Post-stroke hemiplegia rehabilitation: Evolution of the concepts

HAL Id: inserm-03223415

https://www.hal.inserm.fr/inserm-03223415

Submitted on 10 May 2021

HAL is a multi-disciplinary open access

archive for the deposit and dissemination of

sci-entific research documents, whether they are

pub-lished or not. The documents may come from

teaching and research institutions in France or

abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, est

destinée au dépôt et à la diffusion de documents

scientifiques de niveau recherche, publiés ou non,

émanant des établissements d’enseignement et de

recherche français ou étrangers, des laboratoires

publics ou privés.

Post-stroke hemiplegia rehabilitation: Evolution of the

concepts

P. Marque, D. Gasq, E. Castel-Lacanal, X. de Boissezon, I. Loubinoux

To cite this version:

P. Marque, D. Gasq, E. Castel-Lacanal, X. de Boissezon, I. Loubinoux. Post-stroke hemiplegia

re-habilitation: Evolution of the concepts. Annals of Physical and Rehabilitation Medicine, Elsevier

Masson, 2014, 57 (8), pp.520-529. �10.1016/j.rehab.2014.08.004�. �inserm-03223415�

Literature

review

/

Revue

de

la

litte´rature

Post-stroke

hemiplegia

rehabilitation:

Evolution

of

the

concepts

Re´e´ducation

de

l’he´miple´gie

vasculaire

:

e´volution

des

concepts

P.

Marque

*

,

D.

Gasq

,

E.

Castel-Lacanal

,

X.

De

Boissezon

,

I.

Loubinoux

a_{Servicedeme´decinephysiqueetre´adaptation,CHURangueil,1,avenuePoulhes,31059Toulousecedex9,France} b_{Unite´ 825Inserm,CHUPurpan,1,placeBaylac,31059Toulousecedex9,France}

Received6August2014;accepted6August2014

Abstract

Strokerehabilitationhasundergonearevolutionoverthelastthreedecades.Cohortstudieshaveconsistentlyreinforcedtheimportanceof post-strokerehabilitationtostimulaterecovery,buttheconceptsofempiricalmethodsoriginallyproposedbytherapiststorehabilitatethesepatients havenotwithstoodclinicalanalysis.Functionalneuroimagingandanimalmodelshaveunveiledthemechanismsunderlyingfunctionalrecovery and helpedteams understand itslimitationsand improvement modalities.These neuroscience discoveries constitute thegroundsneeded to understandtheemergenceofnewtechnologies:roboticsandvirtualreality.Theobjectiveofthisreviewoftheliteraturewastoselectkeyworksin thisfieldtobetterunderstandcurrenttherapeuticpossibilities.

# _{2014PublishedbyElsevierMassonSAS.}

Keywords:Stroke;Rehabilitation;Neuroimaging;Hemiplegic;Plasticity

Re´sume´

Lare´adaptationdespatientsvictimesd’AVCaconnuuneve´ritablere´volutionaucoursdes3dernie`resde´cennies.Lese´tudesdecohortesont toujourssouligne´ l’inte´reˆtdelare´adaptationapre`sunAVCpourstimulerlare´cupe´ration,maislesconceptsdesme´thodesempiriquesinitialement propose´esparlesthe´rapeutespourre´e´duquercespatientsn’ontpasre´siste´ a` l’analysecliniquecritique.Paralle`lement,lestravauxdeneuroimagerie fonctionnelleetlesmode`lesanimauxontpermistoutd’abordd’objectiverlesme´canismesdelare´cupe´rationfonctionnelle,puisd’encomprendre leslimitesainsiquelesmodalite´sd’ame´lioration.Cesde´couvertesdesneurosciencessontlesbasessanslesquellesonnepeutpascomprendre l’apparitiondesnouvellestechnologies:robotiqueetre´alite´ virtuelle.Cetterevuedelalitte´ratureseproposedefaireunese´lectiondecestravaux cle´squipermettentdemieuxcomprendrelespropositionsactuelles.

# _{2014Publie´ parElsevierMassonSAS.}

Motscle´s: AVC;Re´e´ducation;Imageriefonctionnelle;He´miple´gique;Plasticite´

1. Englishversion 1.1. Introduction

Recovery is aspontaneous phenomenon observed evenin theabsenceofrehabilitation,andhasbeenknowntophysicians

for a long time. In the 19th century, Broca defined vascular syndromeastheonsetofasuddenfocaldeficitimprovingover time,thusenclosinginthedefinitionofstrokeitselfthenotion ofrecovery [1].

Intheeighties,cohortstudiesvalidatedtheimprovementof impairments and autonomy inthe weeks following astroke. The Copenhagen study [2,3] published in the Annals of PhysicalMedicineandRehabilitation(AMPR)intheninetiesis oneofthesestudies.Becauseofitsprospectiveandexhaustive natureintherelativelyhomogenousDanishhealthcaresystem, especiallyintherehabilitationfield,itbecameaflagshipstudy. Thisclinicalstudyevaluatedmotorimpairmentsandautonomy Availableonlineat

ScienceDirect

www.sciencedirect.com

AnnalsofPhysicalandRehabilitationMedicine57(2014)520–529

*Correspondingauthor.Servicedeme´decinephysiqueetre´adaptation,CHU Rangueil,1,avenuePoulhes,31059Toulousecedex9,France.

E-mailaddress:marque.ph@chu-toulouse.fr,philippe.marque@gmail.com (P.Marque).

http://dx.doi.org/10.1016/j.rehab.2014.08.004

1877-0657 #2014PublishedbyElsevierMassonSAS.

Open access under CC BY-NC-ND license.

Open access under CC BY-NC-ND license.

post-stroke,fromthe admissionintotheneurologyunit upto thedischargefromthePM&Rcenter.Resultsfromthisstudy validatedtheexistenceofmotorrecoveryaffecting,contrarily tothenotionstooreadilywidespreadlateron,allcategoriesof hemiplegicpatients, including the most severeones: 40% of patientswithseveremotorimpairmentsduringtheacutephase endupwithmildor moderateimpairmentsaftertherecovery process[3].

Thesecondrelevanceofthisstudywasitsvaluableanalysis of the recovery time course: for motor impairments, most improvementsoccurduringthefirst15weeks,regardlessofthe severityoftheinitialmotordeficit[2].Thiswasaveryoriginal resultatatimewhenitwasthenormtospeakofrecoveryover the span of several years. Furthermore, this rather quick timelinesuggests,byanalogywiththechild’sdevelopment,the notion, still debated today, of critical period. Several other studies validated this motor recovery kinetic and it is now commonlyusedtodefineresidualmotorimpairmentsasstable andchronicaftersix monthspost-stroke[4].

1.2. Rehabilitationimprovesmotorrecovery

Inparallel,sincetheeightiesthenumberofpublicationson theimpactofrehabilitationorphysiotherapyonthisrecovery processhascontinuedtoincrease.Oneofthefirstsystematic reviews of the literature on this topic was published by OttenbacherandJannellin1993[5].Sincethen,several meta-analyseshave beendevotedtothistheme [4–17].Along the years, theconclusions of these reviews have beenconstant. Firstofallthesedifferentstudiescontinuouslybacked-upthe evidenceoftheeffectivenessofrehabilitationonpost-stroke motor recovery: today 467 controlled studies on this topic havebeenidentifiedamountingto25,373patients[12]. Reha-bilitation care can improve the recovery of lower limb

[9,16,17] and upper limb [6,9] impairments, gait [13] and balance [16,17], as well as autonomy in activities of daily living[4,5,11,14].

Forthepast30years,variouspublicationshavereportedthe samefactorspromotingeffectivepost-strokerehabilitation.

Earlymanagementisafactoraffectingtheeffectivenessof post-stroke rehabilitation [4,10,18,19]. The issue of harmful effectsofexerciseduringtheinitialphaseintheintensivecare unitremains,eveniftheresultsoftworecentclinicalstudiesare ratherreassuring[20,21].However,beyondthatfirstweek,all studiesagreethatrehabilitationcareshouldbeinitiated.

Rehabilitation care intensity is also a factor promoting effective stroke rehabilitation [8,9]. In this case again, the optimaldosehasnotbeenclearlydetermined,but30minutesof dailyphysiotherapyhavebeenvalidatedasmoreeffectivethan 30minutesthreetimesaweekandthisdurationseemstobethe minimumamountrecommendedtoobservepositiveeffectson gait and autonomy [5,8,9]. In the framework of constraint-induced movement therapy (CIMT), the daily duration of rehabilitationtherapywassometimesincreaseduptosixhours perday[4,6].

Finally, the medical team effect is an element frequently encountered[4,11,18].Rehabilitationcareinmultidisciplinary

unitsspecializedinneuro-rehabilitationseemsmoreeffective thantheoneproposedbyanon-specializedornon-coordinated rehabilitation unit.

However, these studies do not highlight a stroke-specific physiotherapy technique. This was an essential result inthe context of the nineties. In fact at the time, post-stroke physiotherapywas madeupof different empiricaltechniques often identified by the names of their creator: Bobath, Brunstrom,Vojda,Kabat,Perfetti...Sixteencontrolledstudies have comparedthesedifferent techniquesespecially theones mostoftenused atthetime:BobathandBrunstrom[5,7,17]. These comparisons might seem dated today, yet on a fundamentalleveltheyremainrelativelyinterestingandatthe time, thesetwotechniquesdid seemdramatically opposed.

The first one, the Bobath method, proposes techniques basedonspasticityandtheinhibitionofautomaticreactions, and recommends avoiding any type of strength training exercises, which are suspected of increasing these spastic phenomena considered archaic and noxious for motor recovery. The second one, the Brunstrom method, uses techniques based on repetitive training of the automatic reflexes,includingresistancetraining,whicharebothstrictly prohibited in theBobath method.However, nocomparative studyhasevidencedthesuperiorityofoneofthesetechniques overtheother [7,17].

Atthesametime,severalstudiesreportedconclusionsthat were not in favor of the concepts underlying the Bobath method. Thilmann et al. [22] highlighted that in spastic subjects,thegaininstretchreflexeswasmorealteredthanthe onset threshold of stretch reflexes. The authors did not find stretch reflexes below 358/s., where under that threshold difficultiesweremorerelatedtotheparalysisthanthespasticity itself. Furthermore, additional studies showed that physical exercise, especiallystrength training, which isbanned inthe Bobathmethod,didnotincreasespasticityinachronicmanner

[23,24].

Infact,physiotherapyappearedtobeaneffectivetreatment to promote motor recovery post-stroke, yet the precise modalities of an effective therapy were not properly defined inthenineties.However,atthetime,itseemedquiteclearthat these modalitiesdidnot relyon acquiringspecific,empirical techniques difficulttomasterformostphysiotherapists.

1.3. Uncoveringbrainplasticityafter astroke

Brainplasticityisdefinedasallthemechanismsallowingthe braintoadaptitsownfunctiontoanewsituation.Theevidence oflastingclinicalimprovementspost-strokedidinfactbrought tolightthis typeof adaptation. In theeighties already,some authors [25] after having drawn conclusions from animal studiesindissimilarcontextsexplainedpost-strokefunctional recovery by mechanisms that unveiled redundant neuronal networks, alterations in synaptic effectiveness or dendritic sprouting phenomena.

The appearanceof brainfunctionalneuroimagingvalidated thesemechanismsofmotorrecovery.Theseimagingtechniques

canmeasurethelocalbloodflow,eitherbyusingaradiotraceras in PET scans, or using the paramagnetic properties of oxyhemoglobin in fMRI. These methods assume that in the brainthisbloodflowisdirectlycorrelatedtoneuronalactivation. Sincethedifferencesinbloodflowbetweentwotasksarequite modest,theyneedtaskrepetitionandastatisticalanalysisofthe differences observed to evidence these differences. In fact, resultsarepresentedasamappingofthestatisticaldifferences obtained between two situations: movement and rest for example.PET study of brain activation in casecontrols who recoveredina satisfactorymanner froma subcorticallacunar infarct (LACI) showed a reorganization of the somatotopic arrangementsoftheprimarymotorareasforthesepatientswhen theyperformedoppositemovementofthethumbandtheother fingers.Chollet et al. [26]found an activation affecting both hemisphereswhenmovingthe paretichand.However, Weiller etal.[27]didnotobserveinsimilarpatientsthesameactivation, butratheradisplacementofthehand-relatedcorticalactivation towardsthefacialmotorareas.Beyondtheirinterpretationwhich remainsquitecomplicated,theseresultsvalidateforthefirsttime theexistenceofbrainplasticityinstrokepatients.

Brainimaginghasvalidatedsomemechanismsinvolvedin thesebrainplasticityphenomena.Diaschisis[28]wasthefirst of these mechanisms to be uncovered. The definition of diaschisiscorrespondstoadepressionofbloodflowincertain areas of the brain not directly injured but functionally connected to the necrotic area. In other words, part of the impairment-related symptoms observed in patients with neurovascular lesions is correlated to the phenomenon of deafferentationofbrainareasthatarenotimpairedbutwhich receiveprojections from the injuredarea. Crossed cerebellar diaschisis(CCD)wasthefirstmechanismtobehighlightedand couldexplainthehighlyataxicnatureofsometypesofmotor impairmentsobservedafterlesionsofthemiddlecerebralartery (MCA) territory. The second mechanism evidenced was the discovery of a pre-existing, yet inactive, network able to substituteitselftothedeficientfunction,Whenthisnetworkis locatedinthesamefunctionalareatheterm‘‘redundant’’[29]

iscommonly used, when thissubstitution relies onnetworks locatedfartherawayinareasnormallyusedbyotherfunctions, theterm ‘‘vicarious’’ispreferably used [30].In all cases,in orderfor allthese networkstobecomefunctional,long term potentiationor long termdepression (LTP) or LTD synaptic changesarerequired.

1.4. Brainplasticity: isitalwaysapositivephenomenon? Nevertheless, more than a decade was necessary to understand the real meaning of these changes. First of all, Loubinouxet al. [31,32] described the time course of these changes in patients who recovered in a satisfactory manner. Theyshowed thatthe interpretation of these results was not univocal. In fact, 15 days post-stroke the increase in the activationsobserved were ratherevidenced on the ipsilateral sideofthesolicitedhandwithadecreasedcortexactivationin theinjuredside.Theseimagesappearedtobequitesimilarto theonesdescribedbyChollet10yearsearlier[26].Fourmonths

post-stroke, we canobserve alateral reorganizationof these activationsonthecontralateralside[31,32].Theseactivations remainsignificantlygreaterthaninthecontrolgroup.Oneyear later,theauthorsdidnotfindanydifferencescomparedtothe controlgroup[31,32].

Inparallel,Muraseetal.in2004[33]reportedthenoxious role of interhemispheric inhibition post-stroke. This author showed that in the 20 milliseconds before a unilateral hand movement,therewasamarkedreductionininterhemispheric inhibitionmechanismsnormallyobservedinhealthysubjectsat rest[34].Thisdecreaseisalwayslimitedandinallcasesless importantonthehemiplegicsidevs.thehealthysideforstroke patientsbutalsocomparedtoapopulationofhealthysubjects. Murase et al. developed the concept of maladaptive brain plasticity[33].Thelesiontriggersadecreasedinterhemispheric inhibitionfromtheinjuredhemispheretowardsthehealthyone. Thisreductionleadstoanoveractivationoftheipsilateralmotor networks,whichinturntriggeran increasedinhibition ofthe injuredhemisphere.Thisimbalance mechanism affectingthe interhemisphericinhibitionequilibriummight,atleastinitially, worsentheimpairmentsandexplaintheactivationsobservedby Loubinouxetal.15days post-stroke[31,32].Brainplasticity could represent a negative factor in some circumstances (Fig.1).

1.5. Brainplasticityislimited

Certainly, neuroimaging studies might present several sources of bias: small cohort, over-representation of sub-corticallesionsandmoderateimpairments.However,theywere abletoproposeahierarchicalrankingintherecoveryprocess and set the limits of spontaneous brain plasticity. When the supplementingnetworkisanatomicallyandfunctionallyclose tothe initialinjured network (redundancy)the deficitis less severe and the recovery will be better [35–41]. Conversely, whenthe compensationnetworkislocatedatadistancefrom theinitialnetworktheimpairmentwillbemoresevereandthe recoverywillbepartialorofpoorquality[26,42–44].Inadults with hemiplegia, the use of a compensation network contralateraltothelesionandthusipsilateraltothemovements oftheparetichand,translatesintoapoorqualityrecoveryand thepersistenceofsevereimpairments.Tomodelthisprinciple, some authors proposed a lateralization index [45–47]. This index calculates the ratio between the volumes of activated pixelsinthemotorareascontralateraltothemovementandfor alltheipsilateralorcontralateralmotorareas.Thisindexvaries between 0 and1. An index score closeto 1 supportsproper brainplasticityandpointstoagoodrecovery.Thisindexisnow used in neuroimaging studiesas an intermediate criterion to evaluatetheimpactofsomerehabilitationorpharmacological interventionsonrecovery[48,49].

1.6. Animalmodels inrehabilitation studies

Itisalsointheeightiesthatthefirstanimalmodelsappeared. The first of them was the enriched environment model and concernedtherat[50].Thismodelwasinitiallyusedtostudy

P.Marqueetal./AnnalsofPhysicalandRehabilitationMedicine57(2014)520–529 522

developmental brain plasticity innewborns. This modelwas adaptedtopost-lesionsplasticity.Theenrichedenvironmentis representedbyalargercage, withadiversityof architectural settings and accessories, with several animals in it. This environmentstimulatesthe relationshipsbetweenthe animals andimplicitlyofferstotheratsagreaterrangeofmotoractivity thaninanindividualcage.Ratsplacedafteravascularlesionof the MCA,in cages withan enriched environment presented, afterafewweeks,abetterqualitymotorrecoveryontheclinical tests than rats in individual cages [50]. Furthermore, on a histological level, the density of dendrites per neurons was greaterinratsplacedintheenrichedenvironmentunderlininga moreimportantsprouting[50].

Similarly,arehabilitationmodelwasdevelopedinmonkeys. Preciseandreproducible brainmappingispossiblethanks to corticalmicro-stimulationtechniques.Nudoetal.studiedthe reorganizationofbraintopographyafteravascularlesioninthe handterritoryinadultsquirrelmonkeys[51].Hefirstnotedthat inthehoursfollowingthelesion,areasperipheraltothelesion appeared and responded by contractions of the wrist and fingers.Theseresultsunderlinearapidplasticity,probablyby theunveilingofinactivenetworks.Butafterafewweeks,Nudo et al. observed a disappearance of these stimulation sites answering to the wrist and fingers in favor of other areas answering to the elbow and shoulder [51]. Conversely, in monkeyswherethehealthyupperlimbwasimmobilized,thus makingit mandatoryto usetheimpairedlimb,the numberof areas and the cortical surface, when stimulated, led to an increased response in the hand, wrist and finger muscles

[52,53].Thisstudyunderlinedause-dependentplasticityaswell as competition phenomena between cortical representations. Similar use-dependent cortical representation competition

phenomenaordependentorsensoryafferentswerehighlighted in humans: inupper-limb amputees [54]or rightafter having placedatourniquet[55].Acatchphrasewasthenproposedasa rehabilitationmotto:‘‘useitorloseit’’.

1.7. Sensory stimulation,constraint-inducedtherapy

Constraint-inducedtherapyofthehealthyupperlimbisalso atechniquethathasshoweditseffectivenessinhumans,ona clinicallevelinstudieswithagoodmethodology[56],aswell as on intermediate criteria in functional neuroimaging trials

[40]or TMScortical mappingstudies[57].

Yet constraint-therapy is a technique only concerning a smallportionofpareticpatients.Itisnotmeantforpatientswho are too severely impaired or during the first stages of rehabilitation training.

Passive mobilization is involved in the effectiveness of rehabilitationduringtheinitialstagesandinpatientswhoare completely paralyzed. In fact, PET [58] or EEG/fMRI [59]

neuroimagingstudiesshowedthatpassivemobilizationallowed the activation of motor networks. Furthermore, passive rehabilitation trainingis abletodurablyalterbrainactivation asevidencedbyfMRIinhealthy[60]andhemiplegicsubjects

[61]alike.Theseresultsareinaccordancewithclinicalstudies that showed the positive effect of transcutaneous electrical nerve stimulation (TENS) on motor recovery [62]. Finally, Conforto et al. [63] reported thattwohours of TENS inthe medianterritoryenabledtheimprovementofhandgripfunction inchronichemiplegicpatients.

The effectsof passivemobilizationare probablylimitedif they are not integrated within a finalized movement and an associative learning process. In fact, coupled stimulation

Fig. 1.Principles of new rehabilitation techniques post-stroke. Neuroimaging [31,32] and electrophysiology [33] studies showed an imbalance of the interhemisphericequilibrium,responsibleforanoveractivationofthehealthyhemisphereandunder-activationoftheinjuredhemisphere.Theobjectiveofthese newrehabilitationtechniquesistorestoretheinterhemisphericbalancebystimulatingtheinjuredhemisphereand/orinhibitingthehealthyhemisphere(CIT: constraint-inducedtherapy;NIBS:non-invasivebrainstimulation–rTMS,tDCS).

techniques associating peripheral electrical stimulation and central stimulation via TMS done in healthy [64–66] and hemiplegic[67]patients,showedthatsensorystimulationsare oftenmoreeffectivewhen coupledtoanassociativelearning conditioningtraining.

Afterwards,roboticsandvirtualrealitytechniquesappeared asinterestingtechniquessincethesetoolscanproviderepeated andfinalizedmovementeveninseverelyimpaired patients. 1.8. Post-lesionbrainplasticityand criticalperiod

Finally, studiesconducted inrats, underlinethe notion of critical period similar to the framework of developmental plasticity.Thepositiveeffectsoftheenrichedenvironmentsare no longer perceived if the animal is placed in this type of environmentbeyondweek4post-stroke[68,69].Theseresults obtainedinanimalmodelsaresimilartootherresultsreported inhumansubjects.Therecoverytimelinepost-strokedescribed in a cohort study promotes these mechanisms [2]. A longitudinal study of coupled stimulation in stroke patients withhemiplegia showed thatbrain plasticity capacities were greaterthreemonthsafterthestrokethanat1-yearpost-stroke

[67].Thisnotionofcriticalperioddoesnotimplytheabsence ofbrainplasticityonthelongterm,butratherthatintheweeks following the lesion, facilitation mechanisms and brain plasticityarepredominantcomparedtomechanismsregulating brain plasticity inhibition [69]. These results underline the relevanceofintensivedailyrehabilitationcareduringthefirst fewweekspost-stroke.

Neuroscienceandclinicalstudiesfromtheendofthe20th centuryhave completelyalteredthe waywe approach stroke rehabilitation. We went from rehabilitation care based on empirical methods that were quite analytic, to rehabilitation carefocusedontasksandrepetition:useorlosethefunction. The principles of stroke rehabilitation for patients with hemiplegiaarenowquitesimilartotheonesformotorlearning. Disclosureofinterest

The authors declare thattheyhaveno conflictsof interest concerningthisarticle.

2. Version franc¸aise 2.1. Introduction

Lare´cupe´rationestunphe´nome`nespontane´ quiestobserve´ meˆme en l’absence de re´adaptation et qui est connu des cliniciensdepuislongtemps.Brocaau19esie`clede´finissaitle syndrome vasculaire comme l’apparition d’un de´ficit brutal, focal etqui s’ame´liore avec le temps,incluant donc dans la de´finitionmeˆmedel’AVClanotiondere´cupe´ration[1].

Dans lesanne´es1980,les e´tudesdecohortesontpermisde ve´rifiercetteame´liorationdesde´ficiencesetdel’autonomiedans lessemaines quisuivent unAVC. Lacohorte deCopenhague

[2,3]publie´edanslesAPMRdanslesanne´es1990estunedeces e´tudes.Soncaracte`reprospectif,exhaustif,dansunsyste`mede

sante´ danois relativement homoge`ne en particulier dans le domaine dela re´adaptation, en fontune e´tudeemble´matique. Cette e´tude clinique a e´value´ les de´ficiences motrices et l’autonomieapre`sAVC,del’admissionenunite´ deneurologie jusqu’a` la sortie du centre de re´adaptation. Ses re´sultats confirment l’existence d’une re´cupe´ration motrice touchant, contrairementauxnotions tropsouventre´panduesparlasuite, toutes les cate´gories de patientshe´miple´giques, y compris les plus se´ve`rement atteints initialement : 40 % des patients pre´sentantunde´ficitmoteurtre`sse´ve`rea` laphaseaigue¨ ontun de´ficitle´geroumoyena` l’issueduprocessusdere´cupe´ration[3]. Lesecondinte´reˆtdecettee´tudeestd’avoire´tudie´ lede´cours temporel de cette re´cupe´ration : l’essentiel des progre`s en termes de de´ficiences motrices est effectue´ au cours des 15premie`ressemainesquellequesoitl’importancedude´ficit moteurinitial[2].Ils’agitla` d’unre´sultattouta` faitoriginala` unee´poqueou` ile´taitdecoutumedeparlerdere´cupe´rationsur des pe´riodes de l’ordre de plusieurs anne´es. Par ailleurs, ce de´courstemporelrelativementrapidesugge`reparanalogieavec lede´veloppementdel’enfant,lanotionencorediscute´edenos jours,depe´riodecritique.Plusieursautrese´tudesontconfirme´ cette cine´tique de la re´cupe´ration motrice et il est classique maintenant de de´finir le de´ficit moteur se´quellaire comme stableetchroniquea` partirdu6emois[4].

2.2. Lare´adaptationame´liorelare´cupe´ration

Paralle`lement, le nombre de publications concernant l’influence de lare´adaptation ou de lakine´sithe´rapie surces processus de re´cupe´ration n’a cesse´ d’augmenter depuis les anne´es80.Unedespremie`resrevuessyste´matiquessurcesujet estpublie´e parOttenbacheretJannellen1993[5].Depuisde nombreuses me´ta-analyses ont e´te´ consacre´es a` ce sujet [4– 17].Lesconclusionsdecesrevuesrestentconstantesaufildu temps. Tout d’abord ces diffe´rents travaux n’ont cesse´ de renforcer l’e´vidence de l’efficacite´ de la re´e´ducation sur la re´cupe´ration motrice apre`s AVC : actuellement 467 essais controˆle´speuventeˆtreidentifie´ssurcettequestionrepre´sentant 25373patients[12].Lapriseenchargedere´adaptationpermet d’ame´liorer la re´cupe´ration du de´ficit moteur du membre infe´rieur[9,16,17]etdumembresupe´rieur[6,9],delamarche

[13]etdel’e´quilibre[16,17],del’autonomiedanslesactivite´s delaviequotidienne[4,5,11,14].

Les facteurs d’efficacite´ de la re´adaptation apre`s AVC restent les meˆmes de publications en publications depuis maintenant30ans.

La pre´cocite´ de la prise en charge en re´adaptation est un facteurdel’efficacite´ delare´adaptation[4,10,18,19].Ilexiste certesdesinterrogationssurlecaracte`rede´le´te`redel’exercice aude´coursdelaphaseinitialeensoinsintensifs,meˆmesiles re´sultats de deux essais re´cents sont plutoˆt rassurants

[20,21]. Mais au-dela` de cette premie`re semaine, toutes les e´tudesmontrent quelare´adaptation doiteˆtrede´bute´e.

L’intensite´ delapriseenchargeestaussiunfacteuressentiel de l’efficacite´ de la re´adaptation [8,9]. Ici encore la dose optimale n’est pas clairement de´termine´e, mais 30 min de kine´sithe´rapiequotidiennesontplusefficacesque30mintrois

P.Marqueetal./AnnalsofPhysicalandRehabilitationMedicine57(2014)520–529 524

fois par semaine et semblent eˆtre le minimum recommande´ pourobserveruneefficacite´ delapriseenchargesurlamarche etl’autonomie[5,8,9].Danslecadredesprotocolesdethe´rapie contrainte,ladure´equotidiennedere´adaptationapueˆtreporte´e jusqu’a` 6h parjour[4,6].

Enfinl’effet e´quipe estune´le´mentre´gulie`rementretrouve´

[4,11,18].Lapriseenchargedansdesunite´spluridisciplinaires, spe´cialise´es en neurore´adaptation semble plus efficace que cellepropose´ paruneunite´ dere´adaptationnonspe´cialise´eou noncoordonne´e.

Par contre, ces e´tudes ne permettent pas d’identifier une technique de kine´sithe´rapie spe´cifique a` l’he´miple´gique vasculaire. Il s’agit d’un re´sultat essentiel dans le contexte desanne´es1990.Eneffet,lakine´sithe´rapieapre`sAVCa` cette e´poqueestconstitue´edediffe´rentestechniques empiriquesle plussouventidentifie´esparlenomdeleur cre´ateur:Bobath, Brunstrom, Vojda, Kabat, Perfetti... 16 essais controˆle´s ont compare´ cesdiffe´rentestechniquesetplusparticulie`rementles deuxplusutilise´sa` cettee´poque:BobathetBrunstrom[5,7,17]. Sicescomparaisons peuventapparaıˆtre de´sue`tes aujourd’-hui, elles restent sur le plan fondamental, relativement inte´ressantes.Eneffetsousleursformesdel’e´poque,cesdeux techniquespeuventapparaıˆtrecommeradicalementoppose´es. L’une,lame´thodeBobath,proposedestechniquesd’inhibition delaspasticite´ etdesre´actionsautomatiques,ellerecommande d’e´viter tout exercice en force suspect d’augmenter ces phe´nome`nes spastiques conside´re´s comme archaı¨ques et pe´joratifspourlare´cupe´ration.L’autre,lame´thodeBrunstrom, utilisedes techniques de renforcement des re´actions re´flexes automatiques,y compriscontrere´sistance,quisont rigoureu-sementprohibe´esparlame´thodeBobath.Pourtantaucunessai comparatif n’a permis de mettre en e´vidence de supe´riorite´ d’unedesestechniques surl’autre[7,17].

Paralle`lementplusieurstravauxontpropose´ desconclusions qui n’e´taient pas favorables aux concepts sous-tendant la me´thodeBobath.Thilmannetal.[22]montrentquelegaindu re´flexed’e´tirementestplusmodifie´ queleseuild’apparitiondu re´flexechez le sujet spastique. Il ne retrouve pas de re´flexe d’e´tirementendessousde358/s.Endessousdecesvitessesles difficulte´ssontdoncplusenrapportaveclaparalysiequ’avecla spasticite´.Demeˆme,d’autrestravauxontmontre´ quel’exercice physiqueetenparticulier,lerenforcementmusculairequie´tait pe´nalise´ dansla me´thodeBobath, n’augmentait pasde fac¸on chroniquelaspasticite´[23,24].

La kine´sithe´rapie apparaıˆt donc comme un traitement efficace pour favoriser la re´cupe´ration motrice apre`s AVC, maislesmodalite´spre´cisesdecetteefficacite´ restentencorea` de´finirdanslesanne´es1990.Ilsemblepourtantde´ja` clairque celles-ci ne passent pas par l’acquisition de techniques spe´cifiques, empiriques difficiles a` maıˆtriser pour la plupart deskine´sithe´rapeutes.

2.3. Miseene´videncedelaplasticite´ ce´re´brale apre`s le´sionvasculaire

La plasticite´ ce´re´brale est de´finie comme l’ensemble des me´canismes qui permettent au cerveau d’adapter son

fonctionnementa` unesituationnouvelle.Lamiseene´vidence d’uneame´liorationcliniquedurableapre`sunele´sionce´re´brale vasculaire est en soit de´ja` la mise en e´vidence de ce type d’adaptation. De´ja` dans les anne´es 1980, par de´duction des re´sultats d’e´tudes re´alise´es chez l’animal dans des contextes pourtantsouvente´loigne´s,certainsauteurs[25]proposaientpour expliquerlare´cupe´ration fonctionnelle apre`sle´sion ce´re´brale chez l’adulte, des me´canismes de de´masquage de re´seaux neuronaux redondants,des modificationsd’efficacite´ synapti-quesoudesphe´nome`nesdebourgeonnementsdendritiques.

L’arrive´e de l’imagerie fonctionnelle ce´re´brale a permis d’objectivercesme´canismesdelare´cupe´rationdelamotricite´. Ces techniques d’imagerie permettent de mesurer le de´bit sanguince´re´brallocal soitparl’utilisationd’un radio-isotope dans le cas du TEP, soit par l’utilisation des proprie´te´s paramagne´tique de l’oxyhe´moglobine dans l’IRMf. Ces me´thodes font l’hypothe`se qu’au niveau ce´re´bral ce de´bit sanguinestline´airement corre´le´ a` l’activationneuronale.Les diffe´rences de de´bit entre deux taˆches e´tant modestes, elles ne´cessitentpoureˆtremisesene´videncelare´pe´titiondestaˆches et une analyse statistiques des diffe´rences observe´es. Les re´sultats sont doncpre´sente´s sous formede cartographie des diffe´rences statistiques obtenues entre deux situations : mouvement et repos par exemple. L’e´tude en TEP de l’activationce´re´bralere´alise´echezdescaste´moinsayantbien re´cupe´re´ d’unAVClacunairesous-corticalmontrequ’ilexiste une re´organisation de la somatotopie des aires motrices primaireschezcespatientslorsdelare´alisationdemouvements d’oppositiondupouceetdesautresdoigts. Cholletetal.[26]

retrouventuneactivationinte´ressantlesdeuxhe´misphe`res,lors d’un mouvementde la main pare´tique.De son coˆte´, Weiller et al. [27] observent chez un patient similaire, non pas une activation bilate´rale, mais un de´placement de l’activation corticalecorrespondanta` lamain,verslesairesdelaface. Au-dela` deleur interpre´tationqui restecomplique´e,ces re´sultats objectivent pour la premie`re fois l’existence d’une plasticite´ ce´re´braleapre`sle´sionce´re´bralechezl’adulte.

L’imagerie ce´re´brale a permis aussi de valider certains me´canismes mis en jeu au cours de ces phe´nome`nes de plasticite´ ce´re´brale. Le diaschisis [28] fut le premier de ces me´canismes. Le diaschisis correspond a` la baisse de de´bit sanguin ce´re´bral dans certaines zones ce´re´brales non le´se´es mais relie´es fonctionnellement avec la zone ne´crose´e. En d’autre terme,unepartiedessymptoˆmesde´ficitairesobserve´s chez lespatients victimes de le´sions neurovasculaires est en relation avec un phe´nome`ne de de´affe´rentation d’aires ce´re´bralesnon le´se´esmais recevant desprojectionsissuesde lazonele´se´e.Lediaschisisce´re´belleuxfutundepremiersa` eˆtre misene´videnceetpourraitexpliquerlecaracte`retre`sataxique decertainesformesdede´ficitsmoteursobserve´sapre`sle´siondu territoiredel’ACM.

Le de´masquage d’un re´seau pre´existant mais quiescent capable de supple´er la fonction de´ficitaire fut le second me´canismemis e´vidence.Lorsque cere´seau ce situe dans le meˆme territoire fonctionnel le terme de redondance [29] est habituellementutilise´,lorsquecettesupple´ancefaitappela` des re´seaux plus lointains situe´s dans des territoires initialement

utilise´spard’autresfonctions,letermedevicarianceestpre´fe´re´

[30]. Dans tous les cas, pour que ces re´seaux deviennent fonctionnels des modifications de l’efficacite´ synaptique de typeLTP ouLTDsontne´cessaires.

2.4. Laplasticite´ ce´re´brale:unphe´nome`ne toujours positif?

Ne´anmoins,ilafalluplusd’unede´cenniepourcomprendre lesens exactde ces modifications. Tout d’abord, Loubinoux etal.[31,32]ontde´critlede´courstemporeldeceschangements chez des patients ayant bien re´cupe´re´. Elle montre que l’interpre´tationdesesre´sultatsn’estpasunivoque.Eneffet15j apre`s l’AVC l’augmentation des activations observe´es sont plutoˆt ipsilate´rales a` lamain mobilise´e avec une diminution d’activation dans le cortex du coˆte´ le´se´. Ces images apparaissaient en de´finitive relativement proches de celle observe´epar Chollet 10ansplus toˆt[26].Quatre mois apre`s l’AVConassistea` unerelate´ralisationdesactivations ducoˆte´ controlate´ral[31,32].Cesactivationsrestentsignificativement plusimportantesquedanslapopulationte´moin.Unanplustard cesauteursneretrouventplusdediffe´rencesaveclegroupede te´moin[31,32].

Paralle`lement Murase et al. en 2004 [33] mettaient en e´vidence le roˆle de´le´te`re de l’inhibition interhe´misphe`rique apre`sle´sionce´re´brale.Eneffetcetauteuramontre´ qu’ilexiste dans les20ms qui pre´ce`dent unmouvement unilate´ralde la main,une leve´e des me´canismes d’inhibition interhe´misphe`-rique normalement observe´e chez le sujet sain au repos

[34].Cettediminutionesttoujourslimite´eetdanstouslescas beaucoup moinsimportante du coˆte´ he´miple´gique aussibien parcomparaisonaucoˆte´ sainqu’a` unepopulationdesujetsain. Murase et al. de´veloppent le concept de plasticite´ ce´re´brale inadapte´e [33]. La le´sion de´clenche une diminution de l’inhibition interhe´misphe´rique de l’he´misphe`re le´se´ vers le sain. Cette diminution conduit a` une hyperactivation des re´seaux moteurs ipsilate´raux qui en retour vont augmenter l’inhibitionde l’he´misphe`re le´se´. Ceme´canismede de´se´qui-libredelabalanceinhibitriceinterhe´misphe`rique,pourraitau moins initialement, majorer le de´ficit et expliquer les activations observe´es par Loubinoux et al. 15jours apre`s l’AVC[31,32].Laplasticite´ ce´re´bralepouvaiteˆtreunfacteur ne´gatifdans certainecirconstance.

2.5. Laplasticite´ ce´re´braleestlimite´e

Les e´tudes de neuroimageries pre´sentent certainement plusieurssourcesdebiais:petitepopulation,sur-repre´sentation desle´sions sous-corticales et des de´ficits mode´re´s. Elles ont permispourtantdeproposerunehie´rarchiedanslesprocessus dere´cupe´rationetdefixerleslimitesdelaplasticite´ ce´re´brale spontane´e. Plus lere´seau de supple´ance estproche anatomi-quement et fonctionnellement du re´seau initial le´se´ (redon-dance)pluslede´ficitestle´geretmeilleureseralare´cupe´ration

[35–41].A` l’oppose´ pluslere´seaudecompensationsesituea` distance du re´seau initial plus le de´ficit sera lourd et la re´cupe´ration partielle ou de mauvaise qualite´ [26,42–44].

L’utilisation d’un re´seau de compensation controlate´ral a` la le´sion et donc ipsilate´ral aux mouvements la main le´se´ se traduit chez l’adulte he´miple´gique, par une re´cupe´ration de mauvaise qualite´ et par lapersistanced’un de´ficitimportant. Pour mode´liser ce principe, un indice de late´ralisation a e´te´ propose´ par certains auteurs [45–47]. Cet indicefait le ratio entre les volumes de pixels active´s dans les aires motrices controlate´rales au mouvement et dans l’ensemble des aires motrices ipsi- ou controlate´rales. Cet indice varie entre 0 et 1.Unindiceprochede1signeuneplasticite´ ce´re´bralefavorable et est en faveur d’une bonne re´cupe´ration. Cet indice est maintenant utilise´ dans les e´tudes en neuroimagerie comme crite`reinterme´diairepour e´valuerl’influencesur la re´cupe´ra-tiondecertainesinterventionsre´e´ducativesou me´dicamenteu-ses[48,49].

2.6. Lesmode`lesanimauxde re´e´ducation

C’est dans les anne´es 1980 que sont aussi apparus les mode`les animaux. Le premier de ces mode`les est celui de l’environnementenrichietconcernelerat[50].Ils’agitd’un mode`le initialementutilise´ pour e´tudierlaplasticite´ de´velop-pementale chez le nouveau-ne´.Ce mode`le a e´te´ adapte´ a` la plasticite´ postle´sionnelle.Lemilieuenrichiestrepre´sente´ par une cage de taille plus importante, avec une diversite´ architecturale et d’accessoire, ou` plusieurs individus sont pre´sents. Il s’agit donc d’un environnement qui stimule les relations inter-individus et qui implicitement impose aux individus uneactivite´ motrice plusimportante quedans une cage individuelle. Les ratsplace´s apre`s une le´sionce´re´brale vasculairedel’arte`rece´re´bralemoyenne,dansdescagesavec environnement enrichi pre´sentaient au bout de quelques semaines, une re´cupe´ration motrice de meilleure qualite´ sur les e´preuves cliniques que ceux place´s dans des cages individuelles[50].De plussurleplanhistologiqueladensite´ dedendritesparneuronese´taitplusgrandepourlesratsplace´s en milieu enrichis te´moignant d’un bourgeonnement synap-tiqueplusimportant [50].

Demeˆmeunmode`ledere´e´ducationae´te´ de´veloppe´ chezle singe. Les techniques de microstimulations corticales per-mettent de re´aliser des cartographies ce´re´brales pre´cises et reproductibles.Nudoetal.onte´tudie´ lare´organisationdeses cartographiesapre`slare´alisationd’unele´sionvasculairedans leterritoiredelamainchezdessingesmacaques[51].Ilnote toutd’abordquedanslesheuresquisuiventlale´sion,dessites situe´senpe´riphe´riedelale´sionre´pondantpardescontractions dupoignetetdesdoigts,apparaissent.Cesre´sultatste´moignent d’une plasticite´ rapide, probablement par de´masquage de re´seauxredondantsquiescents.MaisNudoetal.observentau bout de quelques semaines une disparition de ces sites de stimulationre´pondantpourlepoignetetlesdoigts,auprofitde sitesre´pondantpourlecoudeetl’e´paule[51].A` l’opposer,les singeschezlesquelsonimmobiliselemembresupe´rieursain, contraignant ainsi l’usage du membre de´ficitaire, voient le nombre de sites et la surface corticale dont la stimulation entraıˆne une re´ponse au niveau des muscles de la main, du poignetet desdoigts, augmenter [52,53].Cette e´tudemeten

P.Marqueetal./AnnalsofPhysicalandRehabilitationMedicine57(2014)520–529 526

e´vidence une plasticite´ de´pendante de l’usage ainsi que des phe´nome`nesde compe´titionentrelesrepre´sentations cortica-les. Des phe´nome`nes similaires de compe´titions des repre´-sentationscorticales,de´pendantesdel’usageoudesaffe´rences sensitives ont e´te´ mis en e´vidence chez l’homme : chez l’ampute´ du membre supe´rieur [54]ou dans les minutesqui suiventlamiseenplaced’ungarrot[55].Unve´ritablesloganest alorspropose´ commebasedelare´e´ducation:«utilises leou perdsle».

2.7. Stimulationssensitives,the´rapieinduiteparla contrainte

Lathe´rapie induiteparlacontrainte dumembre supe´rieur sain est une technique qui a aussi de´montre´ son efficacite´ l’hommeaussibiensurleplancliniquedans dese´tudesavec une me´thodologie de bon niveau [56], que sur des crite`res interme´diairesdansdesessaisdeneuroimageriefonctionnelle

[40]oudescartographiescorticalesre´alise´esparTMS[57]. Maislathe´rapiecontrainteestunetechniquequis’adressea` une petite fraction des malades pare´tiques. Elle reste inapplicablechezlespatientstropde´ficitaires oupendantles premie`rese´tapesdelare´adaptation.

La mobilisation passive participe de l’efficacite´ de la re´adaptationaucoursdespremie`rese´tapesetchezlespatients totalementple´giques.Eneffet,lestravauxdeneuroimagerieen TEP[58]ou EEG/IRMf [59]ont montre´ quelamobilisation passivepermettaitd’activerlesre´seauxmoteurs.De meˆmela re´alisation d’un entraıˆnement passif est capable de modifier durablementl’activationce´re´braleenIRMfchezlesujetsain

[60]commechezlesujethe´miple´gique[61].Cesre´sultatssont concordants avec des e´tudes cliniques qui montraient l’effet

favorable des TENS sur la re´cupe´ration motrice [62]. Enfin Conforto et al.[63] ont montre´ que la re´alisationde 2h de TENS dans leterritoire du me´dian permettait d’ame´liorer le handgripdepatienthe´miple´giquechronique.

Les effets des mobilisations passives sont probablement limite´ss’ilsnes’inte`grent pasdans unmouvementfinalise´ et dansuneformed’apprentissagedetypeassociatif.Eneffetles techniquesdestimulationscouple´esassociantunestimulation e´lectriquepe´riphe´riqueetunestimulationcentraleparTMSet quionte´te´ re´alise´eschezlesujetsain[64–66]etl’he´miple´gique

[67], montrent que les stimulations sensitives sont plus efficaces lorsqu’elles s’inte`grent dans une forme de condi-tionnementassociatifdecetype.

C’estdans ce sensque lestechniques derobotiques et de re´alite´ virtuelle sont apparues comme des techniques inte´r-essantescarcesontdesoutilsquipermettentdere´pondrea` ces objectifsdemouvementsre´pe´te´setfinalise´sycomprischezdes malades tre`sde´ficitaires.

2.8. Plasticite´ ce´re´bralepostle´sionnelleetpe´riodecritique Enfinlese´tudesre´alise´eschezleratfonte´mergerlanotion de pe´riode critique comme dans le cadre de la plasticite´ de´veloppementale. Les effets positifs de l’environnement enrichine sontplusperc¸ussi l’animalestplace´ dansce type d’environnement au-dela` de la 4e semaine apre`s l’ictus

[68,69].Cesre´sultatsobtenuschezl’animalsontconcordants avec d’autres re´sultats obtenus chez l’homme. Le de´cours temporeldelare´cupe´rationapre`sAVCde´critparlese´tudesde cohortesestenfaveurdecetypedeme´canismes[2].L’e´tude longitudinale des stimulations couple´es chez l’he´miple´gique montredescapacite´sdeplasticite´ beaucoupplusimportantea`

Fig.1. Principesdesnouvellestechniquesdere´e´ducationapre`sAVC.Lestravauxdeneuroimagerie[31,32]etd’e´lectrophysiologie[33]ontmontre´ qu’ilexistaitun de´se´quilibredelabalanceinterhe´misphe´rique,responsabled’unehyperactivationdel’he´misphe`resainetd’unehypoactivationdel’he´misphe`rele´se´.L’objectifdes nouvellestechniquesdere´adaptationestderestaurercettebalanceinterhe´misphe´riqueenstimulantl’he´mipshe`rele´se´ et/oueninhibantl’he´misphe`resain.(CIT: the´rapieinduiteparlacontrainte;NIBS:stimulationsce´re´bralesnoninvasives–rTMS,tDCS).

troismoisqu’a` unanapre`sl’AVC[67].Cettenotiondepe´riode critiquenesignifiepasqu’iln’yaplusdeplasticite´ ce´re´bralea` longterme,maisquedanslessemainesquisuiventunele´sion les me´canismes de facilitation de la plasticite´ ce´re´brale est pre´dominantesurlesme´canismesre´gulateursd’inhibitiondela plasticite´ ce´re´brale[69].Cesre´sultatssoulignentl’inte´reˆtd’une prise en charge quotidienne intensive durant les premie`res semaines.

Lestravauxdeneurosciencesetdese´tudescliniquesdelafin du20esie`clemodifienttotalementl’approchedelare´adaptation apre`s un AVC. Nous sommes passe´s d’une prise en charge base´e sur des me´thodes empiriques globales relativement analytiques a` une re´adaptation oriente´e vers la taˆche et la re´pe´tition:utiliserouperdre lafonction.Lesprincipes dela re´adaptation de l’he´miple´gique vasculaire se rapprochent maintenantbeaucoupplusdecellesdel’apprentissagemoteur (Fig.1).

De´clarationd’inte´reˆts

Lesauteursde´clarentne pasavoirdeconflitsd’inte´reˆtsen relationaveccetarticle.

References

[1]Monod-Broca P. [Paul Broca: 1824–1880]. Annales de chirurgie 2001;126:801–7.

[2]JorgensenHS, NakayamaH,RaaschouHO,Vive-LarsenJ,StoierM, OlsenTS.Outcomeandtimecourseofrecoveryinstroke.PartII:time courseofrecovery.TheCopenhagenStrokeStudy.Archivesofphysical medicineandrehabilitation1995;76:406–12.

[3]JorgensenHS, NakayamaH,RaaschouHO,Vive-LarsenJ,StoierM, OlsenTS.Outcomeandtimecourseofrecoveryinstroke.PartI:outcome. TheCopenhagenStrokeStudy.Archivesofphysicalmedicineand reha-bilitation1995;76:399–405.

[4]Langhorne P, Bernhardt J, Kwakkel G. Stroke rehabilitation. Lancet 2011;377:1693–702.

[5]OttenbacherKJ,JannellS.Theresultsofclinicaltrialsinstroke rehabili-tationresearch.Archivesofneurology1993;50:37–44.

[6]OujamaaL,RelaveI,FrogerJ,MottetD,PelissierJY.Rehabilitationof arm function after stroke. Literature review. Annals of physical and rehabilitationmedicine2009;52:269–93.

[7]KollenBJ,LennonS,LyonsB,Wheatley-SmithL,ScheperM,BuurkeJH, etal.TheeffectivenessoftheBobathconceptinstrokerehabilitation:what istheevidence?Stroke;ajournalofcerebralcirculation2009;40:e89–97. [8]KwakkelG,vanPeppenR,WagenaarRC,WoodDauphineeS,Richards C,AshburnA,etal.Effectsofaugmented exercisetherapytimeafter stroke: a meta-analysis. Stroke; a journal of cerebral circulation 2004;35:2529–39.

[9]Kwakkel G, Wagenaar RC, Twisk JW, Lankhorst GJ, Koetsier JC. Intensity oflegand armtraining afterprimarymiddle-cerebral-artery stroke:arandomisedtrial.Lancet1999;354:191–6.

[10]LanghorneP,HolmqvistLW,EarlySupportedDischargeT.Early sup-porteddischargeafterstroke.Journalofrehabilitationmedicine:official journal oftheUEMS EuropeanBoardof PhysicalandRehabilitation Medicine2007;39:103–8.

[11]PelissierJ.Themanagementofstrokepatients.Conferenceofexpertswith apublichearing.Mulhouse(France),22October2008.Annalsofphysical andrehabilitationmedicine2010;53:124–47.

[12]VeerbeekJM,vanWegenE,vanPeppenR,vanderWeesPJ,HendriksE, RietbergM,etal.Whatistheevidenceforphysicaltherapypoststroke?A systematicreviewandmeta-analysis. PloSone2014;9:e87987.

[13] vandePortIG,Wood-DauphineeS,LindemanE,KwakkelG.Effectsof exercisetrainingprogramsonwalkingcompetencyafterstroke:a sys-tematicreview.Americanjournalofphysicalmedicine&rehabilitation/ AssociationofAcademicPhysiatrists2007;86:935–51.

[14] VanPeppenRP,KwakkelG,Wood-DauphineeS,HendriksHJ,Vander WeesPJ,DekkerJ.Theimpactofphysicaltherapyonfunctionaloutcomes afterstroke:what’stheevidence?Clinicalrehabilitation2004;18:833–62. [15] PollockA,BaerG,CampbellP,ChooPL,ForsterA,Morris J,etal. Physical rehabilitation approaches for the recovery of function and mobilityfollowingstroke.TheCochranedatabaseofsystematicreviews 2014;4:CD001920.

[16] PollockA,BaerG,PomeroyV,LanghorneP.Physiotherapytreatment approachesfortherecoveryofposturalcontrolandlowerlimbfunction following stroke. The Cochrane database of systematic reviews 2003;CD001920.

[17] PollockA,BaerG,PomeroyV,LanghorneP.Physiotherapytreatment approachesfortherecoveryofposturalcontrolandlowerlimbfunction following stroke. The Cochrane database of systematic reviews 2007;CD001920.

[18] KalraL,LanghorneP.Facilitatingrecovery:evidencefororganizedstroke care.Journalofrehabilitationmedicine: officialjournaloftheUEMS EuropeanBoardofPhysicalandRehabilitationMedicine2007;39:97–102. [19] RousseauxM,DaveluyW, Kozlowski R.Value andefficacyof early supporteddischargefromstrokeunits.Annalsofphysicaland rehabilita-tionmedicine2009;52:224–33.

[20] BernhardtJ, Dewey H,Thrift A,Collier J,DonnanG. Avery early rehabilitationtrialforstroke(AVERT):phaseIIsafetyandfeasibility. Stroke;ajournalofcerebralcirculation2008;39:390–6.

[21] LanghorneP,StottD,KnightA,BernhardtJ,BarerD,WatkinsC.Very earlyrehabilitationorintensivetelemetryafterstroke:apilotrandomised trial.Cerebrovasculardiseases2010;29:352–60.

[22] ThilmannAF,FellowsSJ,GarmsE.Themechanismofspasticmuscle hypertonus.Variationin reflexgainoverthetimecourseofspasticity. Brain:ajournalofneurology1991;114:233–44.

[23] SaundersDH,GreigCA,MeadGE,YoungA.Physicalfitnesstrainingfor stroke patients. The Cochrane database of systematic reviews 2009;CD003316.

[24] SaundersDH,SandersonM,BrazzelliM,GreigCA,MeadGE.Physical fitnesstrainingforstrokepatients.TheCochranedatabaseofsystematic reviews2013;10:CD003316.

[25] DombovyML,Bach-y-RitaP.Clinical observationsonrecoveryfrom stroke.Advancesinneurology1988;47:265–76.

[26] CholletF,DiPieroV,WiseRJ,BrooksDJ,DolanRJ,FrackowiakRS.The functionalanatomyofmotorrecoveryafterstrokeinhumans:astudywith positronemissiontomography.Annalsofneurology1991;29:63–71. [27] WeillerC,CholletF,FristonKJ,WiseRJ,FrackowiakRS.Functional

reorganizationofthebraininrecoveryfromstriatocapsularinfarctionin man.Annalsofneurology1992;31:463–72.

[28] SeitzRJ,AzariNP,KnorrU,BinkofskiF,HerzogH,FreundHJ.Therole ofdiaschisisinstrokerecovery.Stroke;ajournalofcerebralcirculation 1999;30:1844–50.

[29] GreenJB.Brainreorganizationafterstroke.Topicsinstrokerehabilitation 2003;10:1–20.

[30] DancauseN.Vicariousfunctionofremotecortexfollowingstroke:recent evidencefromhumanandanimalstudies.TheNeuroscientist:areview journalbringingneurobiologyneurologyandpsychiatry2006;12:489–99. [31] LoubinouxI,CarelC,ParienteJ,DechaumontS,AlbucherJF,MarqueP, etal.Correlationbetweencerebralreorganizationandmotorrecovery aftersubcorticalinfarcts.NeuroImage2003;20:2166–80.

[32] TombariD,LoubinouxI,ParienteJ,GerdelatA,AlbucherJF,TardyJ, etal.AlongitudinalfMRIstudy:inrecoveringandtheninclinicallystable sub-corticalstrokepatients.NeuroImage2004;23:827–39.

[33] MuraseN,DuqueJ,MazzocchioR,CohenLG.Influenceof interhemi-spheric interactions on motor function in chronic stroke. Annals of neurology2004;55:400–9.

[34] FerbertA,PrioriA,RothwellJC,DayBL,ColebatchJG,MarsdenCD. Interhemisphericinhibitionofthehumanmotorcortex.TheJournalof physiology1992;453:525–46.

P.Marqueetal./AnnalsofPhysicalandRehabilitationMedicine57(2014)520–529 528

[35] LoubinouxI,Dechaumont-PalacinS,Castel-LacanalE,DeBoissezonX, MarqueP,ParienteJ,etal.PrognosticvalueofFMRIinrecoveryofhand functioninsubcorticalstrokepatients.Cerebralcortex2007;17:2980–7. [36] WerhahnKJ,ConfortoAB,KadomN,HallettM,CohenLG.Contribution oftheipsilateralmotorcortextorecoveryafterchronicstroke.Annalsof neurology2003;54:464–72.

[37] RouillerEM,OlivierE.Functionalrecoveryafterlesionsoftheprimary motorcortex.Progressinbrainresearch2004;143:467–75.

[38] FridmanEA,HanakawaT,ChungM,HummelF,LeiguardaRC,Cohen LG. Reorganization of the human ipsilesional premotor cortex after stroke.Brain:ajournalofneurology2004;127:747–58.

[39] Johansen-BergH,RushworthMF,BogdanovicMD,KischkaU, Wima-laratnaS,MatthewsPM.Theroleofipsilateralpremotorcortexinhand movementafterstroke.ProceedingsoftheNationalAcademyofSciences oftheUnitedStatesofAmerica2002;99:14518–23.

[40] SchaechterJD,KraftE,HilliardTS,DijkhuizenRM,BennerT, Finkles-teinSP,etal.Motorrecoveryandcorticalreorganizationafter constraint-inducedmovementtherapyinstrokepatients:apreliminarystudy. Neu-rorehabilitationandneuralrepair2002;16:326–38.

[41] NellesG,JentzenW,JueptnerM,MullerS,DienerHC.Armtraining inducedbrainplasticityinstrokestudiedwithserialpositronemission tomography.NeuroImage2001;13:1146–54.

[42] ZemkeAC,HeagertyPJ,LeeC,CramerSC.Motorcortexorganization afterstrokeisrelatedtosideofstrokeandlevelofrecovery.Stroke;a journalofcerebralcirculation2003;34:e23–8.

[43] CareyJR,KimberleyTJ,LewisSM,AuerbachEJ,DorseyL,RundquistP, etal.Analysis offMRIand fingertracking training in subjectswith chronicstroke.Brain:ajournalofneurology2002;125:773–88. [44] SeitzRJ,KnorrU,Azari NP,Herzog H,FreundHJ. Visualnetwork

activationinrecoveryfromsensorimotorstroke.Restorativeneurology andneuroscience1999;14:25–33.

[45] BaronJC,CohenLG,CramerSC,DobkinBH,Johansen-BergH, Lou-binouxI,etal.Neuroimaginginstrokerecovery:apositionpaperfromthe FirstInternational Workshopon Neuroimaging and Stroke Recovery. Cerebrovasculardiseases2004;18:260–7.

[46] WardNS.Functionalreorganizationofthecerebralmotorsystemafter stroke.Currentopinioninneurology2004;17:725–30.

[47] WardNS,CohenLG.Mechanismsunderlyingrecoveryofmotorfunction afterstroke.Archivesofneurology2004;61:1844–8.

[48] ManganottiP,AclerM,FormaggioE,AvesaniM,MilaneseF,BaraldoA, etal.Changesincerebralactivityafterdecreasedupper-limbhypertonus: anEMG-fMRIstudy.Magneticresonanceimaging2010;28:646–52. [49] SenkarovaZ,HlustikP,OtrubaP,HerzigR,KanovskyP.Modulationof

corticalactivityinpatientssufferingfromupperarmspasticityfollowing strokeandtreatedwith botulinumtoxinA:anfMRIstudy.Journalof neuroimaging:officialjournaloftheAmericanSocietyofNeuroimaging 2010;20:9–15.

[50] BiernaskieJ,CorbettD.Enrichedrehabilitativetrainingpromotes im-provedforelimbmotorfunctionandenhanceddendriticgrowthafterfocal ischemicinjury.TheJournalofneuroscience:theofficialjournalofthe SocietyforNeuroscience2001;21:5272–80.

[51] NudoRJ,MillikenGW,JenkinsWM,MerzenichMM.Use-dependent alterationsofmovementrepresentationsinprimarymotorcortexofadult squirrelmonkeys.TheJournalofneuroscience:theofficialjournalofthe SocietyforNeuroscience1996;16:785–807.

[52] NudoRJ,MillikenGW.Reorganizationofmovementrepresentationsin primarymotorcortexfollowingfocalischemicinfarctsinadultsquirrel monkeys.Journalofneurophysiology1996;75:2144–9.

[53] NudoRJ,WiseBM,SiFuentesF,MillikenGW.Neuralsubstratesforthe effectsofrehabilitativetrainingonmotorrecoveryafterischemicinfarct. Science1996;272:1791–4.

[54] KewJJ,RiddingMC,RothwellJC,PassinghamRE, LeighPN, Soor-iakumaranS,etal.Reorganizationofcorticalbloodflowandtranscranial magneticstimulationmapsinhumansubjectsafterupperlimb amputa-tion.Journalofneurophysiology1994;72:2517–24.

[55] Brasil-NetoJP,CohenLG,Pascual-LeoneA,JabirFK,WallRT,Hallett M.Rapidreversiblemodulationofhumanmotoroutputsaftertransient deafferentationoftheforearm:astudywithtranscranialmagnetic stimu-lation.Neurology1992;42:1302–6.

[56] WolfSL,WinsteinCJ,MillerJP,TaubE,UswatteG,MorrisD,etal.Effect ofconstraint-inducedmovementtherapyonupperextremityfunction3to 9monthsafterstroke:theEXCITErandomizedclinicaltrial.JAMA:the journaloftheAmericanMedicalAssociation2006;296:2095–104. [57] LiepertJ,BauderH,WolfgangHR,MiltnerWH,TaubE, WeillerC.

Treatment-inducedcorticalreorganizationafterstrokeinhumans.Stroke; ajournalofcerebralcirculation2000;31:1210–6.

[58] WeillerC,JuptnerM,FellowsS,RijntjesM,LeonhardtG,KiebelS,etal. Brain representation of active and passive movements. NeuroImage 1996;4:105–10.

[59] AlaryF,DoyonB,LoubinouxI,CarelC,BoulanouarK,RanjevaJP,etal. Event-relatedpotentialselicitedbypassivemovementsinhumans: char-acterization, source analysis, and comparison to fMRI. NeuroImage 1998;8:377–90.

[60] CarelC,LoubinouxI,BoulanouarK,ManelfeC,RascolO,CelsisP,etal. Neuralsubstratefortheeffectsofpassivetrainingonsensorimotorcortical representation:astudywithfunctionalmagneticresonanceimagingin healthysubjects.Journalofcerebralbloodflowandmetabolism:official journaloftheInternationalSocietyofCerebralBloodFlowand Metabo-lism2000;20:478–84.

[61] Dechaumont-PalacinS,MarqueP,DeBoissezonX,Castel-LacanalE, CarelC,BerryI,etal.Neuralcorrelatesofproprioceptiveintegrationin thecontralesionalhemisphereofveryimpairedpatientsshortlyaftera subcorticalstroke:anFMRIstudy.Neurorehabilitationandneuralrepair 2008;22:154–65.

[62] SondeL,GipC,FernaeusSE,NilssonCG,ViitanenM.Stimulationwith lowfrequency(1.7Hz)transcutaneouselectricnervestimulation (low-tens)increasesmotorfunctionofthepost-strokepareticarm. Scandina-vianjournalofrehabilitationmedicine1998;30:95–9.

[63] Conforto AB, Kaelin-Lang A, Cohen LG. Increase in hand muscle strengthofstroke patientsaftersomatosensory stimulation.Annalsof neurology2002;51:122–5.

[64] StefanK,KuneschE,CohenLG,BeneckeR,ClassenJ.Inductionof plasticityinthehumanmotorcortexbypairedassociativestimulation. Brain:ajournalofneurology2000;123:572–84.

[65] Ridding MC, Uy J.Changes in motor cortical excitability induced by paired associative stimulation. Clinical neurophysiology: official journal of the International Federation of Clinical Neurophysiology 2003;114:1437–44.

[66] Castel-LacanalE,Gerdelat-MasA,MarqueP,LoubinouxI, Simonetta-MoreauM.Inductionofcorticalplasticchangesinwristmusclesbypaired associativestimulationinhealthysubjectsandpost-strokepatients. Ex-perimentalbrainresearch2007;180:113–22.

[67] Castel-LacanalE,MarqueP,TardyJ,deBoissezonX,GuiraudV,Chollet F,etal.Inductionofcorticalplasticchangesinwristmusclesbypaired associativestimulationintherecoveryphaseofstrokepatients. Neuror-ehabilitationandneuralrepair2009;23:366–72.

[68] BiernaskieJ,ChernenkoG,CorbettD.Efficacyofrehabilitative experi-encedeclineswithtimeafterfocalischemicbraininjury.TheJournalof neuroscience: the official journal of the Society for Neuroscience 2004;24:1245–54.

[69] MurphyTH,CorbettD.Plasticityduringstrokerecovery:fromsynapseto behaviour.NaturereviewsNeuroscience2009;10:861–72.