european journal of paediatric neurology 21 (2017) 75e103

Official Journal of the European Paediatric Neurology Society

Original article Advancing non-invasive neuromodulation clinical trials in children: Lessons from perinatal stroke

Adam Kirton

Departments of Pediatrics and Clinical Neurosciences, Cumming School of Medicine, Hotchkiss Brain Institute and Alberta Children's Hospital Research Institute, 2888 Shaganappi Trail NW, Calgary, AB T3B6A8, Canada

abstract

Keywords: Applications of non-invasive brain stimulation including therapeutic neuromodulation are Neuroplasticity expanding at an alarming rate. Increasingly established scientific principles, including Developmental plasticity directional modulation of well-informed cortical targets, are advancing clinical trial Brain stimulation development. However, high levels of disease burden coupled with zealous enthusiasm Cerebral palsy may be getting ahead of rational research and evidence. Experience is limited in the Perinatal stroke developing brain where additional issues must be considered. Properly designed and Transcranial magnetic stimulation meticulously executed clinical trials are essential and required to advance and optimize Transcranial direct current stimula- the potential of non-invasive neuromodulation without risking the well-being of children tion and families. Perinatal stroke causes most hemiplegic cerebral palsy and, as a focal injury Clinical trials of defined timing in an otherwise healthy brain, is an ideal human model of developmental plasticity. Advanced models of how the motor systems of young brains develop following early stroke are affording novel windows of opportunity for neuromodulation clinical tri- als, possibly directing neuroplasticity toward better outcomes. Reviewing the principles of clinical trial design relevant to neuromodulation and using perinatal stroke as a model, this article reviews the current and future issues of advancing such trials in children. © 2016 European Paediatric Neurology Society. Published by Elsevier Ltd. All rights reserved.

potentially dangerous devices. Examples range from shame- 1. First principles of neuromodulation less promotion of enhanced gaming performance to teenagers clinical trials to do it yourself tDCS machines being made in people's base- ments. Ethical issues specific to the application of brain Non-invasive brain stimulation applications are exploding. stimulation in children must also be considered. For these There is great and justified concern that the rate of growth in reasons, and in order to advance the responsible scientific the number of brains being stimulated is already far exceeding study of neuromodulation in the developing brain, several the level of science that supports the approach. There are principles merit discussion here. major issues of unregulated use across broad and often It is highly unlikely that introducing a focal magnetic field vulnerable populations with immoral marketing of unproven, or local current into a functional area of human cortex will

E-mail address: [email protected]. http://dx.doi.org/10.1016/j.ejpn.2016.07.002 1090-3798/© 2016 European Paediatric Neurology Society. Published by Elsevier Ltd. All rights reserved. 76 european journal of paediatric neurology 21 (2017) 75e103

magically create new, clinically relevant function. Instead, an This raises a final principle consideration of disease spec- endogenous substrate for neuroplasticity that might be altered ificity. As a very common, disabling, and highly studied dis- by such neuromodulation seems a much more likely mecha- ease, major depression carries well-defined diagnostic and nism by which brain stimulation might produce lasting, ther- classification criteria. Despite this, there are innumerable apeutic alterations in brain function. This fundamental factors, both measureable and unknown, that would likely tenet also helps correct for the known and large heterogeneity influence response to neuromodulation. In contrast, autism is between subjects inevitably enrolled in such trials. That a TMS a heterogeneous disorder of social and communication measurement as simple as the rest motor threshold can range development that is likely due to hundreds of different genetic from 20 to over 60% of maximum stimulator output across a disorders in addition to other etiologies. This does not mean sample of normal subjects of the same age and gender points that informed, symptom-specific targeting of cortical regions to an even more enormous inter-subject variability in clinically to enhance other therapies or learning is impossible. Howev- diseased populations. However, if such subjects share funda- er, the breadth of heterogeneity must be acknowledged and mental neuroplasticity mechanisms within their cortex (e.g. adjusted for whenever possible if meaningful trials are to be long term potentiation) and are induced to activate them in the designed. Trials of autism due to one specific mutation bring context of desired, functional activity, the potential for neu- limitations of recruitment and sample size and are still not romodulation is likely greater. ideal; consider the phenotypic variability of tuberous sclerosis In a similar context, an informed cortical target for mod- alone. However, striving for disease specificity whenever ulation is also essential. Identification of such functionally possible will likely advance progress in paediatric neuro- relevant cortical regions is often difficult. As outlined below, modulation trials much faster. Extricating the very specific studies of enhancement of motor learning with brain stimu- forms of perinatal stroke from the more complex world of lation and have often logically targeted the primary motor cerebral palsy for motor learning neuromodulation trials cortex. This has logically extended to clinical populations with provides a practical example. motor disability, targeting the motor cortex and related network components in common populations of motor disability such as adult stroke hemiparesis.1 Importantly, this 2. Perinatal stroke evolving process has not rested on such simplistic anatomical localization alone. Instead, neurophysiological models have You will not likely incur a higher period of risk for ischaemic been developed to first understand what happens to the sys- stroke than the week you are born.4 A term newborn carries a tem of interest in the disease state. These often include large risk >1:3500,5 three-fold higher than a week in the life of a bodies of evidence from preclinical animal models combined diabetic, hypertensive, smoking adult and eight-fold above all with human studies using advanced neuroimaging and other adults.6 An additional 50% of perinatal stroke presents later in neurophysiology tools. Such an example for perinatal stroke infancy.7 Perinatal stroke is the leading cause of hemiplegic will be presented below. cerebral palsy (HCP) and most survivors suffer additional Such models not only identify potential targets but also a neurological sequelae including intellectual disabilities, lan- desired direction for change. For example, the lesioned motor guage impairments, developmental and behavioural disor- cortex may be underactive while the homologous region of the ders, and epilepsy.8e10 Frequent occurrence combined with contralateral, non-lesioned hemisphere may be relatively lifelong morbidity generates large global burdens. Identifica- overactive. Such a model of “imbalanced interhemispheric tion of a causative factor remains elusive in most cases11 and motor inhibition” is probably over simplified but is well sup- with no means of prevention, perinatal stroke and HCP will ported by large volumes of neurophysiological evidence and burden thousands of children for decades to come. has driven the majority of non-invasive brain stimulation An essential first step in improving outcomes from peri- trials in adult stroke.1,2 Recent summative evidence of rTMS natal brain injury is to understand the underlying disease. We therapeutic trials highlights this point by comparing modal- have defined distinct clinical-radiographic perinatal stroke ities and targets across a wide range of such conditions.3 syndromes,11,12 refining perinatal stroke research toward Importantly, each these three principles of modulating an specific disease states. Two main types predominate. These informed target in a specific direction during activation of endoge- are summarized in Fig. 1. We have validated this imaging- nous plasticity are arguably still not well defined in relatively based classification system and demonstrated it's research concrete examples like adult stroke. In fact, such principles applications including the prediction of long-term neurolog- are often not entirely obvious (or even theoretically well ical outcomes,12,13 recognition of novel risk factors,14e16 im- defined) in many other stimulation clinical trials. While such aging markers of disease processes, and new targets for failure should raise immediate concerns of validity, their therapeutic interventions.14,17 Arterial ischaemic strokes (AIS) presence is relatively sparse in the most defined therapeutic are large brain injuries secondary to occlusion of major cere- non-invasive brain stimulation population: adult major bral arteries. Some present at birth with acute seizures (called depression. High frequency rTMS of the dominant dorsolat- symptomatic neonatal AIS) while others are not recognized eral prefrontal cortex (DPFC) is FDA and Health Canada until infancy when hemiparesis becomes evident (called approved and rapidly expanding as an insured service. While arterial presumed perinatal ischaemic stroke).7,18e20 In based on some human evidence of regional dysfunction in contrast, periventricular venous infarctions (PVI) are subcor- this broad, highly connected area with functional implications tical white matter lesions acquired well before birth. Sec- for some symptomology, it could be argued that the ability of ondary to germinal matrix bleeds with subsequent medullary depression to satisfy the above criteria is modest at best. venous infarction, these lesions occur in utero before 34 weeks european journal of paediatric neurology 21 (2017) 75e103 77

Fig. 1 e Developmental plastic motor organization following perinatal stroke. (A) Arterial perinatal strokes (NAIS or APPIS) acquired near term result in damage to cortical and subcortical structures. (B) Periventricular venous infarction (PVI) are acquired in utero prior to 34 weeks gestation resulting in isolated injury to the subcortical white matter. Both lesions damage corticospinal tracts, leading to contralateral hemiparetic CP. (C) Animal and human evidence has constructed working models of developmental motor organization following such early unilateral injury. Control of the weak right hand (W) often relies on both contralateral corticospinal pathways from the left (lesioned) motor cortex (LM1) and ipsilateral projections from the unlesioned (right) motor cortex (RM1, dashed line). These two inputs compete to establish synapses with anterior horn spinal motor neuron pools (circles) during development. Each M1 may also influence the other via IHI (coloured dashed arrows). The relative interhemispheric balance is associated with clinical function with contralateral control associated with better function. Interventions that promote the success of contralateral (or inhibit the success of ipsilateral) upper motor neuron systems to compete for spinal motor neurons could result in better motor function.

gestation.21e26 MRI work by our group and others can detect symptom, present in 30e60% of acute symptomatic these remote foetal bleeds, facilitating accurate diag- NAIS31,38,39 and >80e90% of presumed perinatal ischaemic nosis.12,26,27 Our recent work has begun elucidating the path- strokes including PVI.12,18,19,31 Clinical, laboratory, and EEG ophysiological and plasticity mechanisms in PVI.12,14,27 Our variables are limited in their abilities to predict motor imaging and population-based data12 suggest AIS and PVI outcome,9,10,29,34,40 but neuroimaging has improved the early prevalence are similar. Most relevant here, both AIS and PVI identification of the most affected children.31,33,34,41e44 We injure primary components of the motor system early in life, and others have described how corticospinal tract diffusion resulting in HCP. This commonality of focal motor injury in a MRI in NAIS45,46 and structural MRI in PPIS12 can predict motor healthy brain, combined with distinct differences in lesion outcomes in infancy. This has opened the window for inter- timing and location, makes perinatal stroke the ideal human vention earlier in development. Deficits in language, vision, model for the study of developmental plasticity. cognition, behaviour and epilepsy also occur, present in 20e60%8,10,18,19,47e49 of arterial strokes. The morbidity of perinatal stroke lasts a lifetime, amplifying the burden on 3. Perinatal stroke outcomes child, family, and society.50 Physical disability contributes across this realm of consequences and current interventions Neurodevelopmental deficits occur in ~75% of perinatal stroke have limited efficacy.51 There is therefore an urgent need for survivors.8,10,19,28e34 Hemiparetic CP is the most common new treatment strategies founded upon our best possible term-born cerebral palsy35 and stroke is the leading cause.36,37 understanding of the neurophysiology that underlies the Motor deficits are the most prominent and disabling clinical dysfunction. 78 european journal of paediatric neurology 21 (2017) 75e103

logically overactive contralesional M1 controlling the ipsi- 4. Plastic organization following unilateral lateral weak hand represents a potential therapeutic target perinatal brain injury: an integrated model in hemiparetic CP.55,84 Decreasing pathological over- activity of non-lesioned M1 may enhance motor learning In 1936, Kennard described better outcomes in younger pri- in adult stroke.85e88 In the original Kennard experiments, 52 mates following unilateral motor cortex lesions. This Ken- lesioning the contralesional M1 months after initial injury nard principle has fostered efforts to understand and harness in young primates improved function of the originally age-related plasticity. Common occurrence and focal injury in paretic limb.52 Complex cortical motor circuits within and an otherwise healthy brain makes perinatal stroke an ideal between the hemispheres also mediate neuroplasticity human model. Terms like “repair” and “reorganization” imply including interhemispheric inhibition (IHI)89,90 and intra- the existence of inherent restorative mechanisms that cortical inhibitory and facilatory circuits.65,91,92 Alterations 53 evolutionary models suggest would not exist. Instead, plas- in these have provided insight to adult stroke recov- tic adaptation may represent alterations of normal, ongoing ery.93e97 Each of these model components represent developmental processes occurring after injury. Elegant ani- neurophysiological outcomes measurable with modern mal work and human studies have solidified a model that technologies (below).98 We have measured such systems in creates novel avenues for therapeutic interventions in hemi- children with paediatric99 and perinatal81 stroke, confirm- 54e56 paretic CP. The model consists of 3 primary components ing their relevance to clinical function. (see Fig. 1): The lesioned (A) and non-lesioned (B) motor cortex C. Synaptic competition model. The target of these devel- (and their intra and inter-hemispheric connections) and their oping upper motor neuron systems are the spinal lower influence on spinal motor neuron pools (C). motor neurons, control of which determines function. Continuous competition between contralateral and A. The lesioned hemisphere: contralateral projections to ipsilateral corticospinal tract projections to establish the paretic hand. Adult stroke and animal studies sug- synapses with these cells occurs through development gest that, on average, motor control in the lesioned with eventual contralateral domination and withdrawal hemisphere is associated with better function.57e61 A of ipsilateral projections.54,55 Primate studies confirm a small childhood stroke study showed that recruitment protracted developmental period for such organ- of perilesional motor area may be associated with ization56,100e102 spanning childhood and into adult- higher function.62 We have further correlated the hood.103 Injury likely reduces the normal contralateral presence of contralateral projections from the lesioned advantage, allowing ipsilateral projections to establish hemisphere with better function.63,64 Rare hemiparetic control in more severe hemiparetic CP. CP studies have shown decreased excitability in the Cortical stimulation or inhibition can modulate this pro- lesioned hemisphere65 while maintained motor activa- cess. Pharmacological inhibition of M1 alters spinal motor tions approximate typical motor areas.66 PVI-like le- neuron innervation and enhances ipsilateral pro- sions have associated this contralateral arrangement jections104,105 as seen in hemiparetic CP.55 Daily M1 elec- with better function.67 Small studies have suggested trical stimulation can also preserve the corticospinal rehab-induced clinical gains may be associated with connections normally withdrawn during early develop- increases in lesioned motor cortex activations in ment in cats.106 These findings complement evidence of HCP.68,69 Enhancing motor control in the lesioned activity-dependent enhancement of corticospinal connec- hemisphere should favour improved function. tions in animal,105 histopathological,107 and neurophysio- B. Role of the unlesioned hemisphere: ipsilateral pro- logical studies.71,79 Interventional strategies to enhance jections to the paretic hand. Abnormal projections from contralateral (or inhibit ipsilateral) corticospinal pro- the unlesioned hemisphere to the paretic hand are jections might therefore enhance motor development and common in hemiparetic CP70e75 but inconsistent in function.56 This might be achieved in two ways e stimu- their physiology.70e73,76e78 Ipsilateral projections are lation of the lesioned M1 or inhibition of the unlesioned present in equal proportion at birth but are subse- M1. The latter approach carries stronger animal evidence56 quently withdrawn during normal development.79 with authors concluding that: “activity-dependent pro- Ipsilateral projections likely arise from homologous re- cesses later in development can be harnessed to restore a gions of primary motor cortex (M1).67,76,77,13 Our work more normal pattern of corticospinal connectivity and suggests the same in both childhood64,78 and perinatal function.” stroke.81 In HCP,70e74,77,13 arterial,73e75,82,174 and PVI-like lesions67,73,74 ipsilateral projections are associated with poorer motor function. Over-activity of the non- lesioned M1 may be associated with larger defi- 5. The window of opportunity in cits.67,73,82 In the largest TMS study of perinatal stroke developmental neuroplasticity (n 52), we recently described the neurophysiology of ¼ the contralesional hemisphere, validating many of Modern definitions of CP suggest deficits are static and non- these observations (Zewdie et al., under review). progressive.108 However, neonates with stroke usually Recruitment of ipsilateral projections may therefore demonstrate no observable neurological deficits.109 Asym- represent an example of maladaptive plastic organization metry is not appreciated until 4e6 months12,19 with the full not compatible with normal hand function. A patho severity typically appreciated years later.110 Consistent with european journal of paediatric neurology 21 (2017) 75e103 79

the model above, progressive loss of contralateral control optimize use-dependent changes in brain plasticity and through early development may result in imbalance of cortical function. Such programs should be based on best available control and motor networks, resulting in impaired function. evidence and structured according to standardized elements Both animal100,111 and TMS studies of motor development by such as the TiDIER criteria.105 Consistency of intensity, dosing, ourselves99 and others73,112 agree that plastic motor organi- and methodology can be increased by strict adherence to a zation continues well beyond adolescence. New trials are manual of operations by a limited number of trained thera- proving that the plasticity of even the aged adult brain is pists.144 Structured home programs based on the same prin- modifiable with multiple interventions.113 Whether or not the ciples with ongoing therapist support may be provided during optimal window for interventional modulation suggested in longer-term follow-up periods. Therapy programs should be animal models falls within the earliest years of human designed and delivered by experienced paediatric occupa- development remains to be determined. Before brain stimu- tional therapists. The cost of using such highly qualified lation can be advanced toward this very young age of possible personnel for so many hours of 1:1 therapy may be restrictive. maximal plasticity, the neurophysiology, efficacy, and safety We recently used final year therapy students supervised by 2 of modulatory interventions must first be established in older experienced paediatric OT's to deliver such therapy in camp children. Hence, school age children represent the ideal pop- model with good success, broader knowledge translation to ulation in which brain stimulation and modulation of brain learners, and much lower costs. plasticity should be explored. Before invasive, permanent in- Providing such intensive, goal-directed, evidence-based terventions (e.g. stem cells)114 can be advanced, non-invasive therapy to all participants provides numerous potential ad- means need to demonstrate efficacy of the approach while vantages. These go beyond the primary principle above that elucidating the neurophysiological principles on which they an endogenous substrate for plastic change may well be an are based. essential requirement for brain stimulation to be effective. That all subjects receive individualized, “best available” treatment facilitates randomization to such additional in- 6. Therapeutic neuromodulation in terventions where there is equipoise regarding efficacy and hemiparetic CP minimal, but not zero, risk of adverse events. It creates op- portunity for active, motivated participation focused on aims 6.1. Intensive motor learning, manual therapy that matter to the individual patient. Limitations include the dosage of therapy achieved over focused time frames where Owing to an increasing number of clinical trials, evidence for the optimal number of hours, both total and divided between emerging hemiparetic CP therapies is increasing. Constraint focused and more general training, remain to be determined. induced movement therapy (CIMT) promotes functional use The balance of unimanual versus bimanual and the timing of of an impaired limb by constraint of the less-impaired limb how the two should be integrated is also imprecise and in coupled with repetitive motor practice.113,115e118 In adult need of better evidence. Lastly, the potential psychological stroke, 2 weeks of CIMT can generate gains lasting and social benefits of such group-based participation with years.113,118,119 Multiple paediatric trials support CIMT effec- grouping of participants by developmental level should not be tiveness in hemiparetic CP.51,120e128 Consistent with the underestimated. This may be optimized by inclusion of group model below, functional imaging suggests CIMT shifts motor activities within camp based day programs or, alternatively, function toward the lesioned hemisphere.68,129e132 CIMT- incorporated into smaller time frames within after school induced cortical reorganization has been demonstrated in programs, possibly with the addition of group weekends ses- adults with fMRI130,133e135 and TMS129,136e142 and small hem- sions. When asked, most children with hemiparesis have not iparetic CP studies.68,69,132 CIMT limitations include a some- previously met a peer with the same disability. Working what invasive nature and the exclusion of bimanual learning. together to achieve personal goals alongside similarly Bimanual approaches can also improve function in hemi- affected, motivated peers likely carries large psychosocial paretic CP trials.143,144 Hand-arm Intensive Bimanual Therapy benefits, the more accurate measurement of which is a goal of (HABIT) is an evidence-based, safe, valid, and effective motor future trials. learning therapy in children with hemiparetic CP.124,144e147 As outlined below, we have completed two clinical trials of Efficacy appears equal to CIMT144 but the absence of non-invasive brain stimulation in children with perinatal constraint facilitates functional bimanual motor learning and stroke and hemiparesis. This included a total of 68 partici- removes the complications of casting in our outpatient pop- pants between the ages of 6 and 18 years: 45 in an rTMS trial105 ulation. Comparisons of CIMT and HABIT suggest possible and 24 in a tDCS intervention (unpublished). All participated greater achievement of self-directed goals with HABIT.144 in child-centred, goal-directed, age-appropriate intensive Models for intensive motor learning may combine ap- motor learning programs over 2 weeks. proaches with early CIMT (e.g. week 1) to induce new func- tions in the paretic limb followed immediately (week 2) by 6.2. Non-invasive brain stimulation: repetitive bimanual therapy to encourage incorporation of new function transcranial magnetic stimulation (rTMS) into more functional tasks. How such therapies are delivered has also evolved. Inten- TMS given repeatedly can produce lasting changes in brain sive, camp-based models are increasingly popular, both for function. rTMS studies have established this principle in psychosocial and programming benefits but also to deliver health and disease over the past 20 years148e150 with recent high doses of structured motor learning therapy that may evidence-based summaries of therapeutic efficacy.3 High 80 european journal of paediatric neurology 21 (2017) 75e103

frequency rTMS (~10 Hz) stimulates cortex which both ani- hand function in with chronic subcortical stroke acquired mal151e154 and adult148 stroke studies suggest can facilitate during childhood.63 motor function. Similarly, low frequency rTMS (~1 Hz) typi- We recently completed the PLASTIC CHAMPS trial cally inhibits cerebral cortex though there are exceptions to (Fig. 3).172 This was a factorial, controlled randomized trial both these rules.155e157 rTMS is amenable to randomized, designed to test the ability of contralesional, inhibitory rTMS sham-controlled clinical trials.158 Accumulating evidence and CIMT to enhance motor learning in children with peri- suggests rTMS can modulate neural networks159 to enhance natal stroke and hemiparesis. Forty-five children aged 6e18 motor function in chronic adult stroke.160,161 Limitations of years completed all outcome measures from baseline to 6 rTMS include very focal administration and burdensome, months. There were no serious adverse events or drop outs immobile hardware that prevents simultaneous rehabilitation and tolerability measures were favourable. An additive effect and co-activation of endogenous motor learning systems. of rTMS and CIMT was observed on the primary objective Despite both the high burden of motor disability and outcome of change in AHA at 6 months where the addition of greater brain plasticity in children, rTMS studies have been both therapies more than doubled the chances of a clinically limited. Completed rTMS studies have reported favourable significant improvement (Fig. 2). Subjective, psychosocial and tolerability with no significant adverse events.162e164 Daily quality of life gains were also observed. This trial also rTMS for weeks in animals,165 adults with stroke149,160,166e170 confirmed the feasibility of measuring both baseline AND and our recent paediatric stroke trials63,81,160,171 further sup- post-intervention plastic neurophysiology using advanced port this safety. Evidence from our group and others has imaging and TMS neurophysiology (see below). This trial shown no adverse effect of non-lesioned M1 inhibitory rTMS provides class II evidence for rTMS and CIMT enhancement of on normal (unaffected) hand function in hemiparetic sub- motor learning therapy while supporting the overall safety jects.57,63 We completed the first paediatric rTMS randomized and feasibility of conducting non-invasive stimulation trials trial where 8 days of non-lesional inhibitory rTMS improved in children with perinatal stroke-induced hemiparesis.

Fig. 2 e Timeline of evaluations, interventions, and outcomes. Baseline neurophysiology and motor function studies are performed within a month of intervention. HABIT therapy (90 min) with TDCS or sham (first 30 min) intervention occurs daily for 10 consecutive weekdays. Both Jebsen Taylor hand function test (JT) and safety outcomes (S) are evaluated daily. All outcomes are repeated the week following intervention. Patients enter a home maintenance therapy program with clinical outcomes retested at 2 and 6 months. See text for details. Abbreviations: fMRI: Functional MRI; rsfMRI: Resting state fMRI; TMS: Transcranial magnetic stimulation; TDCS Transcranial direct current stimulation; HABIT: Hand-arm bimanual intensive therapy; AHA: Assisting Hand Assessment; COPM: Canadian occupational performance measure; TST: TDCS safety test; JT: Jebsen Taylor Test of Hand Function; MA: Melbourne Assessment of Unilateral Upper Extremity Function; BB: Box and blocks test; PP: Perdue pegboard test; PMAL: Paediatric Motor Activity Log; QOL: Quality of life measures. european journal of paediatric neurology 21 (2017) 75e103 81

Fig. 3 e PLASTIC CHAMPS trial flow and primary results. (Top) Recruited participants have all motor measures and TMS neurophysiology measured at baseline. They then enter a day camp for 10 consecutive weekdays, all receiving intensive motor learning therapy and randomized to rTMS and CIMT or not. Outcomes are repeated at 1 week and clinical measures at 2 and 6 months. (Bottom) Main results demonstrated additive benefits of rTMS and CIMT on AHA scores at 6 months (left) with large and sustained gains in COPM across all treatment groups (right). (Modified from Kirton et al, Neurology 2016).174

6.3. Non-invasive brain stimulation: transcranial direct healthy adults when administered briefly over the motor current stimulation (tDCS) cortex.180e183 Adult studies have not only demonstrated enhanced motor skill learning with contralateral anodal or ipsi- tDCS applies scalp electrodes (anode and cathode) to generate lateral cathodal tDCS but are also elucidating the mechanisms of weak direct currents (1e2 mA) that induce polarity-dependent neuroplasticity involved.180,181 The duration of effect clearly changes in brain excitability.173 tDCS induces regional, tran- outlasts tDCS interventions by hours to days in a dose dependent sient modulation of resting membrane potential and cortical fashion, confirming a therapeutic potential.85,86,184e186 neuronal excitability.174 In general, anodal stimulation increases Recent trials provide Class I evidence that tDCS can cortical excitability while cathodal stimulation decreases it. enhance motor recovery in adults with chronic stroke (see Modern commercial tDCS systems are painless, inexpensive, Table 1).2,85e87,187,188 Though fundamental mechanisms may and portable, allowing patients to remain mobile during active differ, the same approach outlined above estimulating the rehabilitation. tDCS safety and tolerability in adults is well lesioned, or inhibiting the unlesioned hemisphere, (or both)2 e established with thousands of tested and published safety appears to enhance motor function. For example, Lindenberg guidelines.175e177 Seizure or other serious adverse events have et al.2 studied 20 chronic, hemiparetic stroke patients. Five not been reported. Subjects may report mild, transient tingling or days of structured therapy (60 min) was combined with 30 min itching during the first minute of tDCS current escalation.177 This of bilateral anodal (lesioned M1) and cathodal (non-lesioned sensation can be mimicked in sham experiments, making tDCS M1) tDCS. Outcomes included validated stroke motor function amenable to blinded, randomized trials.178 A published measures and pre and post-intervention fMRI neurophysi- consensus statement endorses the ability of tDCS to enhance ology. Collectively, such studies support the feasibility, safety, motor learning (and other brain functions) in healthy and and potential of our proposed study and provide a solid diseased adults179 but does not even mention the developing foundation for design, stimulation parameters, and outcome brain. tDCS can enhance motor learning in both animals and measures. They also highlight directions for advancement e 82 Table 1 e Summary of tDCS clinical trials in adult stroke. Study Stroke population N Site tDCS Training Outcome measures Results summary Hummel, 200587 Chronic 6 AA 25 cm2; 1 mA; 20 min JTT practice JTT, TMS JTT improved with AA, outlast S stim period. AA increased cortical excitability Hummel, 200588 Chronic 1 AA 25 cm2; 1 mA; 20 min JTT, pinch force, reaction Improved JTT, reaction time, S time, TMS and increase cortical excitability Fregni, 200589 Chronic 6 AA 35 cm2; 1 mA; 20 min JTT practice JTT Improved JTT with AA and CU CU uoenjunlo adarcnuooy2 21)75 (2017) 21 neurology paediatric of journal european S Hummel, 2006267 Chronic 11 AA 25 cm2; 1 mA; 20 Pinch force, reaction time AA shortened reaction time and S increased pinch force relative to S Hesse, 2007268 Subacute 10 AA 1.5 mA; 7 min; 30 days Robot FM 3/10 improved in FM Boggio, 2007269 Chronic 9 AA 5 days JTT Both AA and CU improved JTT CU compared to sham S Celnik, 2009270 Chronic 9 AA 49 cm2; 1 mA; 20 min Key stroke tDCS PNS improved key þ S stroke accuracy compared to PNS sham conditions or tDCS alone. Maintained 1 and 6 days after training. Kim 2009271 Subacute 10 AA 25 cm2; 1 mA; 20 min BBT, finger acceleration AA improved BBT and finger S acceleration to a greater extent than S. BBT improvements lasted at least 60 min post-stim Kim 2010272 Subacute 19 AA 25 cm2; 2 mA; 20 min; 10 30 min of OT FM, BI CU led to greater FM CU days improvements compared to S S Lindenberg 20102 Chronic 20 AA CU 16 cm2; 1.5 mA; 30 min; PT/OT; 5 days FM; WMFT Real tDCS improved FM and þ S 10 days WMFT compared to S. Improvements lasted at least 1 week Mahmoudi 2011195 Subacute to Chronic 10 AA CU 35 cm2, 1 mA; 20 min JTT Improvements with AA CU,

þ þ e

AA AA and CU. No improvements 103 CU with AA (extra-ceph) or S AA Nair 2011273 Chronic 14 CU 1 mA; 30 min; 5 days OT for 60 min/day; 5 ROM, FM, fMRI CU resulted in greater S days improvements in ROM and FM than sham. Effects lasted at least one week. Decreased fMRI activation in contralesional motor region Hesse 2011274 Subacute 96 AA 35 cm2; 2 mA; 20 min; 20 min robot; 30 FM, muscle strength and All groups improved, no effect CU daily sessions tone, BI, BBT between interventions S Bolognini 2011275 Chronic 14 AA CU 35 cm2; 2 mA; 40 min CIMT 14 days JTT, pinch and grip JTT, strength, MALS, FM þ S strength, MALS, FM, TMS improved with tDCS. Increased lesioned M1 excitability Stagg 2012276 Chronic 13 AA 35 cm2, 1 mA; 2 min Reaction time AA improved reaction time, CU and CU to a lesser extent S compared to S. Lindenberg 2012277 Chronic 10 AA CU 1.5 mA; 30 min PT/OT 60 min/day; 10 FM, WMFT Larger improvements in the þ days (re-assess at 5 first week compared to the days) second week Takeuchi 2012278 Chronic 27 1 Hz rTMS 25 cm2; 1 mA; 20 min Finger tapping, pinch force, rTMS-tDCS enhanced pinch AA TMS force, but had no effect on uoenjunlo adarcnuooy2 21)75 (2017) 21 neurology paediatric of journal european rTMS AA finger tapping. Changes in TCI þ Rossi 2013279 Acute 50 AA 35 cm2; 2 mA; 20 min; 5 FM, NIHSS Both groups improved, but no S days difference between AA and S Zimerman 2012280 Chronic 12 CU 25 cm2; 1 mA; 20 min Finger tapping; TMS CU enhance motor skill S compared to S. Correlation between SICI and enhancement. Wu 2013281 Subacute 90 CA (S1) 25 cm2; 1.2 mA; 20 min; PT; 30 min twice a day, MAS, FM, BI More clinical important S (S1) 5x/wk 4 wk 5x/wk 4wk differences with active tDCS Â Â compared to sham post- treatment and at 4 week follow up. Ochi 2013282 Chronic 18 AA 35 cm2; 1 mA; 10 min; 5 Robot-assisted arm FM, MAS, MAL Both tDCS conditions improved CU days training; 5 days FM and MAS, but not MAL Lefebvre 2013283 Chronic 18 AA CU 35 cm2; 1 mA; 30 min Training circuit task Grip force, PPT, circuit task tDCS enhanced circuit task þ S online learning, with retention at 1 week. Improved PPT scores with tDCS Fusco 2013284 Subacute 9 AA CU 35 cm2; 1.5 mA; 15 min 9-hole peg test; pinch force; AA improved 9-hole peg test, þ AA grip strength CU improved strength. No CU improvements with bilateral. S Lefebvre 2014285 Chronic 19 AA CU 35 cm2; 1 mA; 20 min Precision grip movements, Grip and PPT improved with þ S PPT tDCS compared to S Khedr 2013286 Subacute 40 AA 35 cm2; 2 mA; 25 min; 6 Therapy 1 h after tDCS; NIHSS, BI, muscle strength; All groups improved (smaller e

CU days 30 min TMS improvements in muscle 103 S strength). No difference between AA and CU, but both better than S. Increased cortical excitability in lesioned M1 O'Shea 2014287 Chronic 13 AA 35 cm2; 1 mA; 20 min Reaction time AA CU had no effect on þ CU reaction time compared to AA CU sham, whereas AA and CU did þ S quicken RT. (continued on next page) 83 84 Table 1 e (continued) Study Stroke population N Site tDCS Training Outcome measures Results summary

Giacobbe 2013288 Chronic 12 AA 2 mA; 20 min 20 min robotic practice Robot; TMS Movement speed improved. S Smoothness improved with tDCS. tDCS after practice reduced speed. Increased cortical excitability Fusco 2014289 Subacute 16 AA ? PT ? Both AA and S improved S dexterity. No chance in hand force Lee 2014290 Subacute 59 CU 25 cm2, 2 mA; 20 min; 15 VR therapy, 30min/day MAS, manual muscle test, Both groups improved manual 75 (2017) 21 neurology paediatric of journal european S days manual function test, FM, muscle and function and FM CU training BBT, BI and BI scores. Change in FM þ was higher in CU training þ group Cha 2014291 Chronic 20 AA 35 cm2; 1 mA; 20 min; 20 30 min/day BBT; grip strength; FM Both groups improved outcome S days scores. tDCS improved BBT and FM more than training alone Fusco 2014290 Subacute 11 CU 35 cm2; 1.5 mA; 10 min; 45 min 2/day 10 days 9-hole peg test; pinch No added benefit of CU   S 10 days (1/day) strength, FM, grip strength; compared to S BI Au-Yeung 2014292 Chronic 10 AA 35 cm2; 1 mA; 20 min; 1 PPT, pinch strength CU improved PPT, but not pinch CU day strength. No improvement with S AA or S Gillick 2015293 Congenital hemiparesis 13 AA CU 35 cm2; 0.7 mA; 10 min; þ S Goh 2015294 Chronic 10 AA 1 mA; 20 min TMS; Trail making test; Increased cortical excitability 5 Hz-rTMS-A reaction time; pinch force with AA. No change in motor outcomes Di Lazzaro 2014295 Acute 14; 20 AA CU 35 cm2; 2 mA; 40 min; 5 CIMT therapy; 1.5 h ARAT; 9-hole peg test; grip No change in motor outcomes. þ þ S days per day strength; MAL; NIHS; TMS tDCS reduced IHI imbalance at 3 months post stroke. Lefebvre 2015296 Chronic 19 AA CU 35 cm2; 1 mA; 30 min Trained circuit task PPT, circuit task Greater improvement in circuit þ S task and PPT with real tDCS. Sattler 2015297 Acute 20 AA PNS 35 cm2, 1.2 mA; 13 min; 5 JTT, 9-hole peg test, grip Greater improvements at 2 and þ S PNS days strength, FM, hand tapping; 4 week follow up with AA e þ 103 TMS compared to sham. Improvements on secondary outcome measures too with AA compared to S Cho 201556 Chronic 27 AA 35 cm2; 2 mA; 20 min; 18 Mirror therapy or none BBT. FM, JTT, grip strength Control groups improved grip days (3 6weeks) (control); 18 days (3 6 strength and JTT. Greater   weeks) improvements in BBT and grip strength with tDCS Rocha 2016298 Chronic 21 AA 35 cm2; 1 mA; 13 min (for CIMT; 6 h/day for 3 wk FM, MAL, grip strength Increase in FM for active tDCS CU AA) or 9 min (for CU); 3x/ 1 h of PT daily 3x/wk for groups. All groups equally S wk for 4 wks 4 wk improved MAL and grip strength. Straudi 2016299 Acute - Chronic 23 AA CU 35 cm2, 1 mA, 30 min; 10 Robot-assisted therapy; FM, BBT, MAL Both groups improved FM, but þ S days 10 days no difference between AA and S. Patients with chronic subcortical strokes benefited more than those with acute and cortical stroke however. Allman 2016300 Chronic 24 AA 35 cm2; 1 mA; 20 min; 9 Motor training; 1hr per FM, ARAT, WMFT ARAT WMFT showed S days day improvements up to 3 months post-training with AA Triccas 2015301 Subacute e Chronic 22 AA S uoenjunlo adarcnuooy2 21)75 (2017) 21 neurology paediatric of journal european Cunningham 2015302 Chronic 12 AA 35 cm2; 1 mA; 30 min per CIMT therapy; FM; 9-hole peg test; TMS tDCS improved motor þ S stimulation 30 min 2/day x 3/week outcomes. Increased  for 5 weeks excitability in the contralesional hemisphere. Lee 2015303 Chronic 24 AA? ?? Physical therapy FM Real stimulation plus therapy S improved FM scores to a greater extent compared to therapy alone Chen 2016304 Subacute e Chronic 5 AA CU 16 cm2; 1.5 mA; 30 min; 1 h OT/day; 10 days FM, rsfMRI FM scores improved. Increased þ 10 days connectivity between ipsilesional M1 and contralesional PMC Mortensen 2016305 Chronic 15 AA 35 cm2; 1.5 mA; 20; 5 OT; 30 min/day 5 days JTT, grip strength Both groups improved JTT. AA S days improved grip strength compared to sham (no longer present 1wk post). Trend for JTT retention with AA

Legend: AA e anodal stimulation, affected hemisphere, CU e cathodal stimulation, unaffected hemisphere, FM e Fugl-Meyer motor score, AAT e Aachener aphasia test. e 103 85 86 european journal of paediatric neurology 21 (2017) 75e103

larger, more homogeneous populations, evidence-based months post-intervention. The final results were presented at therapy, more integrated neurophysiological outcomes and, the 2016 International Stroke Conference in Los Angeles in of course, studying the more plastic paediatric population. February 2016 and is currently submitted for publication. In We recently demonstrated the ability of tDCS to enhance summary, interventions were well tolerated with no serious motor learning in typically developing children. One previous adverse events and all safety outcomes were satisfied. Sig- study had demonstrated early evidence of safety, tolerability, nificant tDCS effects were observed on COPM, but not AHA and neurophysiological effects of brief tDCS applications to outcomes, suggesting possible benefit and providing impor- motor cortex in children.189 We conducted a 4 arm, blinded, tant data for the execution of a larger, more definitive trial. sham-controlled trial of motor cortex tDCS to enhance motor learning in children aged 6e18 years.190 Subjects repeatedly performed a motor learning task (Purdue Pegboard) with their 7. Systematic approach to clinical trial left hand over three days. They were randomly assigned to design: CONSORT receive 20 min of contralateral anodal (1 mA), ipsilateral cathodal (1 or 2 mA) or sham tDCS during the first 20 min. Any trial is only as good as the methods on which it rests. Learning curves were strongly enhanced with active treat- Thankfully, the importance of this fundamental concept has ment with effects retained 6 weeks later and additional gains evolved over recent decades to carefully define the essential in untrained motor function of both hands (Fig. 4). We believe criteria of valid clinical trials and their reporting. The CON- this provides important proof-of-principle evidence to SORT (Consolidated Standards of Reporting Trials) statement is advance further studies in hemiparetic CP and other disease an evidence-based minimum set of standards for the reporting populations. of randomized clinical trials.194 First initiated in 1996, the cur- Though there is much need for additional investigation, the rent version was last updated in 2010 and the most up to date most informed primary target in hemiparetic children with version including recent modifications are available online. perinatal stroke currently is the contralesional M1. A cathodal These include a 25 item checklist as well as a template flow approach is supported by the above evidence for tDCS effects in diagram to account for participant movement through a trial. normal children190 and probable effect of the similar approach An accompanying “explanation and elaboration” statement with inhibitory rTMS in the same population.172 Stimulation further defines each element and is available in multiple for- over intact (non-lesioned) brain also maximizes safety with mats. Multiple extensions have also been developed to address more predictable distribution of tDCS currents compared to designs beyond simple two group parallel designs such as children with large arterial lesions.191,192 Methods should be cluster, non-inferiority, and pragmatic trials. based in the best accumulated evidence from previous studies Use of the CONSORT guidelines has been associated with of cathodal tDCS in addition to those cited above including improved reporting of clinical trials.195 The website receives adult motor learning,179e181 adult stroke rehabilitation2,86,87,188 >17,000 hits per month. More importantly, over 600 journals and studies of electrode montages in stroke.193 Soft, replace- have endorsed the CONSORT guidelines. While the primary able 25 cm2 electrodes would typically be placed on clean, dry motive for applying the CONSORT guidelines should be the areas of the scalp with the cathode placed over the non- best possible trial, it is fair to say that no future clinical trial lesioned M1. This location would ideally be individualized for will likely be published in any reputable journal without each subject using TMS and/or task fMRI and targeted with careful adherence. It is therefore reasonable to suggest that neuronavigation. Consistent with previous tDCS stroke studies application of the CONSORT guidelines is both an extremely examining electrode placement,193 the reference electrode helpful and scientifically essential process in the design, would be placed over the contralateral orbit. Using available, execution, and reporting of any clinical trial. With this in programmable, and current-controlled stimulators allows mind, the outline of the guidelines will be applied below to automatic ramp up slowly over 30 s to the treatment current, review several essential elements of trial design as they relate typically 1.0 mA tDCS or sham would usually be administered to non-invasive brain stimulation trials in children. Outcome during the first 20e30 min of motor learning therapy and on a measures, the largest single issue, are discussed separately daily basis in a camp-based model. Experienced technicians following this. should apply and monitor all systems, being on hand for im- mediate trouble-shooting. Some tDCS systems can blind the 7.1. Populations administrator though it is only essential that clinicians administering therapy and those performing the outcome Where participant populations exist and how they are measures are blinded to treatment allocation. accessed for recruitment will primarily depend on the inclu- We recently completed a phase 1/2, randomized, sham- sion and exclusion criteria (discussed next). However, before controlled, double-blind, clinical trial to test whether the these criteria are applied, the sample to be screened should be addition of tDCS to intensive motor therapy was associated carefully selected to represent, as best as possible, the popu- with larger gains in clinical function. Participants completed a lation of interest as a whole. Reductions from the true popu- goal-directed, peer-supported after school camp model of lation toward the sample studied will inevitably introduce motor learning therapy for 2 weeks, randomized to receive selection and other biases, the number and calibre of which contralesional, cathodal tDCS (1 mA) or sham during the first decrease the translatability of any ultimate findings propor- 20 min of each day. Primary outcomes were the COPM (sub- tionately. Therefore, the optimal starting point is a large, jective) and AHA (objective) in addition to multiple safety and complete, population-based sample of subjects with the con- tolerability measures (see discussion below) measured out to 2 dition of interest. As mentioned elsewhere, disease-specificity european journal of paediatric neurology 21 (2017) 75e103 87

Fig. 4 e tDCS motor learning trial methods and results. Motor training of the non-dominant hand over 3 days is shown in the top bars with subjects randomized to 1 of 4 tDCS treatment montages (including sham). Active treatment groups showed markedly steeper learner curves (bottom) with sustained effects at 6 weeks. (Modified from Ciechanski and Kirton, Cerebral Cortex 2016).192

will be a major determinant of this population. For example, The yield for clinical trials however was immense, providing a from a population-based sample of cerebral palsy (something large unbiased sample from which candidates could be already difficult to find), perhaps 30% of these will be hemi- randomly recruited, optimizing translatability to the disease paretic and perhaps 75% of these will have confirmed peri- population as a whole. As outlined in the following sections on natal stroke as the cause. selection criteria and recruitment, this apparently very large We began building the Alberta Perinatal Stroke Project (APSP) number shrinks rapidly as these realities are introduced, in 2007. This is a large population-based cohort of MRI- further biasing the sample along the way. confirmed perinatal stroke consented to research now numbering nearly 800 subjects all living within the Canadian 7.2. Selection criteria province of Alberta (population of ~4.2 million). This required an exhaustive review of over 80 ICD codes, 6000 medical charts, Further reduction of the eligible population for recruitment structured imaging review, patient contact and in person clinic will occur with application of inclusion and exclusion criteria. assessments. The effort took over 7 years and is still ongoing. Their selection must rest on the primary research question 88 european journal of paediatric neurology 21 (2017) 75e103

being asked; if this has to be compromised by changing emphasized that the relationship of these factors to the criteria to suit other needs, the question should instead be ability to respond to an intervention are incompletely, if at changed. Once true to the primary question, additional issues all, understood. Therefore, despite these specific consider- of confounders, known response predictors, clinical realities ations, it should be remembered that simple randomization (i.e. minimum function levels), clinical significance or need should usually correct for such differences across treatment (i.e. maximal function levels), safety, and ability to comply groups. must be considered. Criteria should be strictly quantified or Most forms of randomization can be performed simply qualified according to established metrics when possible. using patient codes and computer-based or online software The following provides an example of potentially relevant administered by an unbiased study member such as the inclusion and exclusion criteria for a combined intensive statistician. Concealment of randomization must be assured motor learning and neurostimulation trial in hemiparetic CP. and multiple means are available. Development, storage, and breaking of codes by an unbiased third party may be reason- 7.2.1. Inclusion criteria able. Treatment administrators can sometimes be automati- cally blinded, such as modern, programmable tDCS systems. 1. Symptomatic hemiparetic CP (Paediatric Stroke Outcome Stimulation trials in children are likely moving toward multi- Measure (PSOM) > 0.5) AND Manual Abilities Classification centre studies, introducing additional considerations. System (MACS)196 grade IeIV AND child/parent perceived Randomization in permuted blocks, with the number of limitations in function) blocks matching the number of participating sites, may 2. Clinical and MRI-confirmed perinatal stroke syndrome facilitate even distribution within sites. Multi-site trials on the (NAIS, APPIS, PVI) other hand, may require a central randomization and treat-

3. Active wrist extension >20, finger extension >10 ment allocation process. Expertise from experienced clinical 4. Can lift the affected arm 15 cm above a table surface and trialists and statisticians may be particularly helpful. grasp light objects 5. Term birth (>36 weeks) and current age 5e18 years 7.4. Recruitment 6. Informed consent/assent

Failure to recruit complete samples on time is one of the most 7.2.2. Exclusion criteria common reasons for trials to be delayed or not completed at all. Estimates of attainable sample sizes must be realistic (i.e. 1. Other neurological disorder not related to perinatal stroke pessimistic). Hard numbers from established populations are 2. Multifocal stroke required rather than blind estimates of prevalence based on 3. Severe hemiparesis (no voluntary contraction in paretic published epidemiological data. What may appear as large, hand, MACS level V) easily adequate number of potential subjects often reduces 4. Severe spasticity in the affected limb (Modified Ashworth dramatically when realistic limitations are estimated. Exam- Scale >3) ples include presence of exclusion criteria, inability to confirm 5. Severe developmental delay or other inability to comply all inclusion criteria, geographical factors, failure to recruit with study protocol rates, and attrition with drop-outs possible at all stages (who 6. Unstable epilepsy (>1 seizure/month or >2 medication must be included within intention to treat analyses if occurs changes last 6 months) following randomization). 7. Any TMS171 or MRI contraindication including implanted electronic devices 8. Botox, orthopaedic surgery, or other invasive mechanical 7.5. Informed consent and assent therapy in past 6 months 9. Constraint, brain stimulation or other modulatory therapy Consenting children and parents to novel, experimental trials in past 6 months of brain stimulation requires special attention. Potential benefits may be easily over-estimated by families of disabled 7.3. Randomization children due to lack of alternative therapies, being “impressed” by technology, or other reasons. Risk must also Methods for randomization within clinical trials are complex be fairly disclosed based on best available evidence. Our lab and beyond the scope of this discussion. However, several has completed nearly 3 million stimulations on over 250 fundamental issues to be considered relate predominantly to children with many conditions without serious adverse event. trial design and a priori knowledge of response predictors and This provides direct evidence to families of minimal risk they other potential confounders. Options including block or can trust however such examples are not always available. minimized randomization may help ensure balance of Theoretical risks must be presented in context with estimates treatment allocation across subgroups of patients. Such of relative risk and potential harm. An ability to understand subgroups may be easily defined (e.g. site A versus site B) or randomization may be difficult to communicate and may be more complicated such as predefined factors known or forgotten months later when final results are shared with highly suspected to be associated with the outcomes of in- families. Both parents and children, in our experience, may terest or potential response to treatment. Examples might sometimes harbour feelings of disappointment or even include factors such as age, level of baseline disability, or mistrust when discovering they were in the sham group, even corticospinal tract arrangement. However, it should be when they may have made substantial clinical gains. Honest, european journal of paediatric neurology 21 (2017) 75e103 89

clear, and repeated explanations are all that can reasonably be offered. 8. Outcome measures

Selection of outcome measures is likely one of the most 7.6. Structure and flow essential components of clinical trial design. Standardized, unbiased administration and interpretation by qualified ex- Structuring a complex, multifaceted intervention with perts blinded to treatment allocation and other clinical in- numerous requirements for space, infrastructure, and highly formation is required but often challenging. Additional qualified personnel requires organized structure. An example issues specific to considering neurostimulation in- of the sequence, timeline and flow of our recently completed terventions in children include a relative paucity of vali- controlled trial of tDCS in school-aged children is shown in dated measures, heterogeneous populations (e.g. CP versus Fig. 2. Factors considered include family convenience where hemiparetic CP versus perinatal stroke hemiparesis), and timing of drop offs and pick-ups, time of year (school year variable effects of age and developmental level to name just versus summer), and extracurricular activities may affect a few. However, these limitations are being overcome compliance and participation. Including unique activities may through a variety of creative means to facilitate evaluation diversify therapy while optimizing fun for participants. Ex- of such interventions. Measures used in similar previous amples in our experience include virtual reality, cooking, trials should be considered in anticipation of comparability therapeutic arts, horticultural therapy, and video gaming (e.g. and eventual pooling of data in meta-analyses. At the same rockband). Balancing focused motor training with more gen- time, careful review of how well certain measures were able eral and group activities as well as breaks and relaxation can to demonstrate change (or not) and other limitations in be challenging. It is essential to include input from experts in those same trials may be good reasons to consider addi- paediatric therapy, child life, and subjects and their families tional or alternative outcomes. Outcome measures can be for optimal planning. considered under headings of clinical/functional, safety, and neurophysiological. 7.7. Sham-control and blinding

Effective sham techniques are well established for non- 8.1. Clinical outcomes: motor function invasive stimulation methods. For rTMS, this includes altering coil direction or use of commercially available sham A rigorous, evidence-based approach to clinical motor coils. For tDCS, devices can often be ramped on for all subjects outcome selection should be adopted. Testing should be to produce the typically transient scalp sensations experi- performed by a team of experienced experts that may include enced, and then ramped off for the sham group. Such sham- paediatric occupational and physical therapists, develop- 178 172,190 ming has been proven effective, including in children. mental paediatricians, paediatric neurologists and physiat- Cross-over design trials carry a risk of subjects detecting dif- rists, and clinical trialists. More specific outcomes may ferences when they switch groups and interventions in naı¨ve require additional expertise such as child psychology re- subjects only avoid this potential problem. Ideally, all in- searchers when considering psychosocial and quality of life vestigators, those administering treatments, outcome asses- outcomes. Multidimensional evaluations should consider the sors, parents and children remain blinded. Those body structure and function, activity and participation do- administering tDCS may not be if they are otherwise separated mains of the WHO International Classification of Func- from the outcome measures and therapy. Modern tDCS sys- tioning, Disability and Health (ICF), considering both tems can also be programmed to randomize and administer unimanual and bimanual performance.197 Tools with estab- accordingly, allowing the administrator to remain blinded. lished clinimetric properties in the population of interest (i.e. hemiparetic CP) are selected to evaluate diverse upper ex- 7.8. Analysis and sample size tremity functions relevant to daily living in children.198,199 Measures need to be both child and family friendly Statistical analysis will of course depend on the research including sensitivity to time and fatigue. Measurement of questions to be addressed. In most circumstances, analysis both uni- and bi-lateral function of both limbs are required will be intention to treat, accounting for all subjects ran- for normative data, evaluation of different clinical functions, domized whether they complete the trial or not. Additional and safety (including screening for changes in unaffected per protocol analyses may be performed and reported but this hand function). Testing should be video-taped for quality is not usually done in isolation. Collaborator expertise in assurance, inter-rater validation, and additional offline clinical trials and biostatistical methods are an essential analysis. Motor outcomes are measured at baseline, and 1 requirement. Variables for calculation of sample size should week, 2 months post-intervention. be extrapolated from the most relevant literature. Estimates With these principles in mind, the following three po- are typically based on application of the first order analysis tential primary outcome measures might be considered in a method to the main hypothesis using the primary outcome. clinical trial of intensive motor learning therapy combined Estimated samples should be comparable or proportional to with non-invasive brain stimulation in hemiparetic chil- previous trials in similar populations. dren. An additional list of possible secondary outcomes and All studies need to be registered prior to consent of the first their supporting evidence and attributes are summarized in patient (www.clinicaltrails.gov). Table 2. 90 Table 2 e Examples of potential functional outcome measures for clinical trials of neuromodulation in hemiparetic CP. Measure Target Application and evidence-base Canadian Occupational Performance Measure (COPM) Individual goals for therapy Individualized, family-centred tool identified child and family-perceived difficulties and personal improvement goals for self-care, productivity (school), and activities. The same experienced OT screened potential goals with child and parents with consensus defining those goals that were both functionally relevant and reasonably achievable. Scaled scores for performance and satisfaction of each goal were obtained (1 lowest, 10 highest). Such subjective, patient-centred measures are increasingly considered essential in paediatric hemiparetic CP trials. The COPM is validated for paediatric hemiparetic CP trials across our age range. Sensitivity to change has been established with an increase in COPM scores of 2 units considered clinically significant.  75 (2017) 21 neurology paediatric of journal european Assisting Hand Assessment (AHA) Bimanual function The AHA is an established evidence-based standard for the objective quantification of bilateral hand function in children with hemiparetic CP. Certified occupational therapists administer a structured, play-based assessment of spontaneous bimanual hand use followed by itemized scoring of the video- taped session. This Rasch-built evaluation carries the strongest evidence of inter-rater, intra-rater, and test-retest reliabilities, test-validity, and responsiveness to change for bimanual tasks in hemiparetic CP children within our age range. Sensitivity to change and excellent clinimetric properties have been established in paediatric hemiparetic CP trials. Consistent with validated methods, AHA scores are expressed as logit units with an increase of 5 units considered clinically significant.  Melbourne Assessment of Unilateral Unimanual function The MA is a validated, criterion-referenced functional measure designed to detect change in Upper Limb Function (MA) hemiparetic children. Trained therapists evaluate upper extremity motor tasks with performance ranked for each creating a total maximal score of 122 points (reported as percentage). Validation studies have correlated MA with clinically relevant functional outcomes in school-aged children.

Jebsen Taylor Test of Hand Function (JTTHF) Unimanual function 1. Standardized timed test of unimanual upper extremity functional activities evaluating efficiency of movement. Important for comparability as the most common outcome measure used in adult stroke tDCS clinical trials and recent paediatric CP studies.

AbilHAND Manual function Designed to assess manual function in children with cerebral palsy, the AbilHAND scores 21 common manual activities in children at 3 levels of difficulty with summary logit unit scores. This measure is not specific to hemiparetic children but has been used in other paediatric CIMT trials.

Grip and pinch strength (GS, PS) Motor function GS is a simple, quick measure of hand motor function quantifiable with a dynamometer employed in many paediatric hemiparetic CP and adult rTMS stroke trials. Unaffected hand GS and PS are also potential safety outcomes.

Box and blocks (BB) Dexterity This test of unimanual dexterity will be applied on both sides as in previously reported paediatric CIMT e

trials. Assessment of the unaffected limb provides a safety measure (described above). 103

Purdue pegboard test (PPT) Dexterity 2. Performed bilaterally to measure manual dexterity and is commonly used in paediatric CIMT studies, and was the primary outcome in our healthy tDCS motor learning trial in children.

Paediatric Motor Activity Log (PMAL) Daily activity 3. Commonly used in hemiparetic CP studies with test-retest stability and sensitivity to change. Completed by parents.

PedsQL Cerebral Palsy Module (3.0) Quality of Life Condition specific and validated, age-adjusted child and parental reports assessed domains of daily activities, school activities, movement/balance, pain/hurt, fatigue, eating activities, and speech/ communication.

APSP Parental Outcome Measure (POM) Parental psychology Measures adverse parental psychological outcomes in parents of children with perinatal stroke including unique symptoms of maternal guilt and blame. european journal of paediatric neurology 21 (2017) 75e103 91

8.1.1. Primary objective motor outcome: Assisting Hand during the 2 week intervention phase. Importantly, we will Assessment (AHA) do this bilaterally (2 watches), allowing the generation of a symmetry index between the affected and unaffected limbs. Our preliminary data has established the feasibility This is currently the established standard for the objective and accuracy of these methods in school-aged children quantification of bilateral hand function in children with (Cole et al., unpublished data). hemiparetic CP.199 A Rasch-built evaluation, the AHA carries the strongest evidence of inter-rater, intra-rater, and test- Many additional clinical outcome measures can be retest reliabilities, test-validity, and responsiveness to considered depending on the nature of the trial. Examples, change for bimanual tasks in hemiparetic school-aged including their supporting evidence and primary targets, are children.199e204 Sensitivity to change and excellent clini- summarized in Table 2. metric properties have been established in multiple pedi- atric HCP clinical trials.120,144,205e207 One potential limitation of the AHA is a drop in score when new unimanual functions 8.2. Safety outcomes have not yet been incorporated into the bimanual tasks being measured (i.e. scores may drop despite new function). With limited non-invasive neuromodulation data in the Our trained therapists have successfully executed >200 AHA developing brain, careful and complete application of safety measurements within multiple clinical trials172 with excel- outcomes is paramount within clinical trials. Adult guidelines lent compliance and robust data. and safety reviews are available for both TMS171 and tDCS175 and are certainly applicable. However, issues unique to chil- 8.1.2. Primary subjective motor outcome: Canadian dren need to be screened for and rates of tolerability and po- occupational performance measure (COPM) tential adverse events documented. Safety issues should be identified a priori by experienced investigators but also at arms-length by a data safety and monitoring board (DSMB). Subjective outcome measures are now considered a valid, Interim safety analyses may be incorporated into trial design potentially essential outcome measure in rehabilitation at set time points (e.g. after certain numbers of subjects have 51,208 trials including children with hemiparetic CP. It could completed outcomes). These analyses, as well as any serious be argued that any gains shown in objective tests of motor adverse events, should be reviewed by the DSMB and insti- function (e.g. strength, dexterity) are meaningless if the tutional ethics boards requirements according to established, patients themselves have not perceived the achievement pre-defined protocols. of a personal goal or some other personal satisfaction. For Safety can be considered under the following headings. these reasons, individualized, patient-centered, goal directed tools such as the COPM have been developed and 8.2.1. Serious adverse events validated. Applied by an experienced pediatric OT in combination with child and parents, the COPM can identify child and family-perceived difficulties in categories of self- Typical definitions of SAE are adverse events or reactions 209e213 care, productivity (school), and activities. Validated that results in death, is life-threatening, requires hospi- 214,215 for both school-aged children and pediatric hemi- talisation or prolongation of existing hospitalisation, re- 208,214,216e218 paretic CP trials, the COPM has been a robust sults in persistent or significant disability or incapacity, or 172 measure in our previous perinatal stroke trials. We have is a congenital anomaly or birth defect. Fortunately, SAE also recently characterized how COPM goals are set in this reports in non-invasive brain stimulation have been population and their relationship to success in such exceedingly rare across decades of use. interventional trials (Haspels et al., 2016, unpublished data). 8.2.2. Function-specific adverse effects

8.1.3. Novel “real-life, continuous” motor outcome: actigraphic symmetry index (ASI) Unique undesired outcomes may occur in specific studies. In the case of modulation trials for hemiparetic CP, one example is provided by the known control of both upper No existing motor outcome measure can quantify contin- extremities by the non-lesioned hemisphere. Subse- uous use of the upper extremities during the normal ac- quently, inhibitory stimulation of the contralesional motor tivities of real life. We recently proposed to overcome this cortex might include theoretical consideration of reducing limitation using actigraphy. Lightweight wrist accelerom- hand function in the unaffected hand or in the target eters can constantly measure and store subtle movements affected hand, particularly in those with prominent ipsi- (e.g. fitbit). Such systems can also track movements in lateral corticospinal arrangements. Therefore, primary 219 disabled persons including those with CP. In upcoming safety outcomes in our brain stimulation trials have trials, we will use the Actiwatch2 system to record mean included regular measures of both affected and unaffected movements every 15 seconds for 48 hour epochs of time hand function across time points for each subject and (baseline, 2 months, 6 months) as well as continuously within interim safety analyses. 92 european journal of paediatric neurology 21 (2017) 75e103

8.2.3. Side effects receiving tDCS, scalp itching in 55% was mild, transient, and comparable to sham with no drop-outs. Neither low frequency rTMS or cathodal tDCS of the non-lesioned hemisphere Any intervention capable of having biological effects must decreased function of either hand in children with hemi- also have the risk of side effects. Extrapolating from the paretic cerebral palsy. We believe this data supports that brain much larger volume of evidence from the adult brain stimulation is safe and well tolerated in children. stimulation world and published guidelines,171 a reason- able list of side effects to be specifically screened for can be generated. These include feelings of headache or neck pain 9. Neurophysiological outcomes with TMS with the addition of itching or burning sensa- tions for tDCS. However, symptoms specific to young age Advanced neurotechnologies have greatly advanced the abil- groups may exist, such as TMS-induced vasovagal syncope ity to understand developmental plasticity in real patients. in adolescents.220,221 Each anticipated subjective symptom This includes baseline measures to evaluate natural process can be screened for and quantified including scale severity, and how they relate to function but also the opportunity to duration, need for treatment, etc using a simple form. explore the potential mechanisms of intervention-induced Screening and documentation should be repeated over change. Combining modern technologies allows the compre- time to assess for persistence or tolerance. hensive, integrated study of brain structure and function personalized to each individual. Clinically relevant examples of such integration are increasingly available in adults223 but 8.2.4. Tolerability awaiting full exploration in children. Summarized here are leading applications of single- and paired-pulse TMS, imaging including task and resting state fMRI and diffusion tensor A standardized safety and tolerability evaluation for TMS imaging (DTI), and robotics. in children has been developed63,222 and is easily adapted 190 for different population and modalities including tDCS. 9.1. Transcranial magnetic stimulation (TMS) Subjects are asked to rank order their stimulation experi- ence amongst 7 other common childhood experiences. First described in 1985,224 most TMS studies are published since 2008 but only <3% include children. Focused magnetic Optimizing safety requires attention to each of the above fields applied across neuronal membranes result in focal issues. Brain stimulation studies in children should be per- depolarisations of human cortex according to Faraday's Law. formed by experienced personnel in a secure environment. With narrow spatial resolution, TMS can affect discrete Immediate access to medical care should be available in the functional areas,225 offering non-invasive, painless mapping unlikely occurrence of a serious adverse event. Written stan- of motor systems.67,76,77,226 Tools relevant to the current study dardized operating procedures, for both experimental include “single-pulse” measurements such as motor evoked methods and handling adverse events, should be imple- potentials (MEP) which quantify motor pathway neurophysi- mented and staff tested for their familiarity with these. ology from cortex to muscle. Motor thresholds and stim- Our single centre experience has administered nearly 3 uluseresponse curves reflect corticospinal million stimulations to over 280 children in the past 8 years. excitability.72,222,227 Paired-pulse TMS employs multiple Ages ranged from 8 months to 19 years (median 11.2) with the stimuli to elucidate interactions between different brain vast majority being school-aged. The most common condi- areas.90,92,225,228e232 TMS is safe and well tolerated in chil- tions were perinatal stroke/cerebral palsy (70), mild traumatic dren,162,222,233 supported by animal studies106,151,234e238 and brain injury (TBI, 68), or typically developing (53). Multi- published guidelines.171 Application is safe in adult disciplinary neurophysiological studies included single- and stroke85e87,148,149,160,161,166,228 and supported by our safety paired-pulse TMS methods. Therapeutic clinical trials used work in children with stroke.66,72,81,220 TMS-related seizures repetitive TMS (rTMS) and anodal/cathodal motor cortex have not been described in children162 including administra- tDCS. Prospectively collected safety and tolerability data on all tion over seizure foci.239,240 subjects included the paediatric TMS safety and tolerability Modern technologies include single pulse systems for measure, child and parental interviews, and data safety and excitability and pathways as well as paired-pulse methods monitoring boards. There were no serious adverse events. capable of probing cortical excitatory, inhibitory, and other Tolerability between TMS (402,680 stimulations) and rTMS physiological functions. Neuronavigation systems such as (2.1million stimulations) was comparable and rated similar to Brainsight (Rogue Research, Montreal) allow real time co- a long car ride. Although >100 had brain injuries and/or epi- registration of coil location with individual subject MRI in lepsy, no seizures occurred. Headache following a TMS three dimensions, including overlay of functional imaging neurophysiology protocol was more common in perinatal targets. We have successfully demonstrated each of these stroke (40%) than typically developing (13%) participants but methods in children with stroke.99 Single pulse TMS can was mild and self-limiting. Mild neck pain with was relatively determine bilateral M1 excitability.99,241 Motor evoked poten- common but comparable between perinatal stroke (22%), tials are recorded from hand muscle, often the first dorsal depression (19%), and TBI (19%). Tolerability improved over interosseous (FDI), bilaterally. The motor “hotspot” is defined time with rates of headache, neck pain, and unpleasant over M1 as the area producing consistent, high amplitude tingling decreasing by >50% at repeat testing. Of 51 children MEP. Resting motor threshold (RMT) is then defined as the european journal of paediatric neurology 21 (2017) 75e103 93

stimulator intensity required to produce MEP 50 mV in 5/10 9.3. Diffusion tensor imaging (DTI)  trials. Our studies63,78 suggest stroke side MEP are obtainable in >80% of subjects. Activated motor thresholds are also ob- Diffusion MRI acquires microscopic information on structural tained during ~20% of maximal FDI activation via visual brain connections. DTI measures preferential water molecule feedback. Motor maps can also be created in the lesioned and diffusion (Brownian motion) along axonal axes, providing unlesioned hemisphere. We recently added a TMS Robotic unique structural and physiological information in the system (Axilum, Strasbourg) that should greatly facilitate developing and diseased brain.253 DTI non-invasively mea- execution of such outcomes. sures neural fibre composition, integrity, and orientation Additional potential TMS neurophysiological outcomes while tractography creates three dimensional reconstructions include: (1) Stimulus response curves (rest and active). Bilateral of large white matter tracts.254 DTI can evaluate the motor M1 excitability is quantified by measuring MEP amplitudes system including corticospinal tract integrity.257 In adult across 5 escalating stimulus intensities such as 100/110/120/ stroke, corticospinal tract DTI measures correlate with motor 130/140/150% of RMT. (2) Interhemispheric Inhibition (IHI). Bidi- outcomes.258,259 DTI utility has been demonstrated across rectional IHI (lesioned to unlesioned hemisphere and vice- numerous childhood neurological disorders257 and can be versa) applies stimuli to both M1 in quick succession with synergistically integrated with fMRI and TMS to create methods described elsewhere.63,90,96,242e244 Paired-pulse TMS comprehensive, in vivo models of brain structure and func- applies a suprathreshold conditioning stimulus (CS) con- tion.86 Perinatal stroke DTI studies are few260 though CP ap- tralaterally immediately prior to a test stimulus (TS) with plications261,262 support feasibility. We recently demonstrated interstimulus intervals of 10 and 40 ms. The decrease in TS our ability to evaluate corticospinal tract integrity in perinatal MEP quantifies IHI. (3) Short interval intracortical inhibition (SICI). stroke using DTI.263 M1 SICI can be determined according to established paired- pulse methods.91,245 For example, after suprathreshold CS, 9.4. Robotics SICI effects are typically seen with an interstimulus interval of 2 ms. These and other TMS measures of cortical neurophysi- Robots have further advanced the ability to measure complex ology can define baseline function as well as changes occur- clinical function in disabled patients. Robotic technologies such ring secondary to interventions. as the KINARM (Kinesiological Instrument for Normal and Altered Reaching Movement) can assess limb movements at 9.2. Functional MR (fMRI) and the resting state (rsfMRI) the shoulder and elbow. Unique advantages complimentary to the above modalities are conferred. For example, position sense Functional MRI extrapolates regional brain activity from is probably a major contributor to disability in hemiparetic changes in blood oxygen-dependent signal. Two fMRI mo- patients. However, a lack of objective measures of such func- dalities have provided valuable insight to motor physiology, tion has limited study. Using KINARM, the characteristics of development, and stroke recovery. Task-dependent fMRI position senses dysfunction are increasingly well characterized compares regional activity between on and off states during in adult stroke18,19 We have shown that the same methods are performed motor tasks. Task fMRI has established funda- well tolerated by children where a standardized, validated mental principles of development and stroke recovery assessment of sensorimotor function can include two tasks including the roles of each hemisphere (above). Small hemi- that specifically target proprioception. Position sense can be paretic CP studies have demonstrated similar utility and measured using a position matching task; the robot moves the support our model. However, task fMRI only studies individual affected arm to random spatial locations and the subject then brain areas rather than the integrated neural networks. attempts to mirror match the positions with the unaffected “A connectivity-based system perspective is much closer to the arm. To measure kinaesthesia; the robot moves the affected neurobiology underlying brain physiology compared to analyzing arm. Subjects are required to move the good arm to match anatomically segregated regions (i.e. traditional fMRI).”246 Resting speed, amplitude and direction. Both tasks are completed first state fMRI (rsfMRI) is a recent advance that facilitates such with vision occluded and again without the blindfold to deter- network analysis.247 Simple acquisition of fMRI data during mine their ability to compensate with vision. Using these patient rest can be analyzed by both exploratory and methods, we recently demonstrated the common occurrence hypothesis-driven methods to create models of functional of such deficits in children with perinatal stroke and their as- and effective network connectivity.246,248 Such rsfMRI ap- sociation with clinical function.264 Such robotic outcomes will proaches now provide invaluable insight to brain plasticity be available tool in identifying new modulation targets and mechanisms in adult stroke.246,249,250 Resting state motor assessing outcomes from the interventional trials tat result. networks are measurable in young children251 but are rela- tively unstudied in perinatal stroke. The stark contrast pro- vided by arterial (cortical lesions at term) and PVI (preterm, 10. Future directions subcortical) creates an ideal human model to explore motor network development after perinatal brain injury. Further- The translational pipeline from proof of principle through the more, changes in functional networks likely represent an phases of early stage clinical trials to the accumulation of essential measure of intervention-induced plasticity. multiple, positive, multicentre fully powered trials must be Recently, such approaches have been validated in adults considered. The prolonged nature, and associated high costs, where change induced by brain stimulation interventions are likely comparable to drug development in some regards may be demonstrated using rsfMRI.223,252 though potentially more economical. For example, rTMS for 94 european journal of paediatric neurology 21 (2017) 75e103

adult major depression has driven the field of therapeutic functions in patients with stroke: a meta-analysis. Stroke non-invasive neuromodulation. Despite many positive ran- 2012;43(7):1849e57. domized controlled trials spanning more than a decade, the 2. Lindenberg R, Renga V, Zhu LL, Nair D, Schlaug G. Bihemispheric brain stimulation facilitates motor recovery intervention is only just now becoming an insured service in in chronic stroke patients. Neurology 2010;75:2176e84 many developed countries. The extremely common occur- (1526e632X (Electronic)). rence, poor response rates to current medications, and 3. Lefaucheur J-P, Andre-Obadia N, Antal A, Ayache SS, disabling nature of major depression have likely driven this Baeken C, Benninger DH, et al. Evidence-based guidelines on process relatively quickly. In comparison, treating different the therapeutic use of repetitive transcranial magnetic forms of cerebral palsy in children may not carry the same stimulation (rTMS). Clin Neurophysiol 2014;125(11):2150e206. weight, further slowing the process. This should not http://dx.doi.org/10.1016/j.clinph.2014.05.021. 4. Mineyko A, Kirton A. The black box of perinatal ischemic discourage effort but is a considered reason to optimize the stroke pathogenesis. J Child Neurol 2011;26(9):1154e62. design, exaction, and progress of early clinical trials. 5. Nelson KB. Perinatal ischemic stroke. Stroke 2007;38:742e5 The economics of such interventions in children have not (1524e4628 (Electronic)). been studied and remain difficult to estimate with such a 6. Petty GW, Brown Jr RD, Whisnant JP, Sicks JD, O'Fallon WM, limited number of trials to date. However, camp-based Wiebers DO. Ischemic stroke subtypes: a population-based intensive motor learning is both a likely prerequisite for study of incidence and risk factors. Stroke 1999;30:2513e6 (0039e2499 (Print)). neuromodulation in children with motor disability as well as 7. Raju TN, Nelson KB, Ferriero D, Lynch JK. Ischemic perinatal being an expensive, resource-intensive endeavour. Full day stroke: summary of a workshop sponsored by the National programs may be required to reach adequate dosing. Highly Institute of child health and human development and the qualified personnel are many for such programs, including National Institute of neurological disorders and stroke. skilled therapists to both deliver treatment and administer Pediatrics 2007;120:609e16. outcomes. We have tried to mitigate some of these costs with 8. DeVeber G, MacGregor D, Friefeld S, Chan A, Meanery B, after-school therapy programs and incorporating student Domi T. Neurological outcome in survivors of neonatal arterial ischemic stroke. Pediatr Res 2003;53(4):537A. therapists to administer therapy under the guidance of more 9. Mercuri E, Cowan F, Gupte G, Manning R, Laffan M, experienced therapists. Home-based therapy programs are Rutherford M, et al. Prothrombotic disorders and abnormal increasing in popularity for these reasons and may become neurodevelopmental outcome in infants with neonatal feasible as portable, simple stimulation devices are developed. cerebral infarction. Pediatrics 2001;107(1098e4275):1400e4. Such tDCS systems are already being designed including 10. Sreenan C, Bhargava R, Robertson CM. Cerebral infarction in patient-friendly operation and ability for remote program- the term newborn: clinical presentation and long- term e e ming and monitoring by study investigators. However, home outcome. J Pediatr 2000;137(0022 3476):351 5. 11. Kirton A, deVeber G. Advances in perinatal ischemic stroke. programs may risk removing the substantial psychosocial Pediatr Neurol 2009;40:205e14. benefits young people may receive from participating in camp 12. Kirton A, deVeber G, Pontigon AM, MacGregor D, Shroff M. based programs. These issues will need to be carefully Presumed perinatal ischemic stroke: vascular classification weighed in the design of future trials. predicts outcomes. Ann Neurol 2008;63(4):436e43. Neuromodulation trials in children with cerebral palsy 13. Domi T, deVeber G, Shroff M, Kouzmitcheva E, bring the possibility of precision medicine to paediatric MacGregor DL, Kirton A. Corticospinal tract pre-wallerian degeneration: a novel outcome predictor for pediatric stroke rehabilitation. While predictors of response to such in- on acute MRI. Stroke 2009;40:780e7. terventions have not yet been well defined, there is little doubt 14. Kirton A, Shroff M, Pontigon AM, deVeber G. Risk factors and that individual factors such as age, baseline disability, lesion presentations of periventricular venous infarction vs arterial type or location, and corticospinal tract arrangement likely presumed perinatal ischemic stroke. Arch Neurol play a role.265 A recent review suggested a possible systematic 2010;67:842e8 (1538e3687 (Electronic)). approach to classify hemiparetic children based on clinical 15. Tan M, Shroff M, Pontigon A, deVeber G, Kirton A. Role of (e.g. mirror movements), TMS (corticospinal tract orientation), occipital bone compression of the superior sagittal sinus in the pathophysiology of neonatal cerebral sinovenous and imaging (corticospinal tract integrity) variables.266 These thrombosis. Thromb Haemost 2007;5(2):577. are logical steps that are yet to be validated but carry the po- 16. Kirton A, Armstrong-Wells J, Chang T, deVeber G, Rivkin MJ, tential to bring personalized medicine to the rehabilitation of Hernandez M, et al. Symptomatic neonatal arterial ischemic children with hemiparetic CP. stroke: the international pediatric stroke study. Pediatrics 2011;128:e1402e10 (1098e4275 (Electronic)). 17. Payne ET, Wei XC, Kirton A. Reversible wall enhancement in pediatric cerebral arteriopathy. Can J Neurol Sci Conflict of interest 2011;38:139e40 (0317e1671 (Print)). 18. DeVetten G, Coutts SB, Hill MD, Goyal M, Eesa M, O'Brien B, et al. Acute corticospinal tract Wallerian degeneration is None. associated with stroke outcome. Stroke J Cereb Circ 2010;41(4):751e6. http://dx.doi.org/10.1161/ STROKEAHA.109.573287. references 19. Tan MA, Miller E, Shroff MM, Deveber GA, Kirton A. Alleviation of neonatal sinovenous compression to enhance cerebral venous blood flow. J Child Neurol. 2013 May;28(5):583e8. http://dx.doi.org/10.1177/ 1. Hsu WY, Cheng CH, Liao KK, Lee IH, Lin YY. Effects of 0883073812449689. Epub 2012 Jul 17. repetitive transcranial magnetic stimulation on motor european journal of paediatric neurology 21 (2017) 75e103 95

20. Trauner DA, Chase C, Walker P, Wulfeck B. Neurologic origin in the birth year period 1991e94. Acta Paediatr profiles of infants and children arter perinatal stroke. Pediatr 2001;90:271e7 (0803e5253 (Print)). Neurol 1993;9:383e6. 38. Nelson KB, Lynch JK. Stroke in newborn infants. Lancet 21. Golomb MR, MacGregor DL, Domi T, Armstrong DC, Neurol 2004;3(1474e4422):150e8. McCrindle BW, Mayank S, et al. Presumed pre- or perinatal 39. Nelson KB. Can we prevent cerebral palsy? N Engl J Med arterial ischemic stroke: risk factors and outcomes. Ann 2003;349:1765e9 (1533e4406 (Electronic)). Neurol 2001;50(0364e5134):163e8. 40. Mercuri E, Jongmans M, Bouza H, Haataja L, Rutherford M, 22. Uvebrant P. Hemiplegic cerebral palsy. Aetiology and Henderson S, et al. Congenital hemiplegia in children at outcome. Acta Paediatr Scand Suppl 1988;345:1e100 school age: assessment of hand function in the non- (0300e8843 (Print)). hemiplegic hand and correlation with MRI. Neuropediatrics 23. de Vries LS, Roelants-van Rijn AM, Rademaker KJ, van HI, 1999;30:8e13 (0174e304X (Print)). Beek FJ, Groenendaal F. Unilateral parenchymal 41. deVeber G., Kirton A., CPISR Study Group. Arterial ischemic haemorrhagic infarction in the preterm infant. Eur J Paediatr stroke in children. Ann Neurol n.d.;60(S10):115. Neurol 2001;5:139e49 (1090e3798 (Print)). 42. Goldenberg NA, Knapp-Clevenger R, Manco-Johnson MJ. 24. de Vries LS, Rademaker KJ, Groenendaal F, Eken P, van Elevated plasma factor VIII and D-dimer levels as predictors Haastert IC, Vandertop WP, et al. Correlation between of poor outcomes of thrombosis in children. N Engl J Med neonatal cranial ultrasound, MRI in infancy and 2004;351(1533e4406):1081e8. neurodevelopmental outcome in infants with a large 43. Mercuri E, Anker S, Guzzetta A, Barnett A, Haataja L, intraventricular haemorrhage with or without unilateral Rutherford M, et al. Neonatal cerebral infarction and visual parenchymal involvement. Neuropediatrics 1998;29:180e8 function at school age. Arch Dis Child Fetal Neonatal Ed (0174e304X (Print)). 2003;88(1359e2998):F487e91. 25. Ghi T, Simonazzi G, Perolo A, Savelli L, Sandri F, Bernardi B, 44. Bava S, Archibald SL, Trauner DA. Brain structure in et al. Outcome of antenatally diagnosed intracranial prenatal stroke: quantitative magnetic resonance imaging hemorrhage: case series and review of the literature. (MRI) analysis. J Child Neurol 2007;22:841e7 (0883e0738 Ultrasound Obstet Gynecol 2003;22:121e30 (0960e7692 (Print)). (Print)). 26. Elchalal U, Yagel S, Gomori JM, Porat S, Beni-Adani L, 45. Boardman JP, Ganesan V, Rutherford MA, Saunders DE, Yanai N, et al. Fetal intracranial hemorrhage (fetal stroke): Mercuri E, Cowan F. Magnetic resonance image does grade matter? Ultrasound Obstet Gynecol 2005;26:233e43 correlates of hemiparesis after neonatal and childhood (0960e7692 (Print)). middle cerebral artery stroke. Pediatrics 27. Takanashi J, Barkovich AJ, Ferriero DM, Suzuki H, Kohno Y. 2005;115(1098e4275):321e6. Widening spectrum of congenital hemiplegia: 46. Ganesan V, Ng V, Chong WK, Kirkham FJ, Connelly A. Lesion periventricular venous infarction in term neonates. volume, lesion location, and outcome after middle cerebral Neurology 2003;61:531e3 (1526e632X (Electronic)). artery territory stroke. Arch Dis Child 28. Takanashi J, Tada H, Barkovich AJ, Kohno Y. Magnetic 1999;81(0003e9888):295e300. resonance imaging confirms periventricular venous 47. De Vries LS, Van der Grond J, Van Haastert IC, infarction in a term-born child with congenital hemiplegia. Groenendaal F. Prediction of outcome in new-born infants Dev Med Child Neurol 2005;47:706e8 (0012e1622 (Print)). with arterial ischaemic stroke using diffusion-weighted 29. Kirton A, Wei X. Teaching neuroimages: confirmation of magnetic resonance imaging. Neuropediatrics prenatal periventricular venous infarction with 2005;36(1):12e20. http://dx.doi.org/10.1055/s-2005-837544. susceptibility-weighted MRI. Neurology 2010;74(12):e48. 48. Kirton A, Shroff M, Visvanathan T, deVeber G. Quantified 30. Abels L, Lequin M, Govaert P. Sonographic templates of corticospinal tract diffusion restriction predicts neonatal newborn perforator stroke. Pediatr Radiol 2006;36:663e9 stroke outcome. Stroke J Cereb Circ 2007;38(3):974e80. http:// (0301e0449 (Print)). dx.doi.org/10.1161/01.STR.0000258101.67119.72. 31. deVeber G, Andrew M. Canadian pediatric ischemic stroke 49. Lynch JK. Cerebrovascular disorders in children. Curr Neurol study group cerebral sinovenous thrombosis in children. N Neurosci Rep 2004;4(1528e4042):129e38. Engl J Med 2001;345(0028e4793):417e23. 50. McLinden A, Baird AD, Westmacott R, Anderson P, 32. Fitzgerald KC, Williams LS, Garg BP, Carvalho KS, deVeber G. Early cognitive outcome after neonatal stroke. J Golomb MR. Cerebral sinovenous thrombosis in the neonate. Child Neurol 2007 Sep;22(9):1111e6. Arch Neurol 2006;63:405e9 (0003e9942 (Print)). 51. Westmacott R, MacGregor D, Askalan R, deVeber G. Late 33. Lee J, Croen LA, Lindan C, Nash KB, Yoshida CK, Ferriero DM, emergence of cognitive deficits after unilateral neonatal et al. Predictors of outcome in perinatal arterial stroke: a stroke. Stroke 2009;40:2012e9 (1524e4628 (Electronic)). population-based study. Ann Neurol 52. Taylor TN, Davis PH, Torner JC, Holmes J, Meyer JW, 2005;58(0364e5134):303e8. Jacobson MF. Lifetime cost of stroke in the United States. 34. deVeber G, MacGregor D, Curtis R, Mayank S. Neurologic Stroke 1996;27(9):1459e66. outcome in survivors of childhood arterial ischemic stroke 53. Hoare B, Imms C, Carey L, Wasiak J. Constraint-induced and sinovenous thrombosis. J Child Neurol movement therapy in the treatment of the upper limb in 2000;15(0883e0738):316e24. children with hemiplegic cerebral palsy: a Cochrane 35. Mercuri E, Barnett A, Rutherford M, Guzzetta A, Haataja L, systematic review. Clin Rehabil 2007;21(8):675e85. Cioni G, et al. Neonatal cerebral infarction and neuromotor 54. Kennard MA. Age and other factors in motor recovery from outcome at school age. Pediatrics precentral lesions in monkeys. Am J Physiol 1936;115:137e46. 2004;113(1098e4275):95e100. 55. Trevathan WR. Introduction to evolutionary medicine. In: 36. Mercuri E, Rutherford M, Cowan F, Pennock J, Counsell S, Trevathan WR, Smith EO, McKenna JJ, editors. Evolutionary Papadimitriou M, et al. Early prognostic indicators of medicine, vol. 1. New York: Oxford University Press Inc; 1999. outcome in infants with neonatal cerebral infarction: a p. 3e6. clinical, electroencephalogram, and magnetic resonance 56. Eyre JA, Smith M, Dabydeen L, Clowry GJ, Petacchi E, imaging study. Pediatrics 1999;103(0031e4005):39e46. Battini R, et al. Is hemiplegic cerebral palsy equivalent to 37. Hagberg B, Hagberg G, Beckung E, Uvebrant P. Changing amblyopia of the corticospinal system? Ann Neurol panorama of cerebral palsy in Sweden. VIII. Prevalence and 2007;62(5):493e503. 96 european journal of paediatric neurology 21 (2017) 75e103

57. Eyre JA. Corticospinal tract development and its plasticity reorganization in pediatric hemiplegic patients. after perinatal injury. Neurosci Biobehav Rev 2007;31:1136e49 Neuropediatrics 1997;28:168e74 (0174e304X (Print)). (0149e7634 (Print)). 75. Nezu A, Kimura S, Takeshita S, Tanaka M. Functional 58. Martin JH, Friel KM, Salimi I, Chakrabarty S. Activity- and recovery in hemiplegic cerebral palsy: ipsilateral use-dependent plasticity of the developing corticospinal electromyographic responses to focal transcranial magnetic system. Neurosci Biobehav Rev 2007;31(8):1125e35. http:// stimulation. Brain Dev 1999;21:162e5 (0387e7604 (Print)). dx.doi.org/10.1016/j.neubiorev.2007.04.017. 76. Vandermeeren Y, Sebire G, Grandin CB, Thonnard JL, 59. Werhahn KJ, Conforto AB, Kadom N, Hallett M, Cohen LG. Schlogel X, De Volder AG. Functional reorganization of brain Contribution of the ipsilateral motor cortex to recovery after in children affected with congenital hemiplegia: fMRI study. chronic stroke. Ann Neurol 2003;54:464e72 (0364e5134 Neuro Image 2003;20(1):289e301. (Print)). 77. Yue Cao, Vikingstad EM, Huttenlocher PR, Towle VL, 60. Foltys H, Krings T, Meister IG, Sparing R, Boroojerdi B, Levin DN. Functional magnetic resonance studies of Thron A, et al. Motor representation in patients rapidly reorganization of the human hand sensorimotor area after recovering after stroke: a functional magnetic resonance unilateral brain injury in the perinatal period. Proc Natl Acad imaging and transcranial magnetic stimulation study. Clin Sci U. S. A 1994;91:9612e6. Neurophysiol 2003;114:2404e15 (1388e2457 (Print)). 78. Macdonell RA, Jackson GD, Curatolo JM, Abbott DF, 61. Ward NS, Cohen LG. Mechanisms underlying recovery of Berkovic SF, Carey LM, et al. Motor cortex localization using motor function after stroke. Arch Neurol functional MRI and transcranial magnetic stimulation. 2004;61(0003e9942):1844e8. Neurology 1999;53:1462e7 (0028e3878 (Print)). 62. Calautti C, Leroy F, Guincestre JY, Marie RM, Baron JC. 79. Thickbroom GW, Byrnes ML, Archer SA, Nagarajan L, Sequential activation brain mapping after subcortical stroke: Mastaglia FL. Differences in sensory and motor cortical changes in hemispheric balance and recovery. Neuroreport organization following brain injury early in life. Ann Neurol 2001;12:3883e6 (0959e4965 (Print)). 2001;49(0364e5134):320e7. 63. Rossini PM, Calautti C, Pauri F, Baron JC. Post-stroke plastic 81. Eyre JA, Taylor JP, Villagra F, Smith M, Miller S. Evidence of reorganisation in the adult brain. Lancet Neurol activity-dependent withdrawal of corticospinal projections 2003;2:493e502 (1474e4422 (Print)). during human development. Neurology 2001;57:1543e54 64. Cioni G, Montanaro D, Tosetti M, Canapicchi R, (0028e3878 (Print)). Ghelarducci B. Reorganisation of the sensorimotor cortex 82. Nirkko AC, Rosler KM, Ozdoba C, Heid O, Schroth G, after early focal brain lesion: a functional MRI study in Hess CW. Human cortical plasticity: functional recovery monozygotic twins. Neuroreport 2001;12:1335e40 (0959e4965 with mirror movements. Neurology 1997;48:1090e3 (Print)). (0028e3878 (Print)). 65. Kirton A, Chen R, Friefeld S, Gunraj C, Pontigon AM, 84. Ragazzoni A, Cincotta M, Borgheresi A, Zaccara G, deVeber G. Contralesional repetitive transcranial magnetic Ziemann U. Congenital hemiparesis: different functional stimulation for chronic hemiparesis in subcortical paediatric reorganization of somatosensory and motor pathways. Clin stroke: a randomised trial. Lancet Neurol 2008;7(6):507e13. Neurophysiol 2002;113:1273e8 (1388e2457 (Print)). 66. Kirton A, deVeber G, Gunraj C, Chen R. Cortical excitability 85. Rutten GJ, Ramsey NF, van Rijen PC, Franssen H, van and interhemispheric inhibition after subcortical pediatric Veelen CW. Interhemispheric reorganization of motor hand stroke: plastic organization and effects of rTMS. Clin function to the primary motor cortex predicted with Neurophysiol 2010;121:1922e9 (1872e8952 (Electronic)). functional magnetic resonance imaging and transcranial 67. Berweck S, Walther M, Brodbeck V, Wagner N, Koerte I, magnetic stimulation. J Child Neurol 2002;17:292e7 Henschel V, et al. Abnormal motor cortex excitability in (0883e0738 (Print)). congenital stroke. Pediatr Res 2008;63:84e8. 86. Staudt M. (Re-)organization of the developing human brain 68. Staudt M, Gerloff C, Grodd W, Holthausen H, Niemann G, following periventricular white matter lesions. Neurosci Krageloh-Mann I. Reorganization in congenital hemiparesis Biobehav Rev 2007;31(8):1150e6. acquired at different gestational ages. Ann Neurol 87. Hummel F, Cohen LG. Improvement of motor function with 2004;56:854e63 (0364e5134 (Print)). noninvasive cortical stimulation in a patient with chronic 69. Staudt M, Grodd W, Gerloff C, Erb M, Stitz J, Krageloh-Mann I. stroke. Neurorehabil Neural Repair 2005;19:14e9 (1545e9683 Two types of ipsilateral reorganization in congenital (Print)). hemiparesis: a TMS and fMRI study. Brain 88. Hummel F, Celnik P, Giraux P, Floel A, Wu WH, Gerloff C, 2002;125(0006e8950):2222e37. et al. Effects of non-invasive cortical stimulation on skilled 70. Juenger H, Linder-Lucht M, Walther M, Berweck S, Mall V, motor function in chronic stroke. Brain 2005;128:490e9 Staudt M. Cortical neuromodulation by constraint- (1460e2156 (Electronic)). induced movement therapy in congenital hemiparesis: an 89. Fregni F, Boggio PS, Mansur CG, Wagner T, Ferreira MJ, FMRI study. Neuropediatrics 2007;38:130e6(0174e304X Lima MC, et al. Transcranial direct current stimulation of the (Print)). unaffected hemisphere in stroke patients. Neuroreport 71. Walther M, Juenger H, Kuhnke N, Wilke M, Brodbeck V, 2005;16(0959e4965):1551e5. Berweck S, et al. Motor cortex plasticity in ischemic perinatal 90. Takeuchi N, Tada T, Toshima M, Chuma T, Matsuo Y, stroke: a transcranial magnetic stimulation and functional Ikoma K. Inhibition of the unaffected motor cortex by 1 Hz MRI study. Pediatr Neurol 2009;41(3):171e8. repetitive transcranical magnetic stimulation enhances 72. Farmer SF, Harrison LM, Ingram DA, Stephens JA. Plasticity motor performance and training effect of the paretic hand in of central motor pathways in children with hemiplegic patients with chronic stroke. J Rehabil Med 2008;40:298e303 cerebral palsy. Neurology 1991;41:1505e10 (0028e3878 (1650e1977 (Print)). (Print)). 91. Schambra HM, Sawaki L, Cohen LG. Modulation of 73. Carr LJ, Harrison LM, Evans AL, Stephens JA. Patterns of excitability of human motor cortex (M1) by 1 Hz transcranial central motor reorganization in hemiplegic cerebral palsy. magnetic stimulation of the contralateral M1. Clin Brain 1993;116:1223e47 (Pt 5)(0006e8950 (Print)). Neurophysiol 2003;114:130e3 (1388e2457 (Print)). 74. Maegaki Y, Maeoka Y, Ishii S, Shiota M, Takeuchi A, 92. Pal PK, Hanajima R, Gunraj CA, Li JY, Wagle-Shukla A, Yoshino K, et al. Mechanisms of central motor Morgante F, et al. Effect of low-frequency repetitive european journal of paediatric neurology 21 (2017) 75e103 97

transcranial magnetic stimulation on interhemispheric 112. Bouza H, Rutherford M, Acolet D, Pennock JM, inhibition. J Neurophysiol 2005;94(3):1668e75. Dubowitz LM. Evolution of early hemiplegic signs in full- 93. Mall V, Berweck S, Fietzek UM, Glocker FX, Oberhuber U, term infants with unilateral brain lesions in the neonatal Walther M, et al. Low level of intracortical inhibition in period: a prospective study. Neuropediatrics 1994;25:201e7 children shown by transcranial magnetic stimulation. (0174e304X (Print)). Neuropediatrics 2004;35:120e5 (0174e304X (Print)). 113. Li Q, Martin JH. Postnatal development of corticospinal axon 94. Garvey MA, Ziemann U, Bartko JJ, Denckla MB, Barker CA, terminal morphology in the cat. J Comp Neurol Wassermann EM. Cortical correlates of neuromotor 2001;435:127e41 (0021e9967 (Print)). development in healthy children. Clin Neurophysiol 114. Eyre JA, Miller S, Clowry GJ, Conway EA, Watts C. Functional 2003;114:1662e70 (1388e2457 (Print)). corticospinal projections are established prenatally in the 95. Liepert J, Hamzei F, Weiller C. Motor cortex disinhibition of human foetus permitting involvement in the development the unaffected hemisphere after acute stroke. Muscle Nerve of spinal motor centres. Brain 2000;123:51e64 (Pt 2000;23:1761e3 (0148e639X (Print)). 1)(0006e8950 (Print)). 96. Murase N, Duque J, Mazzocchio R, Cohen LG. Influence of 115. Wolf SL, Winstein CJ, Miller JP, Thompson PA, Taub E, interhemispheric interactions on motor function in chronic Uswatte G, et al. Retention of upper limb function in stroke stroke. Ann Neurol 2004;55:400e9 (0364e5134 (Print)). survivors who have received constraint-induced movement 97. Liepert J, Restemeyer C, Kucinski T, Zittel S, Weiller C. Motor therapy: the EXCITE randomised trial. Lancet Neurol strokes: the lesion location determines motor excitability 2008;7(1):33e40. changes. Stroke 2005;36(1524e4628):2648e53. 116. Daly JJ, Wolpaw JR. Brainecomputer interfaces in 98. Cicinelli P, Pasqualetti P, Zaccagnini M, Traversa R, Oliveri M, neurological rehabilitation. Lancet Neurol 2008;7(11):1032e43. Rossini PM. Interhemispheric asymmetries of motor cortex http://dx.doi.org/10.1016/S1474-4422(08)70223-0. excitability in the postacute stroke stage: a paired-pulse 117. Taub E, Crago JE, Uswatte G. Constraint-induced movement transcranial magnetic stimulation study. Stroke therapy: a new approach to treatment in physical 2003;34:2653e8 (1524e4628 (Electronic)). rehabilitation. Rehabil Psychol 1998;43(2):152e70. 99. Liepert J, Storch P, Fritsch A, Weiller C. Motor cortex 118. Taub E, Uswatte G, Pidikiti R. Constraint-induced movement disinhibition in acute stroke. Clin Neurophysiol therapy: a new family of techniques with broad application 2000;111:671e6 (1388e2457 (Print)). to physical rehabilitation e a clinical review. J Rehabil Res Dev 100. Ziemann U, Corwell B, Cohen LG. Modulation of plasticity 1999;36(3):237e51. in human motor cortex after forearm ischemic nerve block. 119. Taub E, Uswatte G, Elbert T. New treatments in J Neurosci 1998;18:1115e23 (0270e6474 (Print)). neurorehabilitation founded on basic research. Nat Rev 101. Kirton A, deVeber G, Gunraj C, Friefeld S, Chen R. Cortical Neurosci 2002;3(3):228e36. excitability and interhemispheric inhibition in subcortical 120. Wolf SL, Winstein CJ, Miller JP, Taub E, Uswatte G, Morris D, pediatric stroke: chronic reorganization and effects of et al. Effect of constraint-induced movement therapy on contralesional inhibitory rTMS. Brain Stimul 2008;1(3):271. upper extremity function 3 to 9 months after stroke: the 102. Armand J, Olivier E, Edgley SA, Lemon RN. Postnatal EXCITE randomized clinical trial. J Am Med Assoc development of corticospinal projections from motor cortex 2006;296:2095e104 (1538e3598 (Electronic)). to the cervical enlargement in the macaque monkey. J 121. Hakkennes S, Keating JL. Constraint-induced movement Neurosci 1997;17:251e66 (0270e6474 (Print)). therapy following stroke: a systematic review of randomised 103. Koh TH, Eyre JA. Maturation of corticospinal tracts assessed controlled trials. Aust J Physiother 2005;51(4):221e31. by electromagnetic stimulation of the motor cortex. Arch Dis 122. Eliasson AC, Krumlinde-sundholm L, Shaw K, Wang C. Child 1988;63:1347e52 (1468e2044 (Electronic)). Effects of constraint-induced movement therapy in young 104. Chakrabarty S, Martin JH. Postnatal development of the children with hemiplegic cerebral palsy: an adapted model. motor representation in primary motor cortex. J Neurophysiol Dev Med Child Neurol 2005;47(4):266e75. 2000;84:2582e94 (0022e3077 (Print)). 123. Taub E, Ramey SL, DeLuca S, Echols K. Efficacy of constraint- 105. Guzzetta A, Staudt M, Petacchi E, Ehlers J, Erb M, Wilke M, induced movement therapy for children with cerebral palsy et al. Brain representation of active and passive hand with asymmetric motor impairment. Pediatrics movements in children. Pediatr Res 2007;61:485e90 2004;113(1098e4275):305e12. (0031e3998 (Print)). 124. Taub E, Griffin A, Nick J, Gammons K, Uswatte G, Law CR. 106. Martin JH, Lee SJ. Activity-dependent competition between Pediatric CI therapy for stroke-induced hemiparesis in developing corticospinal terminations. Neuroreport young children. Dev Neurorehabilitation 2007;10(1):3e18. 1999;10:2277e82. 125. Willis JK, Morello A, Davie A, Rice JC, Bennett JT. Forced use 107. Martin JH, Kably B, Hacking A. Activity-dependent treatment of childhood hemiparesis. Pediatrics development of cortical axon terminations in the spinal cord 2002;110:94e6 (1098e4275 (Electronic)). and brain stem. Exp Brain Res 1999;125:184e99. 126. Charles J, Gordon AM. Development of hand-arm bimanual 108. Salimi I, Martin JH. Rescuing transient corticospinal intensive training (HABIT) for improving bimanual terminations and promoting growth with corticospinal coordination in children with hemiplegic cerebral palsy. Dev stimulation in kittens. J Neurosci 2004;24:4952e61 (1529e2401 Med Child Neurol 2006;48:931e6 (0012e1622 (Print)). (Electronic)). 127. Deluca SC, Echols K, Law CR, Ramey SL. Intensive pediatric 109. Scales DA, Collins GH. Cerebral degeneration with constraint-induced therapy for children with cerebral palsy: hypertrophy of the contralateral pyramid. Arch Neurol randomized, controlled, crossover trial. J Child Neurol 1972;26:186e90 (0003e9942 (Print)). 2006;21(11):931e8. 110. Bax M, Goldstein M, Rosenbaum P, Leviton A, Paneth N, 128. Sung IY, Ryu JS, Pyun SB, Yoo SD, Song WH, Park MJ. Efficacy Dan B, et al. Proposed definition and classification of of forced-use therapy in hemiplegic cerebral palsy. Arch Phys cerebral palsy, April 2005. Dev Med Child Neurol Med Rehabil 2005;86(11):2195e8. 2005;47(8):571e6. 129. Sakzewski L, Ziviani J, Boyd R. Systematic review and meta- 111. Kirton A, deVeber G. Cerebral palsy secondary to perinatal analysis of therapeutic management of upper-limb ischemic stroke. Clin Perinatol 2006;33:367e86 (0095e5108 dysfunction in children with congenital hemiplegia. (Print)). Pediatrics 2009;123(6):e1111e22. 98 european journal of paediatric neurology 21 (2017) 75e103

130. Brady K, Garcia T. Constraint-induced movement therapy 148. Skold A, Josephsson S, Eliasson AC. Performing bimanual (CIMT): pediatric applications. Dev Disabil Res Rev activities: the experiences of young persons with hemiplegic 2009;15(2):102e11. cerebral palsy. Am J Occup Ther 2004;58(4):416e25. 131. Liepert J, Bauder H, Wolfgang HR, Miltner WH, Taub E, 149. Kuhnke N, Juenger H, Walther M, Berweck S, Mall V, Weiller C. Treatment-induced cortical reorganization after Staudt M. Do patients with congenital hemiparesis and stroke in humans. Stroke 2000;31(0039e2499):1210e6. ipsilateral corticospinal projections respond differently to 132. Schaechter JD, Kraft E, Hilliard TS, Dijkhuizen RM, Benner T, constraint-induced movement therapy? Dev Med Child Neurol Finklestein SP, et al. Motor recovery and cortical 2008;50(12):898e903. reorganization after constraint-induced movement therapy 150. Khedr EM, Ahmed MA, Fathy N, Rothwell JC. Therapeutic in stroke patients: a preliminary study. Neurorehabil Neural trial of repetitive transcranial magnetic stimulation after Repair 2002;16:326e38 (1545e9683 (Print)). acute ischemic stroke. Neurology 2005;65(3):466e8. 133. Wittenberg GF, Chen R, Ishii K, Bushara KO, Eckloff S, 151. Naeser MA, Martin PI, Nicholas M, Baker EH, Seekins H, Croarkin E, et al. Constraint-induced therapy in stroke: Kobayashi M, et al. Improved picture naming in chronic magnetic-stimulation motor maps and cerebral activation. aphasia after TMS to part of right Broca's area: an open- Neurorehabil Neural Repair 2003;17:48e57 (1545e9683 (Print)). protocol study. Brain Lang 2005;93:95e105. 134. Sutcliffe TL, Gaetz WC, Logan WJ, Cheyne DO, Fehlings DL. 152. Pascual-Leone A, Tormos JM, Keenan J, Tarazona F, Cortical reorganization after modified constraint-induced Canete C, Catala MD. Study and modulation of human movement therapy in pediatric hemiplegic cerebral palsy. J cortical excitability with transcranial magnetic stimulation. J Child Neurol 2007;22:1281e7 (0883e0738 (Print)). Clin Neurophysiol 1998;15(0736e0258):333e43. 135. Johansen-Berg H, Dawes H, Guy C, Smith SM, Wade DT, 153. Plautz EJ, Barbay S, Frost SB, Friel KM, Dancause N, Matthews PM. Correlation between motor improvements Zoubina EV, et al. Post-infarct cortical plasticity and and altered fMRI activity after rehabilitative therapy. Brain behavioral recovery using concurrent cortical stimulation 2002;125:2731e42 (0006e8950 (Print)). and rehabilitative training: a feasibility study in primates. 136. Hodics T, Cohen LG. Functional neuroimaging in motor Neurol Res 2003;25:801e10 (0161e6412 (Print)). recovery after stroke. Top Stroke Rehabil 154. Adkins-Muir DL, Jones TA. Cortical electrical stimulation 2005;12(1074e9357):15e21. combined with rehabilitative training: enhanced functional 137. Hodics T, Cohen LG, Cramer SC. Functional imaging of recovery and dendritic plasticity following focal cortical intervention effects in stroke motor rehabilitation. Arch Phys ischemia in rats. Neurol Res 2003;25:780e8 (0161e6412 Med Rehabil 2006;87:S36e42 (12 Suppl 2). (Print)). 138. Liepert J, Miltner WH, Bauder H, Sommer M, Dettmers C, 155. Kleim JA, Bruneau R, VandenBerg P, MacDonald E, Taub E, et al. Motor cortex plasticity during constraint- Mulrooney R, Pocock D. Motor cortex stimulation enhances induced movement therapy in stroke patients. Neurosci Lett motor recovery and reduces peri-infarct dysfunction 1998;250:5e8 (0304e3940 (Print)). following ischemic insult. Neurol Res 2003;25:789e93 139. Liepert J, Graef S, Uhde I, Leidner O, Weiller C. Training- (0161e6412 (Print)). induced changes of motor cortex representations in stroke 156. Teskey GC, Flynn C, Goertzen CD, Monfils MH, Young NA. patients. Acta Neurol Scand 2000;101:321e6 (0001e6314 (Print)). Cortical stimulation improves skilled forelimb use following 140. Liepert J. Motor cortex excitability in stroke before and after a focal ischemic infarct in the rat. Neurol Res 2003;25:794e800 constraint-induced movement therapy. Cogn Behav Neurol (0161e6412 (Print)). 2006;19(1):41e7. 157. Pascual-Leone A, Amedi A, Fregni F, Merabet LB. The plastic 141. Hamzei F, Liepert J, Dettmers C, Weiller C, Rijntjes M. Two human brain cortex. Annu Rev Neurosci 2005;28:377e401. different reorganization patterns after rehabilitative 158. Maeda F, Keenan JP, Tormos JM, Topka H, Pascual-Leone A. therapy: an exploratory study with fMRI and TMS. Modulation of corticospinal excitability by repetitive Neuroimage 2006;31:710e20 (1053e8119 (Print)). transcranial magnetic stimulation. Clin Neurophysiol 142. Liepert J. Transcranial magnetic stimulation in 2000;111:800e5 (1388e2457 (Print)). neurorehabilitation. Acta Neurochir Suppl 2005;93:71e4 159. Chen R, Classen J, Gerloff C, Celnik P, Wassermann EM, (Journal Article). Hallett M, et al. Depression of motor cortex excitability by 143. Ro T, Noser E, Boake C, Johnson R, Gaber M, Speroni A, et al. low-frequency transcranial magnetic stimulation. Neurology Functional reorganization and recovery after constraint- 1997;48(0028e3878):1398e403. induced movement therapy in subacute stroke: case reports. 160. Lisanby SH, Gutman D, Luber B, Schroeder C, Sackeim HA, Neurocase 2006;12(1):50e60. Sham TMS. Intracerebral measurement of the induced 144. Tarkka IM, Kononen M, Pitkanen K, Sivenius J, Mervaalat E. electrical field and the induction of motor-evoked potentials. Alterations in cortical excitability in chronic stroke after Biol Psychiatry 2001;49(5):460e3. constraint-induced movement therapy. Neurol Res 161. Speer AM, Benson BE, Kimbrell TK, Wassermann EM, 2008;30(5):504e10. Willis MW, Herscovitch P, et al. Opposite effects of high and 145. Gordon AM, Schneider JA, Chinnan A, Charles JR. Efficacy of low frequency rTMS on mood in depressed patients: a hand-arm bimanual intensive therapy (HABIT) in children relationship to baseline cerebral activity on PET. J Affect with hemiplegic cerebral palsy: a randomized control trial. Disord 2009;115:386e94 (1573e2517 (Electronic)). Dev Med Child Neurol 2007;49(11):830e8. 162. Takeuchi N, Chuma T, Matsuo Y, Watanabe I, Ikoma K. 146. Gordon AM, Hung YC, Brandao M, Ferre CL, Kuo HC, Friel K, Repetitive transcranial magnetic stimulation of et al. Bimanual training and constraint-induced movement contralesional primary motor cortex improves hand therapy in children with hemiplegic cerebral palsy: a function after stroke. Stroke 2005;36(12):2681e6. randomized trial. Neurorehabil Neural Repair 163. Mansur CG, Fregni F, Boggio PS, Riberto M, Gallucci-Neto J, 2011;25(8):692e702. Santos CM, et al. A sham stimulation-controlled trial of 147. Gordon AM, Chinnan A, Gill S, Petra E, Hung YC, Charles J. rTMS of the unaffected hemisphere in stroke patients. Both constraint-induced movement therapy and bimanual Neurology 2005;64(10):1802e4. training lead to improved performance of upper extremity 164. Quintana H. Transcranial magnetic stimulation in persons function in children with hemiplegia. Dev Med Child Neurol younger than the age of 18. JECT 2005;21:88e95 (1095e0680 2008;50(12):957e8. (Print)). european journal of paediatric neurology 21 (2017) 75e103 99

165. Rossini PM, Desiato MT, Caramia MD. Age-related changes clinical studies in brain stimulation. Clin Neurophysiol of motor evoked potentials in healthy humans: non-invasive 2006;117:845e50 (1388e2457 (Print)). evaluation of central and peripheral motor tracts excitability 181. Reis J, Robertson EM, Krakauer JW, Rothwell J, Marshall L, and conductivity. Brain Res 1992;593:14e9. Gerloff C, et al. Consensus: can transcranial direct current 166. Lin KL, Pascual-Leone A. Transcranial magnetic stimulation stimulation and transcranial magnetic stimulation enhance and its applications in children. Chang Gung Med J motor learning and memory formation? Brain Stimul 2002;25(7):424e36. 2008;1:363e9 (1935e861X (Print)). 167. Liebetanz D, Fauser S, Michaelis T, Czeh B, Watanabe T, 182. Reis J, Schambra HM, Cohen LG, Buch ER, Fritsch B, Zarahn E, Paulus W, et al. Safety aspects of chronic low-frequency et al. Noninvasive cortical stimulation enhances motor skill transcranial magnetic stimulation based on localized proton acquisition over multiple days through an effect on magnetic resonance spectroscopy and histology of the rat consolidation. Proc Natl Acad Sci USA 2009;106(5):1590e5. brain. J Psychiatr Res 2003;37:277e86 (0022e3956 (Print)). 183. Fritsch B, Reis J, Martinowich K, Schambra HM, Ji Y, 168. Naeser MA, Martin PI, Nicholas M, Baker EH, Seekins H, Cohen LG, et al. Direct current stimulation promotes BDNF- Helm-Estabrooks N, et al. Improved naming after TMS dependent synaptic plasticity: potential implications for treatments in a chronic, global aphasia patientecase report. motor learning. Neuron 2010;66:198e204 (1097e4199 Neurocase 2005;11:182e93. (Electronic)). 169. Kauffmann CD, Cheema MA, Miller BE. Slow right prefrontal 184. Boggio PS, Alonso-Alonso M, Mansur CG, Rigonatti SP, transcranial magnetic stimulation as a treatment for Schlaug G, Pascual-Leone A, et al. Hand function medication-resistant depression: a double-blind, placebo- improvement with low-frequency repetitive transcranial controlled study. Depress Anxiety 2004;19:59e62 (1091e4269 magnetic stimulation of the unaffected hemisphere in a (Print)). severe case of stroke. Am J Phys Med Rehabil 2006;85:927e30 170. Klein E, Kolsky Y, Puyerovsky M, Koren D, Chistyakov A, (0894e9115 (Print)). Feinsod M. Right prefrontal slow repetitive transcranial 185. Vines BW, Nair DG, Schlaug G. Contralateral and ipsilateral magnetic stimulation in schizophrenia: a double-blind motor effects after transcranial direct current stimulation. sham-controlled pilot study. Biol Psychiatry 1999;46:1451e4 Neuroreport 2006;17:671e4 (0959e4965 (Print)). (0006e3223 (Print)). 186. Hummel FC, Heise K, Celnik P, Floel A, Gerloff C, Cohen LG. 171. Padberg F, Zwanzger P, Thoma H, Kathmann N, Haag C, Facilitating skilled right hand motor function in older Greenberg BD, et al. Repetitive transcranial magnetic subjects by anodal polarization over the left primary motor stimulation (rTMS) in pharmacotherapy-refractory major cortex. Neurobiol Aging 2010 Dec;31(12):2160e8. http:// depression: comparative study of fast, slow and sham rTMS. dx.doi.org/10.1016/j.neurobiolaging.2008.12.008. Epub 2009 Psychiatry Res 1999;88:163e71 (0165e1781 (Print)). Feb 6. 172. Pascual-Leone A, Rubio B, Pallardo F, Catala MD. Rapid-rate 187. Nitsche MA, Schauenburg A, Lang N, Liebetanz D, Exner C, transcranial magnetic stimulation of left dorsolateral Paulus W, et al. Facilitation of implicit motor learning by prefrontal cortex in drug-resistant depression. Lancet weak transcranial direct current stimulation of the primary 1996;348:233e7 (0140e6736 (Print)). motor cortex in the human. J Cogn Neurosci 2003;15:619e26 173. Wassermann EM. Risk and safety of repetitive transcranial (0898e929X (Print)). magnetic stimulation: report and suggested guidelines from 188. Nitsche MA, Paulus W. Sustained excitability elevations the international workshop on the safety of repetitive induced by transcranial DC motor cortex stimulation in transcranial magnetic stimulation, June 5e7, 1996. humans. Neurology 2001;57:1899e901 (0028e3878 (Print)). Electroencephalogr Clin Neurophysiol 1998;108(1):1e16. 189. Hummel FC, Voller B, Celnik P, Floel A, Giraux P, Gerloff C, 174. Kirton A, Andersen J, Herrero M, Nettel-Aguirre A, Carsolio L, et al. Effects of brain polarization on reaction times and Damji O, et al. Brain stimulation and constraint for perinatal pinch force in chronic stroke. BMC Neurosci 2006;7:73 stroke hemiparesis: the PLASTIC CHAMPS trial. Neurology (1471e2202 (Electronic)). 2016 May 3;86(18):1659e67. http://dx.doi.org/10.1212/ 190. Boggio PS, Nunes A, Rigonatti SP, Nitsche MA, Pascual- WNL.0000000000002646. Epub 2016 Mar 30. Leone A, Fregni F. Repeated sessions of noninvasive brain 175. Zaghi S, Acar M, Hultgren B, Boggio PS, Fregni F. Noninvasive DC stimulation is associated with motor function brain stimulation with low-intensity electrical currents: improvement in stroke patients. Restor Neurol Neurosci putative mechanisms of action for direct and alternating 2007;25:123e9 (0922e6028 (Print)). current stimulation. Neuroscientist 2010;16:285e307 191. Moliadze V, Schmanke T, Andreas S, Lyzhko E, Freitag CM, (1089e4098 (Electronic)). Siniatchkin M. Stimulation intensities of transcranial direct 176. Nitsche MA, Paulus W. Excitability changes induced in the current stimulation have to be adjusted in children and human motor cortex by weak transcranial direct current adolescents. Clin Neurophysiol Off J Int Fed Clin Neurophysiol stimulation. J Physiol 2000;527:633e9. Pt 3(0022e3751 (Print)). 2015;126(7):1392e9. http://dx.doi.org/10.1016/ 177. Brunoni AR, Amadera J, Berbel B, Volz MS, Rizzerio BG, j.clinph.2014.10.142. Fregni F. A systematic review on reporting and assessment 192. Ciechanski P, Kirton A. Transcranial direct-current of adverse effects associated with transcranial direct current stimulation can enhance motor learning in children. Cereb stimulation. Int J Neuropsychopharmacol 2011:1e13 Cortex N. Y N 2016;1991. http://dx.doi.org/10.1093/cercor/ (1469e5111 (Electronic)). bhw114. 178. Nitsche MA, Cohen LG, Wassermann EM, Priori A, Lang N, 193. Kessler SK, Minhas P, Woods AJ, Rosen A, Gorman C, Antal A, et al. Transcranial direct current stimulation: state Bikson M. Dosage considerations for transcranial direct of the art 2008. Brain Stimul 2008;1:206e23 (1935e861X current stimulation in children: a computational modeling (Print)). study. PloS One 2013;8(9):e76112. http://dx.doi.org/10.1371/ 179. Poreisz C, Boros K, Antal A, Paulus W. Safety aspects of journal.pone.0076112. transcranial direct current stimulation concerning healthy 194. Gillick BT, Kirton A, Carmel JB, Minhas P, Bikson M. Pediatric subjects and patients. Brain Res Bull 2007;72:208e14 stroke and transcranial direct current stimulation: methods (0361e9230 (Print)). for rational individualized dose optimization. Front Hum 180. Gandiga PC, Hummel FC, Cohen LG. Transcranial DC Neurosci 2014;8:739. http://dx.doi.org/10.3389/ stimulation (tDCS): a tool for double-blind sham-controlled fnhum.2014.00739. 100 european journal of paediatric neurology 21 (2017) 75e103

195. Mahmoudi H, Borhani HA, Petramfar P, Jahanshahi S, 213. McColl MA, Paterson M, Davies D, Doubt L, Law M. Validity Salehi Z, Fregni F. Transcranial direct current stimulation: and community utility of the Canadian occupational electrode montage in stroke. Disabil Rehabil 2011;33:1383e8 performance measure. Can J Occup Ther Rev Can Ergother (1464e5165 (Electronic)). 2000;67(1):22e30. 196. Schulz KF, Moher D, Altman DG. CONSORT 2010 comments. 214. Carswell A, McColl MA, Baptiste S, Law M, Polatajko H, Lancet Lond Engl 2010;376(9748):1222e3. http://dx.doi.org/ Pollock N. The Canadian occupational performance 10.1016/S0140-6736(10)61879-X. measure: a research and clinical literature review. Can J 197. Plint AC, Moher D, Morrison A, Schulz K, Altman DG, Hill C, Occup Ther Rev Can Ergother 2004;71(4):210e22. et al. Does the CONSORT checklist improve the quality of 215. Hanna K, Rodger S. Towards family-centred practice in reports of randomised controlled trials? A systematic paediatric occupational therapy: a review of the literature on review. Med J Aust 2006;185(5):263e7. parent-therapist collaboration. Aust Occup Ther J 198. Arner M., Beckung E., Eliasson A., Krumlinde-sundholm L., 2002;49(1):14e24. Rosenbaum P., Rosblad B. Manual ability classification 216. Cusick A, Lannin NA, Lowe K. Adapting the Canadian system for children with cerebral palsy. Available at: http:// occupational performance measure for use in a paediatric www.macs.nu/. clinical trial. Disabil Rehabil 2007;29(10):761e6. 199. World Health Organization International classification of 217. Sakzewski L, Boyd R, Ziviani J. Clinimetric properties of functioning, disability and health (ICF). Genevea (Switzerland): participation measures for 5- to 13-year-old children with World Health Organization; 2001. cerebral palsy: a systematic review. Dev Med Child Neurol 200. Gilmore R, Sakzewski L, Boyd R. Upper limb activity 2007;49(3):232e40. http://dx.doi.org/10.1111/j.1469- measures for 5- to 16-year-old children with congenital 8749.2007.00232.x. hemiplegia: a systematic review. Dev Med Child Neurol 218. Wallen M, O'Flaherty SJ, Waugh M-CA. Functional outcomes 2010;52(1):14e21. of intramuscular Botulinum toxin type a and occupational 201. Greaves S, Imms C, Dodd K, Krumlinde-sundholm L. therapy in the upper limbs of children with cerebral palsy: a Assessing bimanual performance in young children with randomized controlled trial. Arch Phys Med Rehabil hemiplegic cerebral palsy: a systematic review. Dev Med 2007;88(1):1e10. Child Neurol 2010;52(5):413e21. 219. Ostensjo S, Oien I, Fallang B. Goal-oriented rehabilitation of 202. Krumlinde-sundholm L, Holmefur M, Kottorp A, preschoolers with cerebral palsyea multi-case study of Eliasson AC. The Assisting Hand Assessment: current combined use of the Canadian occupational performance evidence of validity, reliability, and responsiveness to measure (COPM) and the goal attainment scaling (GAS). Dev change. Dev Med Child Neurol 2007;49(4):259e64. Neurorehabil 2008;11(4):252e9. 203. Krumlinde-sundholm L, Eliasson A-C. Development of the 220. Martin A, Burtner PA, Poole J, Phillips J. Case report: ICF-level assisting hand assessment: a Rasch-built measure intended changes in a preschooler after constraint-induced for children with unilateral upper limb impairments. Scand J movement therapy. Am J Occup Ther 2008;62(3):282e8. Occup Ther 2003;10(1):16e26. 221. Grap MJ, Hamilton VA, McNallen A, Ketchum JM, Best AM, 204. Holmefur M, Krumlinde-sundholm L, Eliasson AC. Interrater Arief NYI, et al. Actigraphy: analyzing patient movement. and intrarater reliability of the assisting hand assessment. Heart Lung 2011;40(3):e52e9. Am J Occup Ther 2007;61(1):79e84. 222. Kirton A, deVeber G, Gunraj C,ChenR.Neurocardiogenic 205. Holmefur M. The Assisting Hand Assessment: continued syncope complicating pediatric transcranial magnetic development, psychometrics and longitudinal use. Stockholm: stimulation. Pediatr Neurol 2008;39:196e7(0887e8994 Karolinska Institutet; 2009. (Print)). 206. Holmefur M, Aarts P, Hoare B, Krumlinde-sundholm L. Test- 223. Gillick BT, Rich T, Chen M, Meekins GD. Case report of retest and alternate forms reliability of the assisting hand vasovagal syncope associated with single pulse transcranial assessment. J Rehabil Med 2009;41(11):886e91. magnetic stimulationin a healthy adult participant. BMC Neurol 207. Eliasson AC, Shaw K, Ponten E, Boyd R, Krumlinde- 2015;15:248. http://dx.doi.org/10.1186/s12883-015-0510-2. sundholm L. Feasibility of a day-camp model of modified 224. Garvey MA, Gilbert DL. Transcranial magnetic stimulation in constraint-induced movement therapy with and without children. Eur J Paediatr Neurol 2004;8(1090e3798):7e19. Botulinum toxin a injection for children with hemiplegia. 225. Sarfeld AS, Diekhoff S, Wang LE, Liuzzi G, Uludag K, Phys Occup Ther Pediatr 2009;29(3):311e33. Eickhoff SB, et al. Convergence of human brain mapping 208. Gilmore R, Boyd R, Sakzewski L. A systematic review of the tools: neuronavigated TMS parameters and fMRI activity in clinimetric properties of upper limb activity measures for 5- the hand motor area. Hum Brain Mapp 2012 to 16-year-old children with congenital hemiplegia. Dev Med May;33(5):1107e23. http://dx.doi.org/10.1002/hbm.21272. Child Neurol 2008;50(Suppl. 4) (Journal Article):66. Epub 2011 Apr 21. 209. Sakzewski L, Ziviani J, Boyd RN. Best responders after 226. Barker AT, Jalinous R, Freeston IL. Non-invasive magnetic intensive upper-limb training for children with unilateral stimulation of human motor cortex. Lancet 1985;1:1106e7 cerebral palsy. Arch Phys Med Rehabil 2011;92:578e84 (0140e6736 (Print)). (1532e821X (Electronic)). 227. Theodore WH. Handbook of transcranial magnetic 210. Wallen M, Ziviani J, Herbert R, Evans R, Novak I. Modified stimulation. Edited by A. Pascual-Leone, N.J. Davey, J. constraint-induced therapy for children with hemiplegic Rothwell, E.M. Wasseran, B.K. Puri, Arnold, London, 2001. cerebral palsy: a feasibility study. Dev Neurorehabil pound 110 sterling, ISBN 0340720093. Epilepsy Behav 2008;11(2):124e33. 2002;3(1525e5050 (Print)):404. 211. Law M, Baptiste S, McColl M, Opzoomer A, Polatajko H, 228. Jang SH, Byun WM, Chang Y, Han BS, Ahn SH. Combined Pollock N. The Canadian occupational performance functional magnetic resonance imaging and transcranial measure: an outcome measure for occupational therapy. magnetic stimulation evidence of ipsilateral motor pathway Can J Occup Ther Rev Can Ergother 1990;57(2):82e7. with congenital brain disorder: a case report. Arch Phys Med 212. Law M, Polatajko H, Pollock N, McColl MA, Carswell A, Rehabil 2001;82:1733e6 (0003e9993 (Print)). Baptiste S. Pilot testing of the Canadian occupational 229. Carr LJ. Development and reorganization of descending performance measure: clinical and measurement issues. motor pathways in children with hemiplegic cerebral palsy. Can J Occup Ther Rev Can Ergother 1994;61(4):191e7. Acta Paediatr Suppl 1996;416:53e7 (0803e5326 (Print)). european journal of paediatric neurology 21 (2017) 75e103 101

230. Duque J, Hummel F, Celnik P, Murase N, Mazzocchio R, connectivity approaches. Brain 2011;134:1264e76 (1460e2156 Cohen LG. Transcallosal inhibition in chronic subcortical (Electronic)). stroke. Neuroimage 2005;28:940e6 (1053e8119 (Print)). 249. Friston KJ, Frith CD, Fletcher P, Liddle PF, Frackowiak RS. 231. Brasil-Neto JP, Cammarota A, Valls-Sole J, Pascual-Leone A, Functional topography: multidimensional scaling and Hallett M, Cohen LG. Role of intracortical mechanisms in the functional connectivity in the brain. Cereb Cortex late part of the silent period to transcranial stimulation of 1996;6:156e64 (1047e3211 (Print)). the human motor cortex. Acta Neurol Scand 1995;92:383e6 250. Stephan KE, Friston KJ. Analyzing effective connectivity with (0001e6314 (Print)). fMRI. Wiley Interdiscip Rev Cogn Sci 2010;1:446e59. 232. Hallett M. Transcranial magnetic stimulation. Negative 251. Rehme AK, Eickhoff SB, Wang LE, Fink GR, Grefkes C. effects. Adv Neurol 1995;67:107e13 (0091e3952 (Print)). Dynamic causal modeling of cortical activity from the acute 233. Terao Y, Ugawa Y. Basic mechanisms of TMS. J Clin to the chronic stage after stroke. Neuroimage Neurophysiol 2002;19:322e43 (0736e0258 (Print)). 2011;55:1147e58 (1095e9572 (Electronic)). 234. Heinen F, Glocker FX, Fietzek U, Meyer BU, Lucking CH, 252. Carter AR, Astafiev SV, Lang CE, Connor LT, Rengachary J, Korinthenberg R. Absence of transcallosal inhibition Strube MJ, et al. Resting interhemispheric functional following focal magnetic stimulation in preschool children. magnetic resonance imaging connectivity predicts Ann Neurol 1998;43:608e12. performance after stroke. Ann Neurol 2010;67:365e75 235. Gilbert DL, Garvey MA, Bansal AS, Lipps T, Zhang J, (1531e8249 (Electronic)). Wassermann EM. Should transcranial magnetic stimulation 253. de Bie HM, Boersma M, Adriaanse S, Veltman DJ, Wink AM, research in children be considered minimal risk? Clin Roosendaal SD, et al. Resting-state networks in awake five- Neurophysiol 2004;115(1388e2457):1730e9. to eight-year old children. Hum Brain Mapp 2012;33:1189e201 236. Valero-Cabre A, Rushmore RJ, Payne BR. Low frequency (1097e0193 (Electronic)). transcranial magnetic stimulation on the posterior parietal 254. Polania R, Nitsche MA, Paulus W. Modulating functional cortex induces visuotopically specific neglect-like syndrome. connectivity patterns and topological functional Exp Brain Res 2006:1e8 (0014e4819 (Print)). organization of the human brain with transcranial direct 237. Valero-Cabre A, Payne BR, Rushmore J, Lomber SG, Pascual- current stimulation. Hum Brain Mapp 2011;32:1236e49 Leone A. Impact of repetitive transcranial magnetic (1097e0193 (Electronic)). stimulation of the parietal cortex on metabolic brain 257. Rollins NK. Clinical applications of diffusion tensor imaging activity: a 14C-2DG tracing study in the cat. Exp Brain Res and tractography in children. Pediatr Radiol 2007;37:769e80 2005;163:1e12 (0014e4819 (Print)). (0301e0449 (Print)). 238. Bolay H, Gursoy-Ozdemir Y, Unal I, Dalkara T. Altered 258. Moller M, Frandsen J, Andersen G, Gjedde A, Vestergaard- mechanisms of motor-evoked potential generation after Poulsen P, Ostergaard L. Dynamic changes in corticospinal transient focal cerebral ischemia in the rat: implications tracts after stroke detected by fibretracking. J Neurol for transcranial magnetic stimulation. Brain Res Neurosurg Psychiatry 2007;78:587e92 (1468e330X 2000;873:26e33. (Electronic)). 239. Eyre JA, Flecknell PA, Kenyon BR, Koh TH, Miller S. Acute 259. Radlinska B, Ghinani S, Leppert IR, Minuk J, Pike GB, Thiel A. effects of electromagnetic stimulation of the brain on Diffusion tensor imaging, permanent pyramidal tract cortical activity, cortical blood flow, blood pressure and damage, and outcome in subcortical stroke. Neurology heart rate in the cat: an evaluation of safety. J Neurol 2010;75:1048e54 (1526e632X (Electronic)). Neurosurg Psychiatry 1990;53:507e13. 260. Heller SL, Heier LA, Watts R, Schwartz TH, Zelenko N, 240. Edgley SA, Eyre JA, Lemon RN, Miller S. Excitation of the Doyle W, et al. Evidence of cerebral reorganization following corticospinal tract by electromagnetic and electrical perinatal stroke demonstrated with fMRI and DTI stimulation of the scalp in the macaque monkey. J Physiol tractography. Clin Imaging 2005;29(4):283e7. http:// 1990;425:301e20 (0022e3751 (Print)). dx.doi.org/10.1016/j.clinimag.2004.09.003. 241. Theodore WH, Hunter K, Chen R, Vega-Bermudez F, 261. Yoshida S, Hayakawa K, Yamamoto A, Okano S, Kanda T, Boroojerdi B, Reeves-Tyer P, et al. Transcranial magnetic Yamori Y, et al. Quantitative diffusion tensor tractography stimulation for the treatment of seizures: a controlled study. of the motor and sensory tract in children with cerebral Neurology 2002;59:560e2 (0028e3878 (Print)). palsy. Dev Med Child Neurol 2010;52:935e40 (1469e8749 242. Tassinari CA, Cincotta M, Zaccara G, Michelucci R. (Electronic)). Transcranial magnetic stimulation and epilepsy. Clin 262. Glenn OA, Ludeman NA, Berman JI, Wu YW, Lu Y, Bartha AI, Neurophysiol 2003;114:777e98 (1388e2457 (Print)). et al. Diffusion tensor MR imaging tractography of the 243. Magnetic stimulation in clinical neurophysiology. vol. 2. pyramidal tracts correlates with clinical motor function in Philadelphia, PA: Elsevier; 2005 children with congenital hemiparesis. AJNR Am J Neuroradiol 244. Daskalakis ZJ, Christensen BK, Fitzgerald PB, Roshan L, 2007;28(9):1796e802. http://dx.doi.org/10.3174/ajnr.A0676. Chen R. The mechanisms of interhemispheric inhibition in 263. Hodge J, Carlson H, Goodyear B, wei x. Kirton A DTI the human motor cortex. J Physiol 2002;543:317e26 quantification of corticospinal tract integrity in perinatal (0022e3751 (Print)). stroke. Proc Neuro DevNet Brain Dev Conf 2011;30. 245. Di LV, Oliviero A, Profice P, Insola A, Mazzone P, Tonali P, et al. 264. Kuczynski AM, Dukelow SP, Semrau JA, Kirton A. Robotic Direct demonstration of interhemispheric inhibition of the quantification of position sense in children with perinatal human motor cortex produced by transcranial magnetic stroke. Neurorehabil Neural Repair 2016 Jan 7. stimulation. Exp Brain Res 1999;124:520e4 (0014e4819 (Print)). pii: 1545968315624781. [Epub ahead of print]. 246. Ferbert A, Priori A, Rothwell JC, Day BL, Colebatch JG, 265. Juenger H, Kuhnke N, Braun C, Ummenhofer F, Wilke M, Marsden CD. Interhemispheric inhibition of the human Walther M, et al. Two types of exercise-induced motor cortex. J Physiol 1992;453:525e46. neuroplasticity in congenital hemiparesis: a transcranial 247. Roshan L, Paradiso GO, Chen R. Two phases of short-interval magnetic stimulation, functional MRI, and intracortical inhibition. Exp Brain Res 2003;151:330e7 magnetoencephalography study. Dev Med Child Neurol (0014e4819 (Print)). 2013;55:941e51 (1469e8749 (Electronic)). 248. Grefkes C, Fink GR. Reorganization of cerebral networks 266. Jaspers E, Byblow WD, Feys H, Wenderoth N. The after stroke: new insights from neuroimaging with corticospinal tract: a Biomarker to categorize upper limb 102 european journal of paediatric neurology 21 (2017) 75e103

functional potential in unilateral cerebral palsy. Front Pediatr 280. Zimerman M, Heise KF, Hoppe J, Cohen LG, Gerloff C, 2015;3:112. http://dx.doi.org/10.3389/fped.2015.00112. Hummel FC. Modulation of training by single-session 267. Hummel FC, Cohen LG. Non-invasive brain stimulation: a transcranial direct current stimulation to the intact motor new strategy to improve neurorehabilitation after stroke? cortex enhances motor skill acquisition of the paretic hand. Lancet Neurol 2006;5:708e12 (1474e4422 (Print)). Stroke 2012;43(8):2185e91. http://dx.doi.org/10.1161/ 268. Hesse S, Werner C, Schonhardt EM, Bardeleben A, Jenrich W, STROKEAHA.111.645382. Kirker SG. Combined transcranial direct current stimulation 281. Wu D, Qian L, Zorowitz RD, Zhang L, Qu Y, Yuan Y. Effects on and robot-assisted arm training in subacute stroke patients: decreasing upper-limb poststroke muscle tone using a pilot study. Restor Neurol Neurosci 2007;25:9e15 (0922e6028 transcranial direct current stimulation: a randomized sham- (Print)). controlled study. Arch Phys Med Rehabil 2013;94(1):1e8. http:// 269. Boggio PS, Rigonatti SP, Ribeiro RB, Myczkowski ML, dx.doi.org/10.1016/j.apmr.2012.07.022. Nitsche MA, Pascual-Leone A, et al. A randomized, 282. Ochi M, Saeki S, Oda T, Matsushima Y, Hachisuka K. Effects double-blind clinical trial on the efficacy of cortical direct of anodal and cathodal transcranial direct current current stimulation for the treatment of major stimulation combined with robotic therapy on severely depression. Int J Neuropsychopharmacol 2007:1e6 affected arms in chronic stroke patients. J Rehabil Med (1461e1457 (Print)). 2013;45(2):137e40. http://dx.doi.org/10.2340/16501977-1099. 270. Celnik P, Paik N-J, Vandermeeren Y, Dimyan M, Cohen LG. 283. Lefebvre S, Laloux P, Peeters A, Desfontaines P, Jamart J, Effects of combined peripheral nerve stimulation and brain Vandermeeren Y. Dual-tDCS enhances online motor skill polarization on performance of a motor sequence task after learning and long-term retention in chronic stroke patients. chronic stroke. Stroke 2009;40(5):1764e71. http://dx.doi.org/ Front Hum Neurosci 2013;6. http://dx.doi.org/10.3389/ 10.1161/STROKEAHA.108.540500. fnhum.2012.00343. 271. Kim DY, Ohn SH, Yang EJ, Park C-I, Jung KJ. Enhancing motor 284. Fusco A, De Angelis D, Morone G, Maglione L, Paolucci T, performance by anodal transcranial direct current Bragoni M, et al. The ABC of tDCS: effects of anodal, bilateral stimulation in subacute stroke patients. Am J Phys Med and cathodal montages of transcranial direct current Rehabil Assoc Acad Physiatr 2009;88(10):829e36. http:// stimulation in patients with stroke-a pilot study. Stroke Res dx.doi.org/10.1097/PHM.0b013e3181b811e3. Treat 2013;2013:837595. http://dx.doi.org/10.1155/2013/ 272. Kim D-Y, Lim J-Y, Kang EK, You DS, Oh M-K, Oh B-M, et al. 837595. Effect of transcranial direct current stimulation on motor 285. Lefebvre S, Thonnard J-L, Laloux P, Peeters A, Jamart J, recovery in patients with subacute stroke. Am J Phys Med Vandermeeren Y. Single session of dual-tDCS transiently Rehabil Assoc Acad Physiatr 2010;89(11):879e86. http:// improves precision grip and dexterity of the paretic hand dx.doi.org/10.1097/PHM.0b013e3181f70aa7. after stroke. Neurorehabil Neural Repair 2014;28(2):100e10. 273. Nair DG, Renga V, Lindenberg R, Zhu L, Schlaug G. http://dx.doi.org/10.1177/1545968313478485. Optimizing recovery potential through simultaneous 286. Khedr EM, Shawky OA, El-Hammady DH, Rothwell JC, occupational therapy and non-invasive brain-stimulation Darwish ES, Mostafa OM, et al. Effect of anodal versus using tDCS. Restor Neurol Neurosci 2011;29(6):411e20. http:// cathodal transcranial direct current stimulation on stroke dx.doi.org/10.3233/RNN-2011-0612. rehabilitation: a pilot randomized controlled trial. 274. Hesse S, Waldner A, Mehrholz J, Tomelleri C, Pohl M, Neurorehabil Neural Repair 2013;27(7):592e601. http:// Werner C. Combined transcranial direct current stimulation dx.doi.org/10.1177/1545968313484808. and robot-assisted arm training in subacute stroke patients: 287. O'Shea J, Boudrias M-H, Stagg CJ, Bachtiar V, Kischka U, an exploratory, randomized multicenter trial. Neurorehabil Blicher JU, et al. Predicting behavioural response to TDCS in Neural Repair 2011;25(9):838e46. http://dx.doi.org/10.1177/ chronic motor stroke. Neuroimage 2014;85:924e33. http:// 1545968311413906. dx.doi.org/10.1016/j.neuroimage.2013.05.096. 275. Bolognini N, Vallar G, Casati C, Latif LA, El-Nazer R, 288. Giacobbe V, Krebs HI, Volpe BT, Pascual-Leone A, Rykman A, Williams J, et al. Neurophysiological and behavioral effects Zeiarati G, et al. Transcranial direct current stimulation of tDCS combined with constraint-induced movement (tDCS) and robotic practice in chronic stroke: the dimension therapy in poststroke patients. Neurorehabil Neural Repair of timing. Neuro Rehabil 2013;33(1):49e56. http://dx.doi.org/ 2011;25(9):819e29. http://dx.doi.org/10.1177/ 10.3233/NRE-130927. 1545968311411056. 289. Fusco A, Iosa M, Venturiero V, De Angelis D, Morone G, 276. Stagg CJ, Bachtiar V, O'Shea J, Allman C, Bosnell RA, Maglione L, et al. After vs. priming effects of anodal Kischka U, et al. Cortical activation changes underlying transcranial direct current stimulation on upper extremity stimulation-induced behavioural gains in chronic stroke. motor recovery in patients with subacute stroke. Restor Brain J Neurol 2012;135(Pt 1):276e84. http://dx.doi.org/ Neurol Neurosci 2014;32(2):301e12. 10.1093/brain/awr313. 290. Fusco A, Assenza F, Iosa M, Izzo S, Altavilla R, Paolucci S, 277. Lindenberg R, Zhu LL, Ruber T, Schlaug G. Predicting et al. The ineffective role of cathodal tDCS in enhancing the functional motor potential in chronic stroke patients using functional motor outcomes in early phase of stroke diffusion tensor imaging. Hum Brain Mapp 2012;33:1040e51 rehabilitation: an experimental trial. Biomed Res Int (1097e0193 (Electronic)). 2014;2014:1e9. http://dx.doi.org/10.1155/2014/547290. 278. Takeuchi N, Tada T, Matsuo Y, Ikoma K. Low-frequency 291. Cha H-K, Ji S-G, Kim M-K, Chang J-S. Effect of transcranial repetitive TMS plus anodal transcranial DCS prevents direct current stimulation of function in patients with transient decline in bimanual movement induced by stroke. J Phys Ther Sci 2014;26(3):363e5. http://dx.doi.org/ contralesional inhibitory rTMS after stroke. Neurorehabil 10.1589/jpts.26.363. Neural Repair 2012;26(8):988e98. http://dx.doi.org/10.1177/ 292. Au-Yeung SSY, Wang J, Chen Y, Chua E. Transcranial direct 1545968311433295. current stimulation to primary motor area improves hand 279. Rossi C, Sallustio F, Di Legge S, Stanzione P, Koch G. dexterity and selective attention in chronic stroke. Am J Phys Transcranial direct current stimulation of the affected Med Rehabil 2014;93(12):1057e64. http://dx.doi.org/10.1097/ hemisphere does not accelerate recovery of acute stroke PHM.0000000000000127. patients. Eur J Neurol 2013;20(1):202e4. http://dx.doi.org/ 293. Gillick BT, Feyma T, Menk J, Krach LE, Usset M, Vaith A, et al. 10.1111/j.1468-1331.2012.03703.x. Safety and feasibility of transcranial direct current european journal of paediatric neurology 21 (2017) 75e103 103

stimulation in pediatric hemiparesis: a randomized type of stroke. Biomed Res Int 2016;2016:5068127. http:// controlled pilot study. Phys Ther 2015 Mar;95(3):337e49. dx.doi.org/10.1155/2016/5068127. http://dx.doi.org/10.2522/ptj.20130565. Epub 2014 Nov 20. 300. Allman C, Amadi U, Winkler AM, Wilkins L, Filippini N, 294. Goh H-T, Chan H-Y, Abdul-Latif L. Aftereffects of 2 Kischka U, et al. Ipsilesional anodal tDCS enhances the noninvasive brain stimulation techniques on corticospinal functional benefits of rehabilitation in patients after stroke. excitability in persons with chronic stroke: a pilot study. J Sci Transl Med 2016;8(330):330re1. http://dx.doi.org/10.1126/ Neurol Phys Ther JNPT 2015;39(1):15e22. http://dx.doi.org/ scitranslmed.aad5651. 10.1097/NPT.0000000000000064. 301. Triccas LT, Burridge JH, Hughes A, Verheyden G, Desikan M, 295. Di Lazzaro V, Rothwell JC. Corticospinal activity evoked and Rothwell J. A double-blinded randomised controlled trial modulated by non-invasive stimulation of the intact human exploring the effect of anodal transcranial direct current motor cortex: corticospinal activity and the human motor stimulation and uni-lateral robot therapy for the impaired cortex. J Physiol 2014;592(19):4115e28. http://dx.doi.org/ upper limb in sub-acute and chronic stroke. Neuro Rehabil 10.1113/jphysiol.2014.274316. 2015;37(2):181e91. http://dx.doi.org/10.3233/NRE-151251. 296. Lefebvre S, Dricot L, Laloux P, Gradkowski W, 302. Cunningham DA, Varnerin N, Machado A, Bonnett C, Janini D, Desfontaines P, Evrard F, et al. Neural substrates underlying Roelle S, et al. Stimulation targeting higher motor areas in stimulation-enhanced motor skill learning after stroke. Brain stroke rehabilitation: a proof-of-concept, randomized, J Neurol 2015;138(Pt 1):149e63. http://dx.doi.org/10.1093/ double-blinded placebo-controlled study of effectiveness and brain/awu336. underlying mechanisms. Restor Neurol Neurosci 297. Sattler V, Acket B, Raposo N, Albucher J-F, Thalamas C, 2015;33(6):911e26. http://dx.doi.org/10.3233/RNN-150574. Loubinoux I, et al. Anodal tDCS combined with radial nerve 303. Lee D-G, Lee D-Y. Effects of adjustment of transcranial direct stimulation promotes hand motor recovery in the acute current stimulation on motor function of the upper phase after ischemic stroke. Neurorehabil Neural Repair extremity in stroke patients. J Phys Ther Sci 2015;29(8):743e54. http://dx.doi.org/10.1177/ 2015;27(11):3511e3. http://dx.doi.org/10.1589/jpts.27.3511. 1545968314565465. 304. Chen JL, Schlaug G. Increased resting state connectivity 298. Rocha S, Silva E, Foerster A, Wiesiolek C, Chagas AP, between ipsilesional motor cortex and contralesional Machado G, et al. The impact of transcranial direct current premotor cortex after transcranial direct current stimulation stimulation (tDCS) combined with modified constraint- with physical therapy. Sci Rep 2016;6:23271. http://dx.doi.org/ induced movement therapy (mCIMT) on upper limb function 10.1038/srep23271. in chronic stroke: a double-blind randomized controlled 305. Mortensen J, Figlewski K, Andersen H. Combined trial. Disabil Rehabil 2016;38(7):653e60. http://dx.doi.org/ transcranial direct current stimulation and home-based 10.3109/09638288.2015.1055382. occupational therapy for upper limb motor impairment 299. Straudi S, Fregni F, Martinuzzi C, Pavarelli C, Salvioli S, following intracerebral hemorrhage: a double-blind Basaglia N. tDCS and robotics on upper limb stroke randomized controlled trial. Disabil Rehabil rehabilitation: effect modification by stroke duration and 2016;38(7):637e43. http://dx.doi.org/10.3109/ 09638288.2015.1055379.