7

Click here to load reader

Mechanisms of functional recovery after stroke: Insights from imaging

  • Upload
    j-c

  • View
    214

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Mechanisms of functional recovery after stroke: Insights from imaging

Pratique Neurologique – FMC 2012;3:160–166Cerebral plasticity; post-stroke

160

Mechanisms of functionalrecovery after stroke:Insights from imaging

Mécanismes et exploration de la plasticité cérébraleaprès AVC : données de l'imagerie

J.-C. Baron a,b

aInserm U894, université Paris 5, Centre de Psychiatrie et Neurosciences, 2, ter rued'Alésia, 75014 Paris, FrancebUniversity of Cambridge, Department of Clinical Neurosciences, Addenbrooke'sHospital, Box 83, Cambridge, CB2 2QQ, UK

The present paper will focus on what

KeywordsStrokePlasticityRecoveryFunctional imagingfMRI

Mots clésAccident vasculairecérébralPlasticité

INTRODUCTION

Following the acute phase of strokewhere major physiological events suchas salvage of the ischemic penumbraunderlie rapid clinical changes and deter-mine final tissue outcome, the subacute-to-chronic phase is characterized by func-tional recovery that evolves despite resid-ual damage. This can involve two entirelydifferent processes:

Récupération

� Imagerie fonctionnellecérébraleImagerie par résonancemagnétique

developing new behavioral strategiesto circumvent the deficit (e.g., usingthe unaffected arm after hemiparesis),a suboptimal type of spontaneousrecovery seen in animals but also inhumans if left unattended;

Corresponding author.J.-C. Baron,Inserm U894, universitéParis 5, Centre dePsychiatrie etNeurosciences, 2, ter rued'Alésia, 75014 Paris,France.E-mail address:[email protected]

genuine restoration – to a variabledegree – of the lost function, whichtherapy aims to enhance.

In the present article, primarily the latterwill be addressed as it is the most usefuland the one subtended by neuronalplasticity.Here we will use a broad definition ofplasticity as any change in neuronal con-nections, circuitry and large-scale net-works, be it at the molecular, cellular orfunctional level, that occurs subsequentto focal brain damage. The ultimate aim ofstudying, and hopefully understanding,the phenomenon of plasticity is toenhance adaptive plasticity, and curbmaladaptive plasticity, by therapeuticmeans in order to improve functionaloutcome.

insights into post-stroke plasticity mecha-nisms have been gained from functionalimaging. Furthermore, we will focus onmotor recovery, because hemiparesis,and particularly hand motor deficit, isnot only the most disabling sequelae ofstroke in functional terms, but also themost frequent (occurring in nearly 50%of stroke survivors) and the slowest torecover. Finally, we will focus on adult-onset stroke, where the type and degreeof plasticity, and hence the issues facingplasticity research, are entirely differentfrom childhood-onset stroke, where hard-wired connections have not crystallizedyet, and hence major changes in brainorganization can take place, such as tak-ing-over of the damaged crossed cortico-spinal tract (CST) by the opposite-sideuncrossed CST, clearly documentedusing functional imaging and single-pulseTMS (Stoeckel and Binkofski, 2010).Thus, children can enjoy very good func-tional outcome despite sometimes mas-sive unilateral brain damage, althoughare at risk of (re)emergence of themasked paralysis in case of recurrentstroke affecting the other hemisphere.Although the bulk of recovery takes placewithin the first month, it usually continuesat a slower pace over several months andin some patients several years. This sim-ple clinical observation suggests that avariety of mechanisms underlie restora-tion of function after stroke.

© 2012 Elsevier Masson SAS. All rights reserved.doi:10.1016/j.praneu.2012.01.008

Page 2: Mechanisms of functional recovery after stroke: Insights from imaging

Pratique Neurologique – FMC 2012;3:160–166 Cerebral plasticity; post-stroke

BRAIN PLASTICITY: A BRIEF SUMMARYOF BASIC MECHANISMS

Physiological plasticityBrain plasticity is a normal process throughout develop-ment, acting to incorporate experience-based knowl-edge (i.e., learning) into neuronal connections in orderto enhance adapted behavior. These processes aremainly based on experience-dependent glutamatergicsynaptic growth and strengthening such as long-termpotentiation (LTP). Synaptic pruning that occurs duringadolescence stabilizes the person's own adaptativebehaviors. During development, acquisition of specificskills such as motor skills, e.g. playing the violin, istherefore associated with growth of synaptic connec-tions relevant to the particular skill, ultimately translatinginto enlarged cortical volume for finger representationsin the primary motor cortex (M1), which is larger theyounger the age at start of skill acquisition. Althoughsynaptic plasticity is less prominent in adulthood, evi-dence from functional imaging of synaptic strengthchanges with experience is now overwhelming, as arechanges in distribution of large-scale network activationduring acquisition of a motor skill until it becomes nearlyautomatic (Karni et al., 1995). Although such plasticpotential declines with aging, it remains active until veryold age in man.

Brain plasticity after focal damageIt was widely believed until about 30 years ago that theadult central nervous system had little or no ability forreorganization and self-repair following focal injury, i.e. isnot plastic. To explain recovery of neurological deficitsfollowing e.g., stroke, processes such as vanishing ofperi-lesional edema/pressure, recovery from diaschisis(an acute functional depression of neural ensemblesremote from but disconnected by the damaged area),reperfusion of the affected territory, and use of alternativecortical representations of the same function and of built-in alternative pathways such as the uncrossedCST, werepostulated (Baron, 2005). Of these mechanisms, onlyreperfusion of the affected territories, provided it occurswithin a few hours of ischemic stroke onset, has a provenmajor role in functional recovery, by rescuing the ischemicpenumbra.Delayed reperfusion does not improve clinicalfunction, and in fact almost consistently occur spontane-ously as an epiphenomenon (Baron, 2005). Pressurefrom vasogenic edema or mass effect from hematomado not cause peri-lesional ischemia or fiber dysfunctionexcept with very large lesions such as malignant MCAinfarction. Diaschisis has been documented in the acutestage of stroke, but in specific cases only such as withcrossed cerebellar diaschisis, which can indeed regressbut seems not to carry direct functional consequences,and thalamo-cortical diaschisis, where regression of

cortical glucose and oxygen hypometabolismmay underlie part of the observed cognitive recovery(Baron, 2005). Marked reductions in glucose/oxygenmetabolism do occur bilaterally after hemisphericstroke, but develop only after a few days and in propor-tion to the volume of the infarct, have no clear clinicalcounterpart and probably simply reflect extensive Wal-lerian degeneration. Regarding the opposite-hemi-sphere uncrossed CST, it does not appear to playany significant role in adult post-stroke hand recovery,as documented by single-pulse TMS (Gerloff et al.,2006). However, unmasking of (latent) cortical repre-sentations is an interesting possibility following corticaldamage (see below).Brain plasticity after focal stroke encompasses innumer-able mechanisms that can affect essentially the entirebrain, i.e., not only the peri-lesional areas but alsoremote areas within the same hemisphere, contralateralhemisphere (via trans-hemispheric connections), cere-bellum and spinal cord (Cramer, 2008; Rossini et al.,2003). These include:

� at the molecular level, changes such as with generegulation and expression, release of neurotransmit-ters, membrane components, receptor density andaffinity, excitatory/inhibitory (i.e., GABA/glutamate)balance;

at the cellular level, such as axonal sprouting fromremaining fibres of a partly damaged pathway, synap-tic growth or loss, experience-dependent changes insynaptic strength, dendritic arborization, and cell pro-liferation (neurogenesis, angiogenesis) with morpho-logical and functional changes (e.g., microglial andastroglial activation with consequent release of cellsignaling molecules);

at the neural circuitry level, such as unmasking oflatent connections, re-distribution of strength of con-nections and of activation within a pre-existing path-way, and recruitment of alternative pathwaysfunctionally homologous but anatomically distinct fromthe damaged ones (e.g., non-pyramidal corticospinalpathways, so-called vicariation).

Of note, individual factors, including genetic build-up(such as BDNF polymorphism), age, development (edu-cation) and environment-dependent, and co-morbiditiessuch as previous stroke, white matter ischemic damageand depression, play a key role in plasticity, accountingfor a large part of the observed variance in individualfunctional outcome despite same damage.Importantly, although the above plastic changes can beadaptative, i.e., serving to restore lost function, they canalso be maladaptive, i.e., hindering restoration of func-tion (Cramer, 2008). Distinguishing these two types ofplasticity is obviously paramount with regards to thera-peutic implications. The role of rehabilitation in modulat-ing some of these changes is crucial. For instance, bydriving experience-dependent synaptic mechanisms,active rehabilitation of the affected arm such as

161

Page 3: Mechanisms of functional recovery after stroke: Insights from imaging

J-C BaronCerebral plasticity; post-stroke

constraint-induced therapy (CIT), has been shown in themonkey with experimental infarction of the M1 handarea to prevent or even reverse "learned non-use'', i.e., a maladaptive plasticity such that the spontaneousexclusive use of the unaffected arm causes regressionof hand area connections in the affected side M1 homo-nculus (Nudo et al., 1996).

BRAIN PLASTICITY AFTER STROKE ANDRECOVERY OF HAND MOTOR DEFICIT:GENERAL NOTIONS

Functional organization of the motorsystem: brief overviewAlthough moving the fingers may appear a simple func-tion, the motor system has extreme organizational com-plexity. It is organized as a network involving not onlyM1, but also the premotor cortex (PM) and the supple-mentary motor area (SMA), which all share somatotopicrepresentation. These areas all have functionally distinctsubdivisions, such as the anterior and posterior M1(BA4a and 4p), the dorsal and ventral PM (PMd andPMv), the rostral and caudal SMA (pre-SMA and SMA-proper). The motor system can be considered "cogni-tive'' inasmuch as it has extensive connections with theprefrontal, parietal and cingulate cortices and the insula(Calautti and Baron, 2003). The SMA is connected withhigher order areas, mainly the prefrontal cortex, and isinvolved in motor act programming, while the PM isinvolved in bimanual motor coordination. M1 contributesonly part of the CST fibres, which also originate from thePM and the post-central cortex (S1). However, distalupper limb function is subtended only by the pyramidaltract, and lesions of the PM result in proximal movementdisturbance only, mainly of an apraxic type. There areconsiderable connections between M1 and S1, contrib-uting to sensori-motor integration, as well as transcal-losal (mainly inhibitory), contributing to bimanualcoordination and proximal limb automatic movements.Other motor systems exist, such as the reticulospinaltract, which is bilaterally organized. In addition, themotor cortices have strong connections with the ipsilat-eral basal ganglia and thalamus as well as the contra-lateral cerebellum.Within M1, there is a mosaic-like representation of upperlimb muscles with extensive intracortical connectivity forfine movement coordination. Functional imaging hasdemonstrated plasticity of the M1 representations innormal adults as they perform – and therefore learn –complex motor tasks (Karni et al., 1995), and this phys-iological cortical plasticity has considerable importancein recovery from focal injury (see below). The uncrossedCST has only marginal physiological function, as shownby TMS studies in normal subjects where only proximalresponses are seen with intense stimuli only.

162

Influence of corticospinal tract damage inpost-stroke motor recoveryOne of the most important factors influencing post-stroke motor recovery is integrity of the pyramidal motoroutput system. In both animals and man, extensive/complete lesions of the precentral gyrus induce perma-nent loss of prehension, while good recovery generallytakes place after partial lesions. Accordingly, theabsence of evoked motor potentials (MEPs) on sin-gle-pulse TMS of ipsilesional M1 in the subacute stageafter stroke is predictive of poor motor recovery (Rossiniet al., 2003). Conversely, detection of MEPs is a markerof recovery, while the severity of CST damage predictsthe response to rehabilitation therapy (Riley et al., 2011).It is estimated that the survival of at least 1/5 of thepyramidal fibers following M1 or CST stroke is neces-sary, though not sufficient, to ultimately ensure restitu-tion of fractionated hand finger movement. Recentdiffusion tensor imaging studies of CST Walleriandegeneration following cortical or subcortical strokehave strengthened these notions (Riley et al., 2011).Hence, within-CST reorganization is a major candidatefor functional recovery of motor control. However, theextent of CST damage accounts for only a fraction of thevariance in motor recovery (Riley et al., 2011), indicatingthat additional factors have amajor role in this process, i.e. for the same degree of CST damage the outcome willdiffer from subject to subject as a function of individualfactors (see above). The role of externally- or self-deliv-ered rehabilitation, as well as its type and intensity, alsoplay an important role. Understanding these factors maylead to new strategies for individualized therapy andhence better stroke outcome for a given damage.

BRAIN PLASTICITY AND MOTORRECOVERY AFTER STROKE: INSIGHTSFROM FUNCTIONAL IMAGING

Thanks to the advent of functional imaging, i.e. brainactivity mapping during behavioral tasks, the biologicalbasis of human post-stroke motor recovery has madeconsiderable progress (Baron et al., 2004; Calautti andBaron, 2003; Cramer, 2008; Dimyan and Cohen, 2011;Rossini et al., 2003). These techniques mainly includePositron Emission Tomography (PET), and morerecently functional Magnetic Resonance Imaging(fMRI), transcranial magnetic stimulation (TMS) andmagneto-encephalography (MEG).

Functional imaging in normal subjectsFunctional imaging studies in normal subjects haveelucidated the functions of the different motor areasby varying the experimental conditions, e.g. rate, force,modality of execution (i.e., motor preparation, initiation,

Page 4: Mechanisms of functional recovery after stroke: Insights from imaging

Pratique Neurologique – FMC 2012;3:160–166 Cerebral plasticity; post-stroke

execution, learning, mental imagery, internally gener-ated vs. externally paced movement, active vs. pas-sive), and established the somatotopy within M1, PMand SMA (Calautti and Baron, 2003). While a simpledominant hand task typically involves the contralateralSM1, PMd and SMA-proper and ipsilateral cerebellum, itwill tend to involve the same areas bilaterally if per-formed by the non-dominant hand, as do also forcefulor complex tasks which in addition may also involveparietal, prefrontal and striatal areas. Even healthyaging is associated with significant, albeit minor,changes in network activation pattern during simplemotor tasks.

Findings in stroke patientsSummarizing across over a hundred functional imagingstudies of post-stroke hand motor recovery, most cross-sectional and some longitudinal, a number of consistentmajor findings have emerged (Buma et al., 2010; Cal-autti and Baron, 2003; Cramer, 2008).

General patternThe main observation from both cross-sectional andlongitudinal studies is that, regardless of the strokelocation, motor recovery is best if the normal patternof network activation for the particular task studied is re-established (Calautti and Baron, 2003). However, evenin the situation of fully recovered normal pattern, signifi-cant changes in connection strength between the maincortical motor areas can be found as ultimate evidenceof underlying plasticity (Grefkes and Fink, 2011; Sharmaet al., 2009). The magnitude of these connectivitychanges correlates with residual motor deficit, suchas assessed by maximal index-thumb tapping rate orother indices of dexterity (Sharma et al., 2009), indicat-ing these changes are adaptative.

Ipsilesional motor cortex

The changes observed in ipsilesional M1 activation,together with their therapeutic implications, differwhether the infarct directly involves M1 or not. Afterpartial damage to M1, peri-infarct activation is a consis-tent observation (Cramer et al., 1997; Jaillard et al.,2005), in agreement with the above-described monkeystudies (Nudo et al., 1996). This peri-infarct activationincreases over time and probably underlies clinicalimprovement, further highlighting the key role of penum-bral salvage in the acute stage. This phenomenon likelyreflects disinhibition/unmasking of pre-existing but nor-mally "silent'' representations in the surround of thelesion (redundancy), and/or progressive activation ofmotor representations normally not devoted to the lostfunction, i.e. supporting more proximal muscles(vicariation).

After subcortical stroke affecting the CST (hence result-ing in M1 de-efferentation), a constant observation in theearly stages after stroke is overactivation of ipsilesionalM1 (Chollet et al., 1991), which tends to return towardsnormal levels over time as motor performance improves(Calautti et al., 2001a). Regardless whether ipsilesionalM1 over-recruitment reflects redundancy or vicariation,it should result in increased output down the remainingCST fibers and hence help to produce the intendedmovement. In turn, neuronal hyperactivity may inher-ently drive plastic changes through experience-depen-dent synaptic strengthening, which would account forthe gradual return of M1 activation towards normal asrecovery proceeds. Congruent with this, consistent dis-placement of activation peak within M1 has beenreported, suggesting reorganization of the motor repre-sentations within the homonculus (Calautti and Baron,2003). It logically follows that stimulating ipsilesional M1after stroke should further enhance motor recovery.Accordingly, intensive affected arm training and excit-atory repetitive TMS (rTMS) onto ipsilesional M1 havebeen applied with significant benefit. In the early stagespost-stroke, the higher ipsilesional M1 activation is, thegreater the gains from ipsilesional rTMS (Ameli et al.,2009). However, in the chronic stage ipsilesional M1activation does not correlate with performance (Calauttiet al., 2007), and the lower it is, the higher the chancethat stimulatory therapy is effective and results in theintended increases in M1 activation. Thus, increasingipsilesional M1 activation is an aim of, and a marker for,effective therapy in the chronic stage. As will be seenbelow, maladaptive overactivation of contralesional M1may be the culprit, through transcallosal inhibition ofipsilesional M1.

Non-primary motor areas

In both cortical and subcortical strokes, overactivation ofthe non-primary motor network bilaterally is anotherconsistent finding in the early stage. This tends to abateas recovery proceeds, returning to normal in those whorecover best but remaining abnormally elevated, partic-ularly in contralesional M1 and PMd, in those with worserecovery (Calautti et al., 2001a). Overactivation of PMdand SMA probably represents enhanced input downnon-pyramidal CST fibers, driven by higher order areas,and contributing to CST as well as non-CST systemsoutput. In addition, contralesional PMd activation maysubtend recovery of/compensation by non-force com-ponents of dexterity such as speed and rythmicity andcoordination of complex movement, as suggested byboth fMRI (Calautti et al., 2010a) and rTMS studies(Johansen-Berg et al., 2002).

Contralesional motor cortex

Overactivation of contralesional M1 following hemi-spheric stroke is an intriguing phenomenon that has

163

Page 5: Mechanisms of functional recovery after stroke: Insights from imaging

J-C BaronCerebral plasticity; post-stroke

attracted considerable interest because of both its elu-sive pathophysiology and its potential therapeutic impli-cations (Calautti et al., 2007; Stoeckel and Binkofski,2010). In recovering strokes, it is present initially buttends to abate over time, with "refocusing'' of activationtowards the ipsilesional M1 (Calautti et al., 2001b; Gerl-off et al., 2006). Importantly, this temporal pattern hasalso been observed in fMRI studies of rats subjected tocortical M1 stroke, showing early contralesional S1 acti-vation with affected side median nerve stimulation, withsubsequent refocusing towards the affected side.Clearly, contralesional M1 activation after stroke doesnot represent activity of the contralesional uncrossedCST, as:

16

the higher this activation the worse the hand recovery(Calautti et al., 2007; Johansen-Berg et al., 2002);

all direct functional connections to hand muscles afterrecovered stroke originate from ipsilesional M1, asshown by EEG-EMG coherence (Gerloff et al., 2006);

and as indicated above, single-pulse TMS of contrale-sional M1 does not evoke any affected hand response(Gerloff et al., 2006).

Likewise, contralesional M1 activation is not significantlyassociated with the presence of contralateral synkine-siae (i.e., mirror movements). One possible interpreta-tion is that it represents loss of inhibition from damagedipsilesional M1. However, although this phenomenon iswell documented using TMS (Rossini et al., 2003), itwould be inconsistent with the situation in subcorticalstroke where ipsilesional M1 is not only present but oftenoveractive. A different hypothesis is that it reflects thepatient's effort to execute the required task, in analogywith normal subjects executing complex tasks (Calauttiand Baron, 2003; Gerloff et al., 2006). In this scenario,contralesional M1 activation would result from increaseddrive from upstream PM and SMA and in turn fromhigher order prefrontal cortex, aiming to send sufficientoutput down the damaged CST. This hypothesis is sup-ported by reports of significant correlation between thedegree of CST damage and the magnitude of contrale-sional M1 activation. Not mutually exclusive, contrale-sional M1 overactivation likely also reflects maladaptiveplasticity. This is supported not only by its negativecorrelation with affected hand performance (Calauttiet al., 2007; Johansen-Berg et al., 2002), but also bythe fact that contralesional M1 inhibition by means ofrTMS tends to improve affected hand motor functions(Lindenberg et al., 2010). This aberrant maladaptiveactivity would exert its functional effects via transcallosalinhibition of ipsilesional M1 (Rossini et al., 2003).

Hemispheric balance of motor cortexactivation

Based on the above, a further conceptual shift is that,rather than considering each side separately, it is thehemispheric balance within this transcallosally

4

connected system that controls performance. This canbe conveniently studied using the Laterality Index (LI), amathematical formula whereby +1 reflects exclusivelyipsilesional (i.e., physiological for simple finger move-ment) and -1 exclusively contralesional activation(Cramer et al., 1997). Studies have shown that the lowerthe LI (i.e., the more unphysiological), the worse theclinical outcome (Calautti et al., 2007; Johansen-Berget al., 2002). Accordingly, the LI returns towards morephysiological values as performance improves over time(Calautti et al., 2010b). The logical speculation then is thatinterventions aiming to restore this balance towardsmorephysiological levels, either by stimulation of ipsilesionalM1 or via inhibition of contralesional M1, or both, shouldresult in improved motor performance. There is accumu-lating evidence supporting this notion, from studies forinstance involving CIT, rehab – or robot – driven activearm training, rTMS (Lindenberg et al., 2010) or pharma-cological agents suchasfluoxetine (Pariente et al., 2001).Crucially, this effect can take place even in the chronicstageof stroke,monthsor yearsafter themotor deficit hasstabilized, indicating that the balance between adaptiveand maladaptive plasticity can reach a plateau which isfunctionally suboptimal but can be shifted towards betteroutcome by appropriate interventions. Another key pointis that to work optimally, such interventions must beassociated with active motor training/rehabilitation prob-ably because these induceeffective use-dependent plas-ticity, therefrom enhanced by other means.

Other cortical areasAnother interesting finding regards the recruitment ofareas normally not engaged in task execution, such asprefrontal, posterior parietal, anterior cingulate and insu-lar. This involvement might reflect compensatory cogni-tive strategies, e.g. attentional or visuo-spatial, whatever`simple' the task may seem. The decreasing recruitmentover time of some of these areas (Calautti et al., 2001a)suggests that recourse to such strategies graduallybecomes less necessary to produce the motor behavior.

CONCLUDING COMMENT

Despite the abovemajor breakthroughs afforded by func-tional brain imaging in understanding plasticity afterstroke and how it underpins functional recovery, muchremains to be done to clarify the exact time-course ofchanges from the acute to the chronic stage, the finemechanisms underlying recovery of the various aspectsof motor behavior and dexterity, and their intimate under-lying processes at the neuronal ensemble, single-cell,synaptic and molecular levels. This effort will be neces-sary to fulfill the ultimate goal to deliver to each strokevictim the best possible care, and hence reach optimalfunctional outcome given each person's brain damage.

Page 6: Mechanisms of functional recovery after stroke: Insights from imaging

Points essentiels

� Il existe un énorme potentiel de plasticité céré-brale après accident vasculaire cérébral (AVC)chez l'adulte.

� Mieux comprendre les mécanismes biologi-ques qui sous-tendent la plasticité cérébrale,et plus particulièrement pourquoi la récupéra-tion fonctionnelle après AVC est si variable d'unpatient à l'autre, permettra de développer denouvelles approches thérapeutiques visantà garantir la meilleure récupération possible.

� La plasticité cérébrale est un phénomène phy-siologique qui emprunte de multiples proces-sus depuis le niveau moléculaire jusqu'auniveau des cartes neuronales. Ces processussont exacerbés après un AVC, mais peuventêtre adaptatifs ou maladaptatifs.

� L'imagerie fonctionnelle cérébrale a mis en évi-dence un certain nombre de cibles pour desinterventions visant à améliorer la récupérationmotrice, y compris au stade chronique.

� Les futures études devront affiner ces connais-sances et mettre en oeuvre des essais clini-ques de façon à transposer celles-ci dans lapratique quotidienne.

Pratique Neurologique – FMC 2012;3:160–166 Cerebral plasticity; post-stroke

DISCLOSURE OF INTEREST

The author declares that he has no conflicts of interestconcerning this article.

REFERENCES

Ameli M, Grefkes C, Kemper F, Riegg FP, Rehme AK, Karbe H, et al.Differential effects of high-frequency repetitive transcranial mag-

netic stimulation over ipsilesional primary motor cortex in corticaland subcortical middle cerebral artery stroke. Ann Neurol2009;66:298–309.

Baron JC. Stroke research in the modern era: images versus dogmas.Cerebrovasc Dis 2005;20:154–63.

Baron JC, Cohen LG, Cramer SC, Dobkin BH, Johansen-Berg H,Loubinoux I, et al. Neuroimaging in stroke recovery: a position

paper from the first international workshop on neuroimaging andstroke recovery. Cerebrovasc Dis 2004;18:260–7.

Buma FE, Lindeman E, Ramsey NF, Kwakkel G. Functional neuroim-

aging studies of early upper limb recovery after stroke: a systematicreview of the literature. Neurorehabil Neural Repair 2010;24:589–608.

Calautti C, Baron JC. Functional neuroimaging studies of motor recov-

ery after stroke in adults: a review. Stroke 2003;34:1553–66.Calautti C, Jones PS, Guincestre JY, Naccarato M, Sharma N, Day DJ,

et al. The neural substrates of impaired finger tapping regularity

after stroke. Neuroimage 2010a;50:1–6.Calautti C, Jones PS, Naccarato M, Sharma N, Day DJ, Bullmore ET,

et al. The relationship between motor deficit and primary motorcortex hemispheric activation balance after stroke: longitudinal

fMRI study. J Neurol Neurosurg Psychiatry 2010b;81:788–92.Calautti C, Leroy F, Guincestre JY, Baron JC. Dynamics of motor

network overactivation after striatocapsular stroke: a longitudinal

PET study using a fixed-performance paradigm. Stroke 2001a;32:2534–42.

Calautti C, Leroy F, Guincestre JY, Marie RM, Baron JC. Sequential

activation brain mapping after subcortical stroke: changes in hemi-spheric balance and recovery. Neuroreport 2001b;12:3883–6.

Calautti C, Naccarato M, Jones PS, Sharma N, Day DD, Carpenter AT,et al. The relationship between motor deficit and hemisphere acti-

vation balance after stroke: A 3 T fMRI study. Neuroimage2007;34:322–31.

Chollet F, DiPiero V, Wise RJ, Brooks DJ, Dolan RJ, Frackowiak RS.

The functional anatomy of motor recovery after stroke in humans: astudywith positron emission tomography. AnnNeurol 1991;29:63–71.

Cramer SC. Repairing the human brain after stroke: I. Mechanisms of

spontaneous recovery. Ann Neurol 2008;63:272–87.Cramer SC, Nelles G, Benson RR, Kaplan JD, Parker RA, Kwong KK,

et al. A functional MRI study of subjects recovered from hemiparetic

stroke. Stroke 1997;28:2518–27.Dimyan MA, Cohen LG. Neuroplasticity in the context of motor reha-

bilitation after stroke. Nat Rev Neurol 2011;7:76–85.Gerloff C, Bushara K, Sailer A, Wassermann EM, Chen R, Matsuoka T,

et al. Multimodal imaging of brain reorganization in motor areas ofthe contralesional hemisphere of well recovered patients aftercapsular stroke. Brain 2006;129:791–808.

Grefkes C, Fink GR. Reorganization of cerebral networks after stroke:new insights from neuroimaging with connectivity approaches.Brain 2011;134:1264–76.

Jaillard A, Martin CD, Garambois K, Lebas JF, Hommel M. Vicariousfunction within the human primary motor cortex? A longitudinalfMRI stroke study Brain 2005;128:1122–38.

Johansen-Berg H, Rushworth MF, Bogdanovic MD, Kischka U, Wima-laratna S, Matthews PM. The role of ipsilateral premotor cortex inhand movement after stroke. Proc Natl Acad Sci U S A2002;99:14518–23.

Karni A, Meyer G, Jezzard P, Adams MM, Turner R, Ungerleider LG.Functional MRI evidence for adult motor cortex plasticity duringmotor skill learning. Nature 1995;377:155–8.

Lindenberg R, Renga V, Zhu LL, Nair D, Schlaug G. Bihemisphericbrain stimulation facilitates motor recovery in chronic strokepatients. Neurology 2010;75:2176–84.

Nudo RJ, Wise BM, SiFuentes F, Milliken GW. Neural substrates forthe effects of rehabilitative training onmotor recovery after ischemicinfarct. Science 1996;272:1791–4.

Pariente J, Loubinoux I, Carel C, Albucher JF, Leger A, Manelfe C,

et al. Fluoxetine modulates motor performance and cerebral acti-vation of patients recovering from stroke. Ann Neurol 2001;50:718–29.

165

Page 7: Mechanisms of functional recovery after stroke: Insights from imaging

J-C BaronCerebral plasticity; post-stroke

Riley JD, Le V, Der-Yeghiaian L, See J, Newton JM, Ward NS, et al.

Anatomy of stroke injury predicts gains from therapy. Stroke2011;42:421–6.

Rossini PM, Calautti C, Pauri F, Baron JC. Post-stroke plastic reorga-

nisation in the adult brain. Lancet Neurol 2003;2:493–502.

166

Sharma N, Baron JC, Rowe JB. Motor imagery after stroke: relating

outcome tomotor network connectivity. AnnNeurol 2009;66:604–16.Stoeckel MC, Binkofski F. The role of ipsilateral primary motor cortex in

movement control and recovery from brain damage. Exp Neurol

2010;221:13–7.