Rev. Bras. Anestesiol. vol.65 número1

REVISTA

BRASILEIRA

DE

ANESTESIOLOGIA

www.sba.com.br

SCIENTIFIC

ARTICLE

Neurotoxic

effects

of

levobupivacaine

and

fentanyl

on

rat

spinal

cord

Yesim

Cokay

Abut

,

Asli

Zengin

Turkmen

,

Ahmet

Midi

,

Burak

Eren

,

Nese

Yener

_,

_Asiye

_Nurten

a_{DepartmentofAnesthesiology,KanuniSultanSuleymanEducationandTrainingHospital,Istanbul,Turkey} b_{DepartmentofPhysiology,FacultyofMedicine,YeniYuzyilUniversity,Istanbul,Turkey}

c_{DepartmentofPathology,FacultyofMedicine,MaltepeUniversity,Istanbul,Turkey}

d_{DepartmentofNeurosurgery,BakirkoySadiKonukEducationandTrainingHospital,Istanbul,Turkey}

Received20January2013;accepted15July2013 Availableonline26October2013

KEYWORDS

Levobupivacaine; Neurotoxicity; Fentanyl

Abstract

Background: Thepurposeofthestudywastocomparetheneurotoxiceffectsofintrathecally administeredlevobupivacaine,fentanylandtheirmixtureonratspinalcord.

Methods:In experiment,therewere fourgroups withmedicationandacontrolgroup.Rats wereinjected15␮Lsalineorfentanyl0.0005␮g/15␮L,levobupivacaine0.25%/15␮Land fen-tanyl0.0005␮g+levobupivacaine0.25%/15␮Lintrathecallyfor fourdays.Hotplatetestwas performedtoassessneurologicfunctionaftereachinjectionat5th,30thand60thmin.Five days afterlastlumbalinjection, spinal cordsectionsbetween theT5 andT6vertebral lev-elswereobtainedforhistologicanalysis.Ascorebasedonsubjectiveassessmentofnumber ofeosinophilicneurons---Redneuron---whichmeansirreversibleneuronaldegeneration.They reflecttheapproximatenumberofdegeneratingneuronspresentintheaffectedneuroanatomic areasasfollows:1,none;2,1---20%;3,21---40%;4,41---60%;and5,61---100%deadneurons.An overallneuropathologicscorewascalculatedforeachratbysummatingthepathologicscores forallspinalcordareasexamined.

Results:In the results of HPT, comparing the control group, analgesic latency statistically prolongedforallfourgroups.

Inneuropathologicinvestment,thefentanylandfentanyl+levobupivacainegroupshave sta-tisticallysignificanthighdegenerativeneuroncountsthancontrolandsalinegroups.

Conclusions: Theseresultssuggestthat,whenadministeredintrathecallyinrats,fentanyland levobupivacainebehavesimilarforanalgesicaction,butfentanylmaybeneurotoxicforspinal cord.Therewasnosignificantdegenerationwithlevobupivacaine,butfentanylgrouphashad significantdegeneration.

© 2013SociedadeBrasileirade Anestesiologia.Publishedby ElsevierEditoraLtda.Allrights reserved.

∗_{Correspondingauthor.}

E-mail:[email protected](Y.C.Abut).

PALAVRAS-CHAVE

Levobupivacaína; Neurotoxicidade; Fentanil

Efeitosneurotóxicosdelevobupivacaínaefentanilsobreamedulaespinhalderatos

Resumo

Justificativa:Oobjetivodesdeestudofoicompararosefeitosneurotóxicosdaadministrac¸ão porviaintratecaldelevobupivacaína efentanilesuasmisturassobreamedulaespinhalde ratos.

Métodos: Oexperimentocompreendeuquatrogruposquereceberammedicamentoeumgrupo controle.Osratosforamsubmetidosàinjec¸ãodesalina(15␮L)oufentanil(0,0005␮g/15mL), levobupivacaínaa0,25%(15␮L)efentanil(0,0005␮g+levobupivacainea0,25%/15␮L)porvia intratecaldurantequatrodias.Otestedeplacaquentefoiusadoparaavaliarafunc¸ão neurológ-icaapóscadainjec¸ãonosminutos5,30e60.Cincodiasapósaúltimainjec¸ãolombar,secc¸õesda medulaespinalentreosníveisvertebraisT5eT6foramobtidasparaanálisehistológica.Usamos umescorecombasenaavaliac¸ãosubjetivadonúmerodeneurônioseosinofílicos(neurônios ver-melhos),oquesignificadegenerac¸ãoneuronalirreversível.Essesneurôniosrefletemonúmero aproximadodeneurôniosem degenerac¸ãopresentesnasáreasneuroanatômicasafetadasda seguinteforma:1=nenhum;2=1-20%;3=21-40%;4=41-60%e5=61-100%neurôniosmortos. Umescoreneuropatológicoglobalfoicalculadoparacadaratopelasomadosescorespatológicos paratodasasáreasexaminadasdamedulaespinhal.

Resultados: Nosresultados doTPQ,comparandoogrupo controle,alatênciaanalgésica foi estatisticamenteprolongadaparatodososquatrogrupos.

Eminvestimentoneuropatológico,osgruposfentanilefentanyl+levobupivacaine apresen-taramdegenerac¸ãoneuronalemcontagenssignificativamentemaisaltasqueosgruposcontrole esalina.

Conclusões:Estesresultados sugeremquefentanilelevobupivacaína,quandoadministrados porviaintratecalemratos,secomportamdeformasemelhanteàac¸ãoanalgésica,mas fen-tanilpodeserneurotóxicoparaamedulaespinhal.Nãohouvedegenerac¸ãosignificativacom levobupivacaína,masogrupofentanilapresentoudegenerac¸ãosignificativa.

©2013SociedadeBrasileiradeAnestesiologia.PublicadoporElsevierEditoraLtda.Todosos direitosreservados.

Introduction

Increasing laboratory evidences1---5 _{suggest that all local}

anestheticsarepotentiallyneurotoxic,andthatneurologic impairment after neuraxial blockade may result from a directneurotoxiceffectofdrugs.Today,commercially avail-ablebupivacaine is a racemic mixture of S (−) and R(+) enantiomers.ItsisolatedS(−)enantiomerlevobupivacaine hasalowerpotentialfor producingtoxicityinthecentral nervoussystem andcardiovascularsystemthan doesR(+) bupivacaineinanimalsandhumans.6---8

Inclinicalsetting,themainpurposeofspinoaxial admin-istrationofopioidsistoreducethelocalanestheticdosage, tomaximize efficacy and tominimize side effects of the involveddrugswhoseactionsitesareinthecentralnervous system. Lipophilic opioids suchas fentanylare commonly administeredspinallyinadults.Thereisminimalpublished report specifically addressing the histologic, physiologic, or clinical evidence of neurotoxicity with spinal fentanyl administration.9,10

In the present study we investigate that the repeated bolusintrathecalinjectionoffentanyl,levobupivacaineand theirmixturecouldbeneurotoxic forspinal cordona rat model.

Methods

The protocol was approved by the Animal Research and UseCommitteeofIstanbulUniversity(25/02/2010Number:

26).AllexperimentswereperformedinDETAE(Department ofNeuroscience,InstituteofExperimentalMedicine, Istan-bul University, Istanbul). The experiment was conducted on Wistar albino rats 6---8 months old and (240---320g). Animals were divided into five groups of 8 animalseach. No drug was injected to control group (Group Control). After positioning prone and shaving, under aseptic con-ditions, without anaesthetizing, the following drugs were injected intrathecally once a day at the same hour, for four days, through the L4-5 intervertebal space: isotonic saline (Group Saline), fentanyl 50␮g/mL (Group Fen-tanyl),levobupivacaine2.5mg/mL(GroupLevobupivacaine) or fentanyl 50␮g/mL+levobupivacaine 2.5mg/mL (Group Fentanyl+Levobupivacaine).Solutionswerepreparedfrom fentanyl citrate(additive free) (Fentanyl-Janssen, Janssen --- Cilag, Belgium) and levobupivacaine hydrochloride (Chirocaine-Sigma Chemical, Steinheim, Germany). Solu-tions were diluted with sterile isotonic saline (Serum Physiologique 0.9%-Galen Deva-Kocaeli, Turkey) (Table 1). Allsolutionswerepreparedandinjectedatroom tempera-ture(20---24◦_{C).Becausechronicallyimplantedintrathecal} catheters can induce damage in tissue,11,12 _we

Table1 Densityof CSF,isotonic saline, levobupivacaine andfentanylat37␲C.

37◦_C

CSF 1.000646±0.000086

Saline 0.99951±0.00001

Levobupivacaine2.5mg/mL 0.99985±0.00002

Fentanyl 0.99333±0.00002

majorityofanimals,lickingthehindpawswasobserved;in

remainder, jumping the end point, cut off timewas 15s.

Behavioral tests were performed by a neuroscientist who

wasblindedthegroupassignment. Motorfunction (MF)of

theposteriorlimbswasassessedbybilaterallygrading the

motorblockas:0,none;1,partiallyblocked;and2,

com-pletelyblocked.13_{Motorblockadewasgradedasnonewhen}

therathadnovisiblelimbweaknessandnormalgait;as par-tiallyblockedwhenthelimbwasabletomovebutnotable tosupportthenormal;andascompletelyblockedwhenthe limb wasflaccid, withnodetectable resistanceto exten-sionofthelimbs.Theanimalswereexamined30minbefore andaftereachinjection.Ratshavinganyproblemwithtail movementormotordysfunctioninthehindlimbswerenot used in the experiments.The following parameters were measuredandrecordedovera2hperiod:sensoryblocked, as determined by the response to the hemostat pinch test.Aftereachinjection,theyweremaintainedona12-h light---darkscheduleandhousedwithfreeaccesstofoodand water.

Histological

evaluation

Afterthelastfunctionalexamination,theratswerekilled byintraperitonealhighdosages(100mg/kg)pentobarbital. Spinal cordwasexcised bya neurosurgeonblinded tothe group assignment and to the results of behavioral mea-surements. In order to discover cranial spread of local anesthetics,sectionswhichobtainedfromT5-6spinalcord wereusedforqualitativeevaluation.Spinalcordwasfixedin neutral(10%)bufferedformalinfor7days.Tissuesexposed toformalin,alcohol,Xyloidandparaffinwithtissuefollow machine (Thermo Shandon Exelsior ES) and embedded in paraffinby routinetechniques. The spinalcordwassliced at 2␮m intervals with the help of rotary type micro-tome (ThermoShandon Finesse325). Tissues stainedwith hematoxylinandeosin, andevaluatedby lightmicroscopy (OlympusCX31)byapathologistblindedtothegroup assign-mentand totheresultsof behavioralmeasurements. The primary neuropathologic alteration seen in the rats was oneofacuteeosinophilicneurondegeneration.14_Degreesof

neuropathologicalterationswithinagivenanatomicregion werescoredbasedonsubjectiveassessmentofnumberand distributionof eosinophilicneurons ---Red neuron---which meansirreversibleneuronaldegeneration.Theyreflectthe approximatenumberofdegeneratingneuronspresentinthe affectedneuroanatomicareasasfollows:1,none;2,1---20%; 3, 21---40%; 4, 41---60%; and 5, 61---100% dead neurons. An overall neuropathologic scorewascalculated for each rat bysummatingthepathologicscoresforallspinalcords,10 areasexaminedforeverypreparation.

Statistical

analysis

TheresultsoftheHPTwereevaluatedusingone-way anal-ysisofvariance(ANOVA),followedbyDunnet’stestforpost hocevaluationtocompareallgroupswithcontrolgroup.

Forexaminationoftolerance,2nd-3rdand4thdaysHPT latency values were compared with the first day results ofeachgroup,andANOVAfollowedbyDunnet’s testwere performed.MFwasnotanalyzedbecauseallanimalshave 0motorfunctionlevels.Degreesofneuropathologic alter-ationswithinspinalcordwereanalyzedwithKruskalWallis followed by Mann Whitney-U test. Significant difference testingwasp<0.05onANOVA.

Results

Allrats completed the experimentand included the data analysis. All animals recovered fully, were awake and actively mobile and eating and drinking normally after 30minofinjection.Duringtheexperimentmotorblockhas notbeen observedin anyoftherats.Noanimalshad sus-tainedvisibleinjuryorbleedinginthespinalcordwhenthe spinalcordexcisedattheendoftheexperiment.

HPT latencies prolonged for all groups comparing the controlgroupwithoutanymotorblock.

The result of HPT latency on the first, second, third andfourth day,at 5th,30thand60thmin, canbeseen in Table2.In5thmin,HPT latencies--- comparingtocontrol group--- wereprolonged statisticallysignificantly insaline (p<0.02), fentanyl and fentanyl+levobupivacaine groups (p<0.05).In30thmin,HPTlatencieswerestatistically sig-nificantlyincreased---comparingtocontrolgroup---insaline (p<0.02),fentanyl(p<0.01)andfentanyl+levobupivacaine (p<0.02). It was found that HPT values for levobupi-vacaine (p<0.01) and fentanyl+levobupivacaine (p<0.02) group were increased significantly different from control valuesin60thmin.

The result of HPT latency on the second day, at 5th, 30th and 60th min, have prolonged in all groups when compared to control group. This was found significant for saline (p<0.01) group and also fentanyl and fen-tanyl+levobupivacaine(p<0.05)groupsin5thmin.In30th min, there was no statistically significant difference in prolongedlatencies. The increment ofHPT values in fen-tanyl+levobupivacainegroupat 60thmin wassignificantly differentfromcontrolgroupinday2(p<0.01).

TheresultofHPTlatencyonthethirdday,at5th,30th and60thmin,inlevobupivacainegroupwassignificantly pro-longedin5thminwhencomparedtocontrolgroup(p<0.05). A significant increase in 60th min in fentanyl (p<0.05), levobupivacaine (p<0.05) and fentanyl+levobupivacaine (p<0.02)groupswasfound.

Table2 Theeffectofrepeateddrugtreatmentontolerancedevelopment.

Group n Hotplatelatency(s)

Day1 Day2

5min 30min 60min 5min 30min 60min

Control 8 1.3±0.3 1.3±0.2 1.3±0.2 1.1±0.1 1.3±0.2 1.4±0.3

Saline 8 5.3±1.3 3.9±0.3 2.9±0.6 3.9±0.7 2.1±0.3a _2.3±0.3

Fentanyl 8 4.9±0.7 4.4±0.9 1.8±0.3 3.1±0.6b _{2.5±0.5 2.1±0.4}

Levobupivacaine 8 3.9±1.0 3.3±0.5 3.5±0.8 2.8±0.5 2.4±0.5 2.4±0.5 Fentanyl+levobupivacaine 8 4.9±1.1 3.8±0.7 3.4±0.4 2.9±0.3 3.1±0.9 3.4±0.5

Group n Hotplatelatency(s)

Day3 Day4

5min 30min 60min 5min 30min 60min

Control 8 1.3±0.1 1.4±0.1 1.3±0.2 1.3±0.2 1.3±0.2 1.4±0.2

Saline 8 2.7±0.6 2.1±0.1a _{2.6±0.2 3.1±0.6 3.6±0.3 2.8±0.4}

Fentanyl 8 2.5±0.2c _2.5

±0.4 3.5±0.6d _2.6

±0.3c _3.0

±0.5 3.6±0.5e

Levobupivacaine 8 2.8±0.6 3.0±1.0 3.4±0.6 3.3±1.2 3.3±0.9 3.6±0.7 Fentanyl+levobupivacaine 8 1.8±0.3b _{3.0±0.5 3.8±0.8 3.8±0.8 2.9±0.4 3.4±0.4}

n:numberofanimals.

Valuesareexpressedintermsofmean±SE. a_{p<0.001comparedtoDay1,30thminvalue.} b _{p<0.02comparedtoDay1,5thminvalue.} c _{p<0.01comparedtoDay1,5thminvalue.} d _{p<0.05comparedtoDay1,60thminvalue.} e _{p<0.02comparedtoDay1,60thminvalue.}

InFig.1,itcouldbeseenthattherepeateddrug applica-tionsshowedanalgesic effects.Control groupshowedthat therewas nochange onanalgesic responsetime for four days.Insaline group, we observedstatistically significant shortanalgesic response timeonsecond and third day at 30th min values (p<0.001). In fentanyl group, analgesic responsetimewasfoundshorteronsecond(p<0.02),third (p<0.01) and fourth (p<0.01) day at 5th min and third (p<0.05)andfourth(p<0.02)dayat60thmin.In levobupi-vacainegroup,therewasnochange,onanalgesicresponse timeforfourdays.Infentanyl+levobupivacainegroup, anal-gesicresponsetimewasfoundshorteratthe5thminofthird day(p<0.02).

Spinal cordsneuropathologic analyses ofall groups are shownin Fig.2. Inneuropathologic evaluation, degenera-tiveneuronscoring wasincreasedstatisticallysignificantly infentanylgroupincomparisontocontrolandsalinegroups (p<0.05). Fentanyl+levobupivacaine group showed high degenerative neuron scores than control, saline and lev-obupivacainegroups(p<0.01).

Discussion

SinceBier and Hildebrantinitially performed spinal anes-thesia with cocaine in 1898, the history of spinal local anesthetic usage in humans has been followed by widespreadapplication withlittleornocontrolledtesting forneurotoxicity.15_{In1985,Readyetal.evaluatedthe}

neu-rotoxic effects of single injections of local anesthetics in rabbits.Theyreportedthat histopathologicalchangesand

neurologic deficitsoccurred withhigher concentrations of tetracaine (1%) and lidocaine (8%).16,17 _{In the past, local}

anestheticsolutionsinclinicallyadministereddoses,rarely inducedneurologicinjury,andtheobservationofneurotoxic effectswouldrequirelargerdosageof drugs.To produced injury,aratmodelinwhichlocalanesthetics were contin-uouslyinfusedhas beendesigned byDrasneretal.In this study,withoutclinicallyrelevant,thefunctionalimpairment and morphologic damage were observed.18 _Previous

stud-iesdesignedbyKofkeindicatedthatwhentheyweregiven systemically, opioidscanproduce limbicsystem hyperme-tabolismandbraindamageinquitelargedoses.19---21_In2000,

thedensityandbaricityofthemixturesusedinspinalblocks weredeterminedforthefirsttimeinBrazil.22

Abovethelight ofthis studies,we wantedtoobserve, someintrathecaldrugsiftheychronicallypenetratespinal cord,couldbeneurotoxicornot.Inourstudy,theresultsof theHPThaveindicatedthat,therewerenomotorblockbut significantantinociceptive-analgesiceffectsinallgroups.

In5thmin,HPTlatencies---comparingtocontrolgroup ---wereprolongedstatisticallysignificantlyinsaline(p<0.02) group. Similarly after repeated injections, its analgesic responsetimewasdecreasedlikeotheropioidanalgesicsin salinegroup.Wecouldnotexplainthatwhysalinebehave likeananalgesicsolution.

Afterrepeatedapplications,fentanyl groupwas devel-oped tolerance to theanalgesic effectbut this tolerance wasnotdevelopedwithlevobupivacainegroup.

0 1 2 3 4 5 6 7

Control Saline Fenta

Levo

Fenta+Levo

Control Saline Fenta

Levo

Fenta+Levo

A

B

C

D

0 1 2 3 4 5 6 7

Control Saline Fenta

Levo

Fenta+Levo

Control Saline Fenta

Levo

Fenta+Levo

5 min 30 min 60 min a

a

a

a

a aa

aaa

bb bb

bbb

b bb

cc ccc

ccc

c _c cc

c ccc ccc

Figure1 Theeffectsofdrugapplicationsonthe1st(A),2nd(B),3rd(C)and4th(D)dayresultsofHPTlatency. Alldataarepresentedasmean±SE.

a_p<0.05,aa_p<0.02,aaa_{p<0.01accordingto5thminvaluetocontrolgroup.} b_p<0.05,bb_p<0.02,bbb_{p<0.01accordingto30thminvaluetocontrolgroup.} c_p<0.05,cc_p<0.02,ccc_{p<0.01accordingto60thminvaluetocontrolgroup.}

thepatientposture,compositionof solution,typeof nee-dle,levelandspeedofinjection,volume,viscosity,inferior venacava obstruction andpregnancy.23---27 _{Butthe baricity}

andtemperature oflocalanesthetics arethemost impor-tantfactorsofthedistributionofthelocalanestheticinto subarachnoidspace.28---30

ThedensityofthehumanCSFisnotuniform,anditcan varywithage,gender,pregnancy,andseveraldiseases.The relationshipbetweenthedensityofthelocalanestheticand theCSF,knownasbaricity.

Temperature changes also, affect the distribution of localanesthetics.When thelocalanesthetics areinjected into subarachnoid space(generally in room temperature

20---24◦_{C)thetemperatureofthelocalanestheticreachesan} equilibriumwiththebodytemperature(37◦_{C)veryquickly,} beforebeingfixedatthenerverootsandat37◦_{C,allisobaric} anestheticsbecomehypobaricsolutions.

Adjuvants arefrequently addedtolocal anesthetics to improve anesthesia and prolong postoperative analgesia. Opioids (morphine, fentanyl, and sufentanil) and cloni-dine showed to be hypobaric at 37◦_{C and, when added} tolocal anesthetics, they reduce the density of the new solution, making it more hypobaric, according to some studies,31,32_{butitdoesnotseemtohaveanyeffecton}

0 1 2 3

Fenta+Levo Levo

Fenta Saline

Control

Neuropathologic score

*

**

Figure2 Theeffectsofrepeateddrugapplicationsonspinal cord.

Dataarepresentedasmean±SE.

*p<0.05comparedwithcontrolandsalinegroup.

**p<0.01compared with control, saline and levobupivacaine group.

Figure3 Arrowshowseosinophilicneuronofratspinalcord infentanylgroup.

In ourstudy,oursolutions’ temperature (inroom tem-perature 20---24◦_{C) and their opioid mixture can affect} therostral spread of drugs. But this mechanism does not explain the excessive eosinofilic neuron count, especially somegroups.

In clinical practice, there are many studies that local anestheticsolutionsweredilutedwithisotonicsterilesaline as we did in our study.36---39 _{Rostral spread of drugs can}

beexplainedwiththe hypobaricityof oursolutions, how-ever, our neuropathologic results --- against the results of Fukushima---showedthatwhenthedrugswereusedevenin verylowanalgesicdoses(withoutmotorblock)theyhad per-manentneurotoxiceffectsinthoracicspinalcord.40_Because

itshows the non-recoverable neurondamage,the eosino-phylicneuroncounthasbeenused---asneurotoxicitysignsin ourstudy,insteadofnon-specificfindingssuchas vacuoliza-tion,edemaandinvasionofmacrophage,picnoticnuclei.41

Wecan see eosinophylic-degenerativeneuron in Fig.3.It hadbeendeterminedbyKofkebefore.14

In neuropathological investment degenerative neuron scoreinspinalcordwasfoundsignificantlyhighinfentanyl

andfentanyl+levobupivacainegroupthancontrolgroupand salinegroup.Alsofentanyl+levobupivacainegrouphas sig-nificantlyhigherscorethanlevobupivacainegroup.

Indeed,oneofthemajoraimsofthisstudywasto deter-mine the neuropathologic changes on spinal cord, after it is chronicallyexposed to intrathecal drugs. Chronically implanted intrathecal catheters characteristically induce damageincontrolanimals,sowepreferrepeatedinjection techniqueinsteadofintrathecalcatheter.11,12

Ourdataconfirmthatfentanylandlevobupivacainecan causespinalcorddamageinratswhentheywereinjected for4days,eveninanalgesicdoses.

At the end, our study tried to explain different sides of neurotoxicity. It is an animal study and has behavioral component.Neuropathologicmethodologyisdifferentand may be more objective than previous studies, and also our study is based on the in vitro CSF-local anesthetic distribution-baricitystudies.Newstudiesshouldbeplanned with electron microscopyor behavioral studies may show that the long term penetration of local anesthetics may cause neuropathological degenerations on spinal cord, in future.

Conflicts

of

interest

Theauthorsdeclarenoconflictsofinterest.

References

1.SakuraS,ChanVW,CirialesR,etal.Theadditionof7.5%glucose doesnotaltertheneurotoxicityof5,5lidocaineadministered intrathecallyintherat.Anesthesiology.1995;82:236---40. 2.SakuraS,BollenAW,CirialesR,etal.Localanesthetic

neuro-toxicitydoesnotresultfromblockadeofvoltage-gatedsodium channels.AnesthAnalg.1995;81:338---46.

3.HashimotoK,SakuraS,BollenAW,etal.Comparativetoxicity ofglucoseandlidocaineadministeredintrathecallyintherat. RegAnesthPainMed.1998;23:444---50.

4.SakuraS,KiriharaY,MugurumaT,etal.Thecomparative neuro-toxicityofintrathecallidocaineandbupivacaineinrats.Anesth Analg.2005;101:541---7.

5.YamashitaA,MatsumotoM,MatsumotoS,etal.Comparisonof theneurotoxiceffects onthespinalcordoftetracaine, lido-caine,bupivacaine,andropivacaineadministeredintrathecally inrabbits.AnesthAnalg.2003;97:512---9.

6.AbergG. Toxicologicaland localanaesthetic effects of opti-callyactiveisomersoftwolocalanaestheticcompounds.Acta Pharmacol.1972;31:273---86.

7.McLeod GA, Burke D. Levobupivacaine anaesthesia. 2001;56:331---41.

8.GautierP,deKockM,Hubertyl,etal.Comparisonoftheeffects ofintrathecalropivacaine, levobupivacaine,and bupivacaine forcaesareansection.BrJAnaesth.2003;91:684---9.

9.AllenJW,HoraisKA,TozierNA,etal.Opiatepharmacologyof intrathecalgranulomas.Anesthesiology.2006;105:590---8. 10.BaharM,CohenML,GrinshpoonY,etal.Aninvestigationofthe

possibleneurotoxiceffectsofintrathecalmidazolamcombine withfentanylintherat.EJA.1998;15:695---701.

11.Bahar M, Rosen M, Vickers MD. Chronic cannulation of the intradural or extradural space in the rat. Br J Anaesth. 1984;56:405---10.

13.KanekoM,SaitoY, KiriharaY, etal. Synergistic antinocicep-tiveinteraction afterepidural coadministration of morphine andlidocaineinrats.Anesthesiology.1994;80:137---50. 14.KofkeWA, Garman RH, Garman R, et al. Opiod

neurotoxic-ity:fentanyl-inducedexacerbationofcerebralischemiainrats. BrainRes.1999;818:326---34.

15.BierA.VersucheuberCocainisirungdesRuckenmarkes.DtschZ Chir.1899;5151:361.

16.Ready LB, Plumer MH, Haschke RH, et al. Neurotoxicity of intrathecal local anesthetics in rabbits. Anesthesiology. 1985;63:364---70.

17.Hodgson PS, Neal JM, Pollock JE, et al. The neurotox-icity of drugs given intrathecally (Spinal). Anesth Analg. 1999;88:797---809.

18.DrasnerK,SakuraS,ChanVWS.Persistentsacralsensorydeficit inducedbyintrathecallocalanestheticinfusionintherat. Anes-thesiology.1994;80:847---52.

19.KofkeWA,GarmanRH,TomWC,etal.Alfentanil-induced hyper-metabolism,seizure,and histopathologyinratbrain.Anesth Analg.1992;75:953---64.

20.Kofke WA, Garman RH, Janosky J, et al. Opioid neurotox-icity: neuropathologic effects in rats of different fentanyl congeners and the effects of hexamethonium-induced nor-motension.AnesthAnalg.1996;83:141---6.

21.KofkeWA,GarmanRH,StillerRL,etal.Fentanyldose-response relationinrats.AnesthAnalg.1996;83:1298---306.

22.CangianiLM.Determinac¸ão dadensidadeedabaricidadedas misturas para anestesia subaracnóidea. Rev Bras Anestesiol. 2000;50:92---4.

23.GreeneNM.Distributionoflocalanestheticsolutionswithinthe subarachnoidspace.AnesthAnalg.1985;64:715---30.

24.Stienstra R, Greene NM. Factors affecting the subarachnoid spreadoflocalanestheticsolutions.RegAnesth.1991;16:1---6. 25.Carpenter RL, Hogan QH, Liu SS, et al. Lumbosacral

cere-brospinalfluidvolumeis theprimarydeterminant ofsensory blockextentanddurationduringspinalanesthesia. Anesthesi-ology.1998;89:24---9.

26.ConnollyC,McLeodGA,WildsmithJA.Spinalanaesthesiafor Caesareansectionwithbupivacaine5mgml±1inglucose8or 80mgml±1.BrJAnaesth.2001;86:805---7.

27.McLeod GA.Density of spinal anaesthetic solutions of bupi-vacaine, levobupivacaine,and ropivacainewith and without dextrose.BrJAnaesth.2004;92:547---51.

28.LuiAC,PolisTZ,CicuttiNJ.Densitiesofcerebrospinalfluidand spinalanaestheticsolutions insurgicalpatientsatbody tem-perature.CanJAnaesth.1998;45:297---303.

29.HorlockerTT,WedelDJ.Density,specificgravity,andbaricityof spinalanestheticsolutionsatbodytemperature.AnesthAnalg. 1993;76:1015---8.

30.StienstraR,GielenM,KroonJW,etal.Theinfluenceof tem-peratureandspeedofinjectiononthedistributionofasolution containingbupivacaineand methylenebluein aspinal canal model.RegAnesth.1990;15:6---11.

31.ParlowJL,MoneyP,ChanPS,etal.Additionofopioidsaltersthe densityandspreadofintrathecallocalanesthetics?Aninvitro study.CanJAnaesth.1999;46:66---70.

32.Hare GM, Ngan JC. Density determination of local anaes-theticopioidmixturesfor spinalanaesthesia.CanJAnaesth. 1998;45:341---6.

33.PattersonL,AveryN,Chan P,etal.Theadditionoffentanyl doesnotaltertheextentofspreadofintrathecalisobaric bupi-vacaineinclinicalpractice.CanJAnaesth.2001;48:768---72. 34.Imbelloni LE, Moreira AD, Gaspar FC, et al. Assessment

of the densities of local anesthetics and their combination withadjuvants. Anexperimental study.Rev BrasAnestesiol. 2009;59:154---65.

35.Nicol ME, Holdcroft A. Density of intrathecal agents. Br J Anaesth.1992;68:60---3.

36.FaustA,FournierR,VanGesselE,etal.Isobaricversus hypo-baricspinalbupivacainefortotalhiparthroplastyinthelateral position.AnesthAnalg.2003;97:589---94.

37.HellerAR,ZimmermannK,SeeleK,etal.Modifyingthebaricity oflocalanestheticsforspinalanesthesiabytemperature adjust-mentmodelcalculations.Anesthesiology.2006;105:346---53. 38.SrivastavaU,KumarA,SaxenaS,etal.Spinaanaesthesiawith

lignocaneandfentanyl.IndianJAnaesth.2004;48:121---3. 39.Ben-DavidB,Solomon E,LevinH, etal.Intrathecal fentanyl

withsmalldosedilutebupivacaine:betteranesthesiawithout prolongingrecovery.AnesthAnalg.1997;85:560---5.

40.FukushimaS, TakenamiT,YagishitaS, etal.Neurotoxicityof intrathecallyadministeredfentanylinaratspinalmodel.Pain Med.2011;12:717---25.