destructivebehavioragainsttheirtail.Theselfmuti-latingratsrequiredincrementalpainmedicationconsistingofBuprenex(buprenorphine)twiceadayaswellastopi-calapplicationofabittertastingmedication(ItchRelief,Petgold),tosurfacesofthetailtodiscouragetheautotomybehavior.Thickerfloorbeddingmarkedlydecreasedtheautotomybehavior.Epidural/periduralscarringwasgrosslyevidentatthetimeofdissectionofthegrossspecimen.Thespinalnerveattheoperativelevelonthediscinjurysidewasgrosslyscarred,andwasadherenttothediscandadjacentpedicle.Thespinalnerveonthenon-discinjurysidewaslessadherentinmostcases.Theproliferativescarringprocesswasprogressivebetweenthethirdandeighthweeksofthestudy.Histologicalstudiesemployedhardtissueprocessing

Fig.2.Theseareaphotographofarepresentativeexampleofa8weeksnon-operative(normal)andapostlaminectomy(untreated)ratthathadsustainedaunilateraldiscinjuryontherightside.Thephotoontherightdemonstratesanasymmetrictailcontractureconcavetowardthediscinjurysideinthelaminectomygroup.

Fig.3.ThisisaphotographofanaxialsectionhardtissueprocessingforMasson–Goldnerstainingconfirmingthelocationofperiduralandperineuralfibrosisshownbytheredarrowfromoperativelaminectomyuntreatedgroupspecimen.

forMasson–Goldnerstainingconfirmedthepresenceofperiduralandperineuralfibrosis(Fig.3).Thecollagencon-tentoftheperiduralspace,asdemonstratedbybiochemicalanalysisforhydroxyproline,wasmoderatelyincreasedatthethirdweekanddramaticallyincreasedfurtherattheeighthweek,theeighthweekgroupbeingapproximatelyfourtimesgreater.ThedifferencebetweenthethirdandeighthweekgroupshadP=0.0011(Fig.4).Thedifferencebetweenoperated(untreated)andnon-operative(normal)groupsissignificantatbothtimeperiods,withPvalues0.0003and0.0001,respectivelyat3and8weeks.Additionally,theshamoperativegroupshowedsignificantdifferencesbe-tweentheoperativegroupat8weeks.Behavioraltestingof

Fig.4.ThisgraphillustratestheoutcomeofhydroxyprolineanalysisfortotalcollagencontentofthedissectedspecimensillustratedinFig.1at0,3and8weeksincludingthenon-operative(normal),shamandoperative(laminectomy,untreated)groups.

J.B.Massieetal./JournalofNeuroscienceMethods137(2004)283–289287

Fig.5.Thisgraphillustratesthe“escaperesponses”(hindlimbwithdrawal)tologarithmicforcegradedvonFreyhairpressureonhindpawsensoryreceptiveareaoftheright(discinjuryside)L5spinalnerve.Measurementsweremadeatweeklyintervalsbeginningat3weeks.

Fig.6.Thisgraphillustratesthe“escaperesponses”(hindlimbwith-drawal)tologarithmicforcegradedvonFreyhairpressureonhindpawsensoryreceptiveareaoftheleft(non-discinjuryside)L5spinalnerve.Measurementsweremadeatweeklyintervalsbeginningat3weeks.

theipsilateraldiscinjuryside(untreatedgroup)usingthevonFreyhairtestingprocedureshowedtactileallodyniaoftheL5sensoryreceptiveareaofthehindpawwithaPvalues0.0001comparedtothenon-operative(normalcontrol)andtheshamoperativegroup(Fig.5).Thenon-discinjuryside(untreatedgroup)showedatendencytoincreasedsensitivitytovonFreytesting,butnotatlevelsofsignificance(Fig.6).4.Discussion

Clinicalexperiencesupportstheviewthatthesizeofthefailedpostlaminectomypatientgroupisnotasmallnumber(Atlasetal.,1996;Chiballetal.,2000;Devulderetal.,1995;Elimimianetal.,1999;Nachemsonetal.,2004;Pearce,2000;Reveletal.,1996).ThesocioeconomiccostsintheUSaloneforbackrelatedinjuriesarecalculatedinthetensofbillionsofdollarsannually(AAOS,1999).Itishighlyunlikelythatcorrectivesurgicalstructuralapproachesalonecanpreventorcorrectthechronicpostlaminectomy

nociceptiveprocessinaconsiderablepercentageofso-called“failedback”cases.

Becauseofrecentrapidadvancesinthestudyofpharma-cologyofpain,thereisanaddedemphasisfortheneedtodevelopreliablepreclinicalmodelsofthepostlaminectomysyndrome.Preclinicalmodelsthatcanmimicthespinalfa-cilitationseeninpostlaminectomypainstatesareneededtoprovidetheinsightsintotheunderlyingnociceptivepro-cesses.Theimportanceofreliablepreclinicalmodelsofthissyndromeisthattheyprovideanopportunitytobegintoun-ravelthesequenceofcomplexnociceptiveprocessinginthespinalnerve,theDRG,andthespinalcordleadingtospinalfacilitation.Suchmodelsofferthepromiseofdevelopmentofimprovedpharmaceuticalstrategiesofdrugselectionanddesigntostrategicallytargetthevariousnociceptivecas-cades.

Asinthehumanpostlaminectomysyndrome,asequenceofeventsevocativeofnociceptiveresponsesisinducedbythesurgicalprocedureemployedinthemodel.Thesestepsinclude:(1)theinitialspinalnerveinjurysecondarytotensionappliedtothenerveatthetimeofitsretractiontoexposethediscspace;(2)theproinflammatoryprocessesin-stitutedaspartofthewoundhealingcascade;(3)latespinalnervescarringandtetheringtothediscandpedicleastheproliferativescarprocessmatures,resultinginaneuropathicprocessanalogoustoentrapmentandscarringofperipheralnervesatothersites(Burcheiletal.,1993;Campbell,2001;Carltonetal.,1994;Millesi,1993).Itispossiblethatno-cioceptivecascadescharacteristicofeachoftheseinjuriesmaybeuniqueanditislikelythatpharmaceuticalstrate-gieswillbedifferentineachcase.Itisclearfromseveralstudiesthatdifferentinsultstoperipheralnervesproducedifferenttypesofspinalfacilitation(Honoreetal.,2000b).Responsestoinflammatoryinsultsareclearlydifferentfrommechanicalinjurytothenerveinmouseandratmodels,forexamplecancernociceptiveprocessesaredifferentthaneitherofthese(Honoreetal.,2000a,2000b,2000c).

Adiscinjuryisproducedintheproposedmodeltocor-respondtothesurgicalprocedureforherniatednucleuspulposusinhumanlaminectomysurgery.TNF-␣andothercytokinesarereleasedbynucleuspulposusatthetimeofdiscinjury(OlmarkerandLarsson,1998;OlmarkerandRydevik,2001;Olmarkeretal.,1995).Experimentalevi-denceprovidedbyOlmarkerandMyers(1998);OlmarkerandRydevik(1998);Igarashietal.(2000)andothers(Burkeetal.,2002;Byrodetal.,1998;Igarashietal.,2000;Kawakamietal.,1996,1997;Kayamaetal.,1996,1998;Randetal.,1997)hasprovidedclearevidenceoftheimpor-tanceofthenucleuspulposusproinflammatoryconstituentsinspinalpainprocessing.Inthemodeldescribed,aunilat-eraldiscinjuryperformedatL5–6inducestactileallodyniaoftheipsilateralL5sensoryreceptivearea.Theresponseislessintenseonthecontralateralside.Itisofinterestthatsomecrossovernociceptiveeffecttotheuninjuredsidehasbeenobservedinotherunilateralspinalnerveinjurymodels(DeLeoandColburn,1995).

288J.B.Massieetal./JournalofNeuroscienceMethods137(2004)283–289

Tactileallodyniainthemodelisevidentat3weeksandcontinuesthroughtheendofthe8weekstudy.Mildnerverootscarringisnotedat3weeksandprogressesthroughtheendofthestudy.Animportantfeatureofthemodelisthepersistenceofthepainbehavioralresponsethrougheightweeks,supportingourproposaloftherelevanceofthemodeltochronicclinicalpostlaminectomypainstates.Therelativeimportanceofnerverootscarringandtetheringinsustainingtheallodyniaresponseremainstobeestablished,butintuitivelyisworthyofcarefulstudy.Thereisclearneedtoevaluatethesequentialneurochemicalandcytokineprofilechangesatvariousstagesofthepostlaminectomyhealingprocesstobetterunderstandthetimerelatednociceptiveprocesses.Suchstudiesobviouslycannotbeperformedinhumanclinicaltrials.Thevalueofthepreclinicalmodeldescribed,isthatitprovidesasuperiortoolforthestudyofmechanismsandpotentialtreatmentsoftheunderlyingpostlaminectomynociceptiveprocessesleadingtospinalfacilitation.

5.Conclusion

Wedescribearatlaminectomymodelthatmimicstheproliferativefibroplasiasoftenencounteredduringrepeatlaminectomyinhumans.Themodeldemonstratesevidenceofparaspinousmusclespasmwithtailcontracturetowardthediscinjuryside,prominentnerverootscarringandad-herencetotheadjacentdiscandpedicle.Itdemonstratesincreasedtactileallodyniaatasignificantlevelonlyonthenerveinjuryside.Thechronicnatureofthepainbehav-ioralresponsesevokedsupportsrelevanceofthemodeltopostlaminectomychronicpainstates.Themodeldescribedshowspromiseforstudyofthemechanismsofdevelop-mentofspinalnociceptivefacilitationpostlaminectomy.Itshouldbeofvalueinprovidingcomparisonswithotherlowlumbarnerveinjurymodels.Additionally,itshouldbeoffurtheruseinassessmentofoutcomesafterapplicationofanti-inflammatorydrugsorotherpharmaceuticalagentspotentiallyusefulfortargetingnociceptivecascades.Themodelisproposedasatoolofutilitytowardtheultimategoalofpreventionormitigationofcertaincasesoftheclinical“failedback”syndrome.

Acknowledgements

VeteransAdministrationMeritGrantMRS0003.References

AAOS.MusculoskeletalConditionsintheUnitedStates.Rosemont,IL,

AmericanAcademyofOrthopaedicSurgeons;1999.p.182.

AtlasSJ,SingerDE,KellerRB,PatrickDL,DeyoRA.Applicationsof

outcomesresearchinoccupationallowbackpain:theMaineLumbarSpineStudy.AmJIndMed1996;29:584–9.

BurcheilKJ,JohansTJ,OchoaJ.Painfulnerveinjuries:bridgingthe

gapbetweenbasicneuroscienceandneurosurgicaltreatment.ActaNeurochirSuppl1993;58:131–5.

BurkeJG,WatsonRW,McCormackD,DowlingFE,WalshMG,Fitz-patrickJM.Intervertebraldiscswhichcauselowbackpainsecretehighlevelsofproinflammatorymediators.JBoneJointSurgBr2002;84:196–201.

ByrodG,RydevikB,NordborgC,OlmarkerK.Earlyeffectsofnucleus

pulposusapplicationonspinalnerverootmorphologyandfunction.EurSpineJ1998;7:445–9.

CampbellJN.Nervelesionsandthegenerationofpain.MuscleNerve

2001;24:1261–73.

CarltonSM,LekanHA,KimSH,ChungJM.Behavioralmanifestationsof

anexperimentalmodelforperipheralneuropathyproducedbyspinalnerveligationintheprimate.Pain1994;56:155–66.

ChaplanSr,BachFW,PogrelJW,ChungJM,YakshTL.Quantitative

assessmentoftactileallodyniaintheratpaw.JNeurosciMethods1994;53:55–63.

ChiballJT,TaitRC,MerysSc.Disabilitymanagementoflowbackpain

injuriesbyemployer-retainedphysicians:ratingandcosts.AmJMed2000;38:529–38.

DeLeoJA,ColburnRW.Theroleofcytokinesinnociceptionandchronic

pain.In:WeinsteinJN,GordonS,RosemeontIL,editors.Lowbackpain.AmericanAcademyofOrthopaedicSurgeons;1995.p.163–85.DeLeoJA,WinkelsteinBA.Physiologyofchronicspinalpainsyn-dromes:fromanimalmodelstobiomechanics.Spine2002;27:2526–37.

DeLeoJA,YezierskiRP.Theroleofneuroinflammationandneuroimmune

activationinpersistentpain.Pain2001;90:1–6.

DevulderJ,BogartL,CastilleF,MoermanA,RollyG.Relevanceof

epidurographyandepiduraladhesiolysisinchronicfailedbacksurgerypatiens[seecomments].ClinJPain1995;11:147–50.

DixonWJ.Efficientanalysisofexperimentalobservations.AnnuRev

PharmocolToxicol1980;20:441–62.

ElimimianJU,SmithPA,IluoreA.Musculoskeletaldisordersandthe

consumptionofdisabilityproducts.HealthMarkQ1999;17:83–98.HablerH,EschenfelderS,LiuXG,JanigW.Sympathetic-sensorycoupling

afterL5spinalnervelesionintheratanditsrelationtochangesindorsalrootganglionbloodflow.Pain2000;87:335–45.

HonoreP,MenningPM,RogersSD,NicholsML,MantyhPW.Neuro-chemicalplasticityinpersistentinflammatorypain.ProgBrainRes2000a;129:357–63.

HonoreP,RogersSD,SchweiMJ,Salak-JohnsonJL,LugerNM,

SabinoMC,etal.Murinemodelsofinflammatory.Neuroscience2000b;98:585–98.

HonoreP,SchweiJ,RogersSD,Salak-JohnsonJL,FinkeMP,Ramnaraine

ML,etal.Cellularandneurochemicalremodelingofthespinalcordinbonecancerpain.ProgBrainRes2000c;129:389–97.

HuntJL,WinkelsteinBA,RutkowskiMD,WeinsteinJN,DeLeoJA.

Repeatedinjurytothelumbarnerverootsproducesenhancedme-chanicalallodyniaandpersistentspinalneuroinflammation.Spine2001;26:2073–9.

IgarashiT,KikuchiS,ShubayevV,MyersRR.VolvoAwardwinnerin

basicsciencestudies:exogenoustumornecrosisfactor-alphamimicsnucleuspulposis-inducedneuropathology.Molecular,histologic,andbehavioralcomparisoninrats.Spine2000;25:2975–80.

KawakamiM,TamakiT,HashizumeH,WeinsteinJN,MellerST.The

roleofphospholipaseA2andnitricoxideinpain-relatedbehaviorproducedbyanallograftofintervertebraldiscmaterialtothesciaticnerveoftherat.Spine1997;22:1074–9.

KawakamiM,TamakiT,WeinsteinJN,HashizumeH,NishiH,Meller

ST.Pathomechanismofpain-relatedbehaviorproducedbyallograftsofintervertebraldiscintherat.Spine1996;21:2101–7.

KayamaS,KonnoS,OlmarkerK,YabukiS,KikuchiS.Incisionof

theanulusfibrosisinducesnerverootmorphologic,vascular.Spine1996;21:2539–43.

J.B.Massieetal./JournalofNeuroscienceMethods137(2004)283–289

289

KayamaS,OlmarkerK,LarssonK,Sjogren-JanssonE,LindahlA,Ryde-vikB.Culturedautologousnucleuspulposuscellsinducefunctionalchangesinspinalnerveroots.Spine1998;23:2155–8.

KimSH,ChungJM.Anexperimentalmodelforperipheralneuropa-thyproducesbysegmentalspinalnerveligationintherat.Pain1992;50:355–63.

MantyhPW.Amechanismbasedunderstandingofcancerpain.Pain

2002;96:1–2.

MillesiH.Forty-twoyearsofperipheralnervesurgery.Microsurgery

1993;14:228–33.

NachemsonA,ZdeblickTA,O’BrienJP.Lumbardiscdiseasewith

discogenicpain.Whatsurgicaltreatmentismosteffective?Spine2004;21:1835–8.

NachemsonAL.Evaluationofresultsinlumbarspinesurgery.Acta

OrthopScandSuppl1993;251:130–3.

NachemsonA.Lumbarspineinstability:outcomeandrandomizedcon-trolledtrials.BullHospJtDis1996;55:166.

OlmarkerK,BlomquistJ,StrombergJ,NanmarkU,ThomsenP,Ry-devikB.Inflammatogenicpropertiesofnucleuspulposis.Spine1995;20:665–9.

OlmarkerK,LarssonK.Tumornecrosisfactoralphaandnucleus-pulposus-inducednerverootinjury.Spine1998;23:2538–44.

OlmarkerK,RydevikB.Newinformationconcerningpaincausedby

herniateddiskandsciatica.Exposuretodisktissuesensitizesthenerveroots.Lakartidningen1998;95:5618–22.

OlmarkerK,RydevikB.Selectiveinhibitionoftumornecrosisfactor-alphapreventsnucleuspulposus-inducedthrombusformation,intraneuraledema,andreductionofnerveconductionvelocity:possibleimplica-tionsforfuturepharmacologictreatmentstrategiesofsciatica.Spine2001;26:863–9.

OlmarkerK,MyersRR.Pathogenesisofsciaticpain:roleofherniatednucleuspulposusanddeformationofspinalnerverootandorsalrootganglion.Pain1998;78:99–105.

PearceJM.Aspectsofthefailedbacksyndrome:roleoflitigation.SpinalCord2000;38:63–70.

RandN,ReichertF,FlomanY,RotshenkerS.Murinenucleuspulposus-derivedcellssecreteinterleukins1-beta,-6,and-10andgranulocyte-macrophagecolony-stimulatingfactorincellculture.Spine1997;22:2598–601,discussion602.

RevelM,AuleleyGR,AlaouiS,NguyenM,DuruozT,Eck-MichaudS,etal.Forcefulepiduralinjectionsforthetreatmentoflumbosciaticpainwithpost-operativelumbarspinalfibrosis.RevRhumEnglEd1996;63:270–7.

WoessnerJrJF,BasheyRI,BoucekRJ.Collagendevelopmentinheartandskinofthechickembryo.BiochimBiophyActa1967;140:329–38.YakshTL.Pharmacologyofspinaladrenergicsystemswhichmod-ulatespinalnociceptiveprocessing.PharmacolBiochemBehav1985;22:845–58.

YakshTL,HuaXY,KalchevaI,Nozaki-TaguchiN,MarsalaM.Thespinalbiologyinhumansandanimalsofpainstatesgeneratedbypersistentsmallafferentinput.ProcNatlAcadSciUSA1999;96:7680–6.