Stroke rehabilitation using noninvasive cortical stimulation: motor deficit

Review

THEMED ARTICLE y Stroke For reprint orders, please contact [email protected]

Stroke rehabilitation using noninvasive cortical stimulation: motor deficit Expert Rev. Neurother. 12(8), 949–972 (2012)

Samar S Ayache‡1,2, Wassim H Farhat‡1,2, Hela G Zouari1,2,3, Hassan Hosseini1,4, Veit Mylius1,5, and Jean-Pascal Lefaucheur*1,2 Université Paris-Est-Créteil, Faculté de Médecine, EA 4391, Créteil, France 2 Assistance Publique, Hôpitaux de Paris, Hôpital Henri Mondor, Service de Physiologie, Explorations Fonctionnelles, Créteil, France 3 CHU Habib Bourguiba, Service d’Explorations Fonctionnelles, Sfax, Tunisie 4 Assistance Publique, Hôpitaux de Paris, Hôpital Henri Mondor, Service de Neurologie, Créteil, France 5 Department of Neurology, Philipps University Marburg, Marburg, Germany *Author for correspondence: Tel.: +33 1 4981 2694 Fax: +33 1 4981 4660 [email protected] 1

Authors contributed equally.

‡

www.expert-reviews.com

Noninvasive cortical stimulation (NICS) has been used during the acute, postacute and chronic poststroke phases to improve motor recovery in stroke patients having upper- and/or lowerlimb paresis. This paper reviews the rationale for using the different NICS modalities to promote motor stroke rehabilitation. The changes in cortical excitability after stroke and the possible mechanisms of action of cortical stimulation in this context are outlined. A number of open and placebo-controlled trials have investigated the clinical effect of repetitive transcranial magnetic stimulation (rTMS) or transcranial direct current stimulation (tDCS) of the primary motor cortex in patients with motor stroke. These studies attempted to improve motor performance by increasing cortical excitability in the stroke-affected hemisphere (via high-frequency rTMS or anodal tDCS) or by decreasing cortical excitability in the contralateral hemisphere (via low-frequency rTMS or cathodal tDCS). The goal of these studies was to reduce the inhibition exerted by the unaffected hemisphere on the affected hemisphere and to then restore a normal balance of interhemispheric inhibition. All these NICS techniques administered alone or in combination with various methods of neurorehabilitation were found to be safe and equally effective at the short term on various aspects of poststroke motor abilities. However, the long-term effect of NICS on motor stroke needs to be further evaluated before considering the use of such a technique in the daily routine management of stroke. Keywords: cortical excitability • motor deficit • neuromodulation • stroke • theta burst stimulation • transcranial direct current stimulation • transcranial magnetic stimulation

Stroke is the leading cause of long-term disability among adults in industrialized countries [1,2] . Despite recent progress in rehabilitation techniques, recovery of motor function after stroke is usually incomplete [2,3] . More than 60% of stroke survivors suffer from persistent neurological deficits and impaired dexterity that cause significant impact on their daily living activities (dressing, eating and self-care) and independence [3] . Thus, the development of new strategies to promote stroke recovery in addition to the classical rehabilitation methods is desirable. Noninvasive cortical stimulation (NICS) is an appealing technique for this purpose since several investigations have proved its potential role in activating and modulating cortical excitability in humans, particularly in association with motor training [4,5]. In this paper, the authors review the current knowledge about the efficacy of NICS techniques in promoting poststroke motor recovery. The PubMed database was searched for relevant articles concerning the application of any 10.1586/ERN.12.83

NICS technique in stroke. The terms ‘transcranial magnetic,’ ‘theta burst,’ ‘paired associative,’ or ‘direct current’ and ‘stimulation,’ ‘motor,’ ‘stroke’ and ‘patient’ were used. Reviews and editorials were excluded. All study designs were accepted, including open label studies. On 1 May 2012, 66 ‘therapeutic’ studies were retained for analysis, excluding purely neurophysiological evaluation of single NICS protocols applied in stroke patients (Tables 1 & 2) [6–8]. Dysfunction of interhemispheric connectivity following motor stroke

In the acute phase of stroke, the initial motor deficit relates to the dysfunction or disruption of the motor corticospinal output, as assessed by the recording of motor-evoked potentials (MEPs) to transcranial magnetic stimulation (TMS) [9] or MRI with diffusion tensor imaging tractography [10] . The preservation or recovery of corticospinal tract integrity allows a good outcome and correlates to motor function [11–13] ,

© 2012 Expert Reviews Ltd

ISSN 1473-7175

949

950

Number and type of Target, coil type patients Control Stimulation Number of pulses per Results condition frequency session and number and intensity of sessions

600 pulses, one session


Contralesional Vertex 1 Hz, 100% 12 (postacute stimulation RMT subcortical MCA stroke: M1, F8 2 years after stroke)

Takeuchi et al. (2005)

Boggio et al. (2006)

Fregni et al. (2006)

Kirton et al. (2008)

1 Hz, 90% RMT

Contralesional Tilted M1, F8 active coil


Improvement of pinch acceleration, but not of maximum pinch force, immediately after rTMS, lasting less than 30 min

1500 pulses, one session (followed by motor training)


100 pulses, 14 sessions (circular coil)

1200 pulses, eight sessions

[58]

[59]

Increase in the excitability of the affected motor cortex, improvement in acceleration of the affected hand and enhancement of the effect of motor training on pinch force

[69]

[70]

[63]

[68]

[54]

[53]

Improvement of movement kinematics of the affected hand. Enhancement of the effect of motor training on pinch force, lasting at least 7 days after rTMS and motor training. Bilateral change in motor cortex excitability (MEP size)

Reduction of spasticity, without change in movement and behavioral scores

Improvement of motor performance (MAUEF, grip strength, PPT, HRFT, IHM) of the affected hand in children with chronic stroke

1200 pulses, five sessions Improvement of motor performance (JJT, sRT, cRT, PPT), increasing over time during the treatment period and lasting at least 2 weeks after treatment completion. No change in cognitive functions

1200 pulses, two sessions Improvement of motor performance. No change in spasticity and mood



Ref.

AI: Activity index scale; AMT: Active motor threshold; ARAT: Action research arm test; BBT: Box and Block test; BI: Barthel index; CIT: Constraint-induced therapy; cRT: Choice reaction time; cTBS: Continuous theta burst stimulation; F8: Figure-of-eight coil; FAC: Functional ambulatory category; FMA: Fugl–Meyer assessment; FMA-LL: Fugl–Meyer assessment for the lower limbs; FMA-UL: Fugl–Meyer assessment for the upper limbs; fMRI: Functional magnetic resonance imaging; FNMS: Functional neuromuscular stimulation; FT: Finger tapping frequency; HG: Handgrip force; HRFT: Halstead–Reitan finger tapping; ICF: Intracortical facilitation; ICI: Intracortical inhibition; IHI: Interhemispheric inhibition; IHM: In-hand manipulation; IPAS: Interventional paired associative stimulation; ISI: Interstimulus interval; iTBS: Intermittent theta burst stimulation; JTT: Jebsen–Taylor hand function test; KG: Key-grip force; M1: Primary motor cortex; MAL: Motor activity log; MAS: Modified Ashworth scale; MAUEF: Melbourne assessment of upper extremity function; MBI: Modified Barthel index; MCA: Middle cerebral artery; MEP: Motor-evoked potential; MI: Motricity index; MRC: Medical Research Council score; mRS: Modified Rankin scale; NIHSS: National Institute of Health score scale; OT: Occupational therapy; PPT: Purdue pegboard test; PT: Physical therapy; RMT: Resting motor threshold; rTMS: Repetitive transcranial magnetic stimulation; SIS: Sensibility impairment score; SMA: Supplementary motor area; sRT: Simple reaction time; SSS: Scandinavian stroke scale; tDCS: Transcranial direct current stimulation; TEMPA: Test évaluant la performance des membres supérieurs des personnes âgées; WFMT: Wolf motor function test.

1 (chronic cortical stroke: 8 months after stroke)

1 Hz, 90% RMT

20 (chronic subcortical Contralesional Tilted active coil stroke: >6 months after M1, F8 stroke)


Málly et al. 64 (chronic cortical (2008) stroke: several years after stroke)

1 Hz, 100% RMT

Contralesional Sham coil M1, F8

10 (chronic cortical or subcortical stroke: 12 months after stroke)

Kakuda et al. (2010)

24 (postacute/chronic cortical or subcortical stroke: 0.5–58 months after stroke)

1 Hz, 110– 120% RMT

Contralesional Tilted M1, F8 active coil

40 (postacute/chronic stroke: 1–50 months after stroke)

Emara et al. (2010)

Theilig et al. (2011)

1 Hz, 90% RMT

Contralesional None M1, F8

1 Hz, 110– 120% RMT

10 (10 bilateral 1 Hz/10 Hz rTMS) (chronic subcortical stroke: >6 months after stroke)

Contralesional Tilted 60 (chronic cortical or M1, F8 active coil subcortical stroke: >2 months after stroke)


Emara et al. (2009)

Improvement of functional status (AI) at 2 weeks after rTMS, for subcortical but not cortical strokes

Control Stimulation Number of pulses per Results condition frequency session and number and intensity of sessions

Low-frequency rTMS over the contralesional unaffected motor cortex: chronic stroke (cont.)

Study (year)

Table 1. Therapeutic studies using repetitive transcranial magnetic stimulation and theta burst stimulation in stroke patients with motor deficit (cont.).

Review Ayache, Farhat, Zouari, Hosseini, Mylius & Lefaucheur

Expert Rev. Neurother. 12(8), (2012)

Number and type of Target, coil patients type


www.expert-reviews.com

1 Hz, 90% RMT 1 Hz, 90% RMT

1 Hz, 90% RMT

1 Hz, 90% RMT



204 (chronic stroke: 5 ± 4.5 years after stroke)

14 (chronic stroke: 87 ± 48 months after stroke)

Contralesional None 9 (9 rTMS + tDCS; M1, F8 chronic subcortical stroke: >6 months after stroke)

Contralesional Tilted 24 (chronic stroke: >6 months after stroke) M1 (leg area), active coil F8




Wang et al. (2012)

600 pulses, ten sessions (followed by 30-min PT)

1000 pulses, one session (rTMS alone or combined with anodal tDCS over the affected hemisphere)

1200 pulses, 22 sessions (combined with 120-min OT), preceded by botulinum toxin injection in spastic upper limb muscles 4 weeks before

1200 pulses, 22 sessions (combined with 120-min OT)

10 (chronic stroke: Contralesional None >6 months after stroke) M1, F8

1 Hz (preceded 600 pulses (preceded by 600 priming pulses), one by 6 Hz rTMS priming), 90% session of stimulator output

Safety study. No conclusion for efficacy

[51]


Carey et al. (2008)

[145]

[62]

Improvement of a pinching motor training task for the paretic hand and decreased bimanual coordination, correlated to a reduced IHI from the unaffected to the affected hemisphere. Combined rTMS + tDCS protocol only enhanced the effect of motor training on pinch force Improvement in gait spatial symmetry, walking ability and motor function (FMA-LL), associated with a reduced asymmetry in MEP size

[84]

[76]

[64]

Ref.

Improvement of motor performance (MAL, FMA, but not WFMT) and reduction in spasticity (MAS) up to 4 weeks after rTMS

Improvement of motor performance (WFMT, FMA) up to 4 weeks after rTMS, regardless of the initial severity of motor deficit

1500 pulses, ten sessions Improvement of motor performance (trained tests: JTT, PPT, (followed or preceded by BBT, KG and untrained motor exercise: tip-pinch, powergrip) up to 3 months after active rTMS, associated with a 45-min PT) reduced IHI from the unaffected to the affected cortex, a short-lasting increase in RMT in the unaffected hemisphere and a long-lasting decrease in RMT in the affected hemisphere. In almost all cases, results were better when PT was performed after rTMS session

6-Hz-primed low-frequency rTMS over the contralesional unaffected motor cortex: chronic stroke

1 Hz, 90% RMT

Contralesional Tilted M1, F8 active circular coil

30 (chronic stroke: 6–88 months after stroke)

Avenanti et al. (2012)

Low-frequency rTMS over the contralesional unaffected motor cortex: chronic stroke (cont.)

Study (year)


Noninvasive cortical stimulation & motor stroke recovery

Review

953

954




11 (chronic stroke: 70 ± 40 months after stroke)


1 Hz (preceded by 6 Hz rTMS priming), 90% RMT

Tilted active coil Tilted active coil

Lesioned M1, F8

48 (acute MCA stroke: 5–15 days after stroke)

9 (9 sham; acute stroke: Lesioned M1, 6–29 days after stroke) F8

Khedr et al. (2010)

Sasaki et al. (2011) 10 Hz, 90% RMT

10 Hz, 90% RMT

3 Hz, 130% RMT or 10 Hz, 100% RMT

10 Hz, 90% RMT

3 Hz, 130% RMT

3 Hz, 120% RMT

1000 pulses, ten sessions Improvement in finger motor tasks for the paretic side with enhanced fMRI activation in the affected hemisphere (combined with motor compared with sham condition practice)

1000 pulses, five sessions Improvement of motor performance (HG, FT) compared with sham condition

[108]

[99]

[102]

Improvement of motor strength, clinical outcome (mRS, NIHSS) and affected motor cortex excitability. No significant difference between 3 and 10 Hz rTMS

750 pulses, five sessions

[97]

[101]

[103]

Improvement of motor performance (PPT, FT, NIHSS) and cortical excitability (AMT, MEP size), lasting up to 3 months, but less pronounced after 3 Hz rTMS over the affected hemisphere than 1 Hz rTMS over the unaffected hemisphere

Improvement of clinical outcome (SSS, NIHSS, BI)

1000 pulses, ten sessions Improvement of motor performance (MI-arm score, FMA-UL, BBT, grip strength) of the paretic upper limb immediately (combined with motor after rTMS. No change for the lower limb (MI-leg score, practice) FMA-LL) and for functional scores (FAC, MBI)

900 pulses, five sessions

300 pulses, ten sessions (combined with PT)

[86]

[85]

Ref.


Tilted active coil

Tilted active coil

Lesioned M1, F8

28 (postacute cortical or subcortical stroke: 1 year after stroke)

Malcolm et al. (2007)

20–25 Hz, 110–130% RMT

Tilted active coil

Lesioned M1, 7 (chronic cortical or subcortical MCA stroke: F8 1–5 years after stroke)

Lomarev et al. (2007)

10 Hz, 80% RMT

Tilted active coil

Lesioned M1, F8

15 (chronic cortical or subcortical stroke: 6–41 months after stroke)

Kim et al. (2006) 160 pulses, one session (combined with motor practice)


High-frequency rTMS over the lesioned motor cortex: chronic stroke

Study (year)



Review

955

956


9 (chronic stroke: 3–9 years after stroke)

Lesioned M1, F8 None

IPAS 0.1 Hz (35 ms ISI), 115% RMT

cTBS, 80% AMT cTBS, 80% AMT

Contralesional Sham coil M1 or S1, F8


12 (chronic MCA stroke: >1 year after stroke)


Meehan et al. (2011)

Talelli et al. (2012)

Sham coil

Sham coil

Lesioned M1, F8

Lesioned M1, F8

10 (chronic subcortical stroke: 7–86 months after stroke)

41 (Chronic MCA stroke: >1 year after stroke)

Ackerley et al. (2010)


iTBS, 90% AMT

iTBS, 90% AMT

No change in motor behavior or electrophysiological parameters (cortical excitability) for the paretic hand

Improvement of motor performance (reduced sRT) immediately after iTBS associated with increased motor cortex excitability (RMT, AMT, MEP size). No change in grip strength

No change in motor performance (PPT, JTT, grip and pinch force)

Improved motor performance of the paretic hand (movement time, velocity, acceleration, time to initiate movement, WFMT)

600 pulses, ten sessions (followed by PT)

No change in motor performance (PPT, JTT, grip and pinch force)

Improvement of grip-lift kinetics for the paretic hand with 600 pulses, one session (followed by training of a increased excitability of lesional M1, without any change in motor performance (ARAT) precision grip task)


300 pulses, ten sessions (followed by PT)

600 pulses, one session (combined with motor practice)

Improvement of grip-lift kinetics for the paretic hand with 600 pulses, one session (followed by training of a decreased excitability of contralesional M1, but deterioration of motor performance (ARAT) associated with a precision grip task) concomitant reduction of lesional M1 excitability


Improvement in gait parameters and MEP size

[114]

[112]

[111]

[114]

[113]

[112]

[111]

[96]

Ref.


Sham coil

Lesioned M1, F8



iTBS, 80% AMT

cTBS, 90% AMT


10 (chronic subcortical stroke: 7–86 months after stroke)

Ackerley et al. (2010)

iTBS over the lesioned motor cortex: chronic stroke

cTBS, 80% AMT




cTBS over the contralesional unaffected motor cortex: chronic stroke

Uy et al. (2003) 180 IPAS, 20 sessions (4 weeks)


IPAS over the contralesional unaffected motor cortex: chronic stroke

Study (year)


Review Ayache, Farhat, Zouari, Hosseini, Mylius & Lefaucheur

Expert Rev. Neurother. 12(8), (2012)


but experimental and imaging data suggest that the reorganization of motor circuits in the cortex plays a more important role than the processes of neural repair following stroke [14] . For example, it has been shown that the hand motor function correlated well with measures of corticospinal tract integrity in the first weeks after stroke, but rather with measures of intracortical excitability in the unaffected hemisphere at 3 months poststroke [15] . Thus, in the acute stage of stroke, motor performance is reduced by damage to corticospinal output directly due to stroke lesion, while the mechanisms of motor recovery in the following weeks and months rather depend on reorganization in alternative cortical networks, notably involving interhemispheric connections and balance. The activities of motor cortices in both hemispheres are modulated by transcallosal interhemispheric inhibitory projections that might be beneficial for the execution of various types of movement but might also be deleterious in neurological disorders such as stroke [16,17] . Although there is a controversy about whether transcallosal interhemispheric influences are primarily excitatory or inhibitory [18] , a number of studies have shown that, following stroke, this is an inhibitory balance between the two hemispheres, which is altered. Stroke reduces the activity of the affected motor cortex, including its transcallosal inhibitory projections to the homologue motor cortex of the unaffected hemisphere. This phenomenon leads to an elevated inhibitory drive from the intact hemisphere, which is disinhibited, to the lesioned one, which is overinhibited, worsening the reduction of neuronal activities in the area of stroke and impeding functional motor recovery [16,19,20] . The disinhibition of the unaffected hemisphere varies with time, decreasing as a function of motor recovery [21] , and with stroke location, being more marked in patients with cortical stroke [22,23] , who also had a poorer functional prognosis, as compared with patients with subcortical stroke. In addition, the hyperexcitability of the unaffected hemisphere may be enhanced by an overuse of the normal upper and lower limbs, compared with the paretic body side. Cortical excitability studies, based on paired-pulse TMS methods, confirmed that intracortical facilitation was increased in the unaffected hemisphere, whereas intracortical inhibition was bilaterally reduced, but more markedly in the affected hemisphere [22,24–28] . However, the reduced intra cortical inhibition in the contralesional motor cortex may represent an adaptive process supporting recovery [28] . On this basis, one can conceptualize that the inhibition of neural activities in the unaffected hemisphere or the reactivation of the lesioned one would compensate the interhemispheric unbalance and restore the excitability of the affected hemisphere, ensuring motor function rehabilitation. The changes in motor cortex excitability after stroke, as assessed by TMS techniques, can be used to determine prognosis factor of motor recovery, in conjunction with data provided by clinical scores and neuroimaging [29] . TMS assessment based on MEP recordings might even have better positive predictive value than diffusion tensor imaging tractography [30,31] . The functional correlates of TMS excitability studies in stroke patients have been summarized in a previous report [32] . Briefly, a good outcome can be predicted by the persistence of MEPs in the paretic limb in www.expert-reviews.com

Review

response to the stimulation of the affected hemisphere in the first week after stroke [33–37] , whereas hyperexcitability of the unaffected hemisphere correlates with a poorer outcome [38] . This unbalance of excitability between both hemispheres decreases in parallel with functional improvement in the first months after stroke [36,39,40] . Modulation of cortical excitability by NICS

NICS techniques are mainly represented by transcranial direct current stimulation (tDCS) on the one hand and repetitive TMS (rTMS) on the other hand, including the conventional paradigms of low-frequency stimulation (applied as a single continuous train) and high-frequency stimulation (applied as intermittent trains regularly separated by pauses) and new paradigms, such as theta burst stimulation (TBS). The methodological aspects and mechanisms of action of these techniques are described elsewhere [41,42] and will not be detailed. Briefly, rTMS is a neurostimulation technique that can produce sustained changes in cortical excitability, mainly depending on stimulation frequency. When applied to the motor cortex of healthy humans, low-frequency rTMS (