Perinatal stroke causes lifelong neurological disability and most hemiparetic cerebral palsy (CP). With morbidity spanning diverse aspects of a child's life and lasting for decades, global impact is large, including 10000 Canadian children. With pathophysiology poorly understood and prevention strategies non-existent, the burden of hemiparetic CP will persist. Limited treatments lead to loss of hope for children and families, necessitating exploration of new therapies. The investigators have evidence that the investigators have a durable new treatment for perinatal stroke, combining non-invasive neurostimulation and child-centred intensive rehabilitation. Via the CHILD-BRIGHT SPOR national network, the investigators will execute a multicentre trial to prove this treatment can improve function in children with perinatal stroke and hemiparetic CP. Using novel advanced technologies not available elsewhere in the world, the investigators will explore how developmental plasticity determines function and response to neuromodulation therapy. This patient oriented effort will advance personalized, precision medicine in pediatric neurorehabilitation to improve outcomes for disabled children and their families.

Aim 1: Establish the ability of tDCS to enhance motor function in children with perinatal stroke.

Hypothesis: Addition of tDCS to intensive motor learning therapy in children with perinatal stroke and hemiparesis will increase the probability of achieving clinically significant gains in function.

Aim 2: Understand the developmental motor neurophysiology of perinatal stroke and the changes that occur during intensive motor learning and the effects of tDCS.

Hypothesis: Therapy-induced functional improvements are associated with increased motor control in the lesioned (stroke) hemisphere as measured by: (a) enlargement of lesioned hemisphere motor maps (robotic TMS), (b) increased functional connectivity between lesioned motor and sensory cortices (rsfMRI), (c) strengthened correlations between lesioned motor cortex NAA and clinical function (MRS), and (d) improved position sense of the affected upper extremity (KINARM robot)

Methods This is a multicentre, randomized, sham-controlled, double blind, phase III clinical trial.

Population. Participants will be recruited through established programs by expert investigative teams at 4 sites. For Calgary (Alberta Children's Hospital) and Edmonton (Stollery Children's Hospital and Glenrose Rehabilitation Hospital), subjects will be identified via the Alberta Perinatal Stroke Project (APSP); a population-based research cohort of >1000 MRI-confirmed perinatal stroke subjects with a proven track record of recruitment to clinical trials71. Toronto subjects will be recruited from established CP research cohorts at Holland Bloorview Kids Rehabilitation Hospital. All sites have established clinical research programs in childhood CP and experience running relevant interventions including camp-based intensive motor learning programs.

Intensive Motor Learning Camp. Participants will complete child-centered, age-appropriate, goal-directed 2-week intensive motor learning camps (summers of 2017, 2018, and 2019). These will be full day programs in peer-supported environments. All sites are experienced in such motor learning interventions. Fidelity of all interventions will be facilitated by standardized operating procedures, videotaping, and scheduled site visits. CIMT during the first week will be followed by bimanual therapy during the second. Tasks will be graded and selected according to relative function with increasing complexity across the spectrum. Tasks will be both symmetrical and asymmetrical, geared to pre-set goals and age-appropriate activities of daily living and individual interests. Intensive motor learning will be focused within 2 hours each day of goal-directed therapy working 1:1 with a dedicated OT. The remaining hours will include gross motor (1.5), group motor (1.5), lunch (1), social snack (0.5x2), and fun breaks (0.5). Total dose is therefore 7.5 hours per day (75 hours total). These activities include upper limb motor activities focused on general activities of daily living rather than individualized goals. Following the 10 day intervention, children will receive a structured home program based on the same principles with ongoing therapist support if needed in the form of phone calls, surveillance, and documentation (home program log) for 6 months.

Randomization and Blinding. Randomization will occur in permuted blocks of 2 to ensure even distribution within sites. Concealment of randomization will be achieved using online methods. After turning on tDCS, those randomized to sham will have their machine automatically ramp down after 30 seconds while those randomized to active treatment will remain at 1.0mA. All subjects will experience the same sensations and such shaming has been proven effective, including in children131. Treating therapists, outcome assessors, parents and children are blinded to treatment allocation. Following day 1 and day 10 tDCS session, participants will be asked first to guess which treatment they received: real, sham or no idea and why. If they choose no idea, they will then be asked to choose only between real and sham and give their reasons why.

Intervention: tDCS. The primary intervention will be cathodal (inhibitory) tDCS over the contralesional M1. Rationale includes evidence from a phase 1 trial (Kirton 2016, submitted), phase 2 rTMS trial, pediatric tDCS motor learning trial as well as adult stroke rTMS and tDCS trials and developmental animal studies. Inhibitory stimulation over intact, contralesional brain also maximizes safety with more predictable distribution of tDCS currents. Based on previous studies of cathodal tDCS in motor learning, adult stroke rehabilitation and studies of electrode montages in stroke and children140, the following methods will be employed. Soft, replaceable 25cm2 electrodes (Soterix, NYC) will be placed on clean, dry areas of the scalp. The cathode will be placed over the contralesional M1, precisely mapped for each patient using neuronavigated (Brainsight2, Rogue Research, Montreal QU) MRI-TMS co-registration over the hotspot for the contralateral first dorsal interosseous muscle. Consistent with previous tDCS stroke studies, the reference electrode will be placed over the contralateral orbit. The current-controlled model stimulator (Soterix, NYC) will automatically ramp up slowly over 30 seconds to the treatment current of 1.0 mA. tDCS or sham will be administered each day during the first 30 minutes of the daily 1:1 therapy sessions. Co-investigators experienced in pediatric tDCS will perform on-site training and ongoing quality assurance according to standard operating procedures.

Clinical Outcomes: Motor Function. Expert co-investigators have designed a rigorous approach to motor outcome assessment, overcoming some of the limitations of existing hemiparetic CP trials. Multidimensional evaluations will assess body structure and function, activity and participation domains of the WHO International Classification of Functioning, Disability and Health (ICF). Tools with established clinimetric properties were selected to evaluate diverse upper extremity functions relevant to daily living in children. Measures had to be both child/family friendly and time sensitive. The investigators will measure uni- and bi-lateral function for normative data, evaluation of bimanual functions, and safety (including screening for changes in unaffected hand function). Tests will be performed by the same experienced, blinded, non-treating site OT and video-taped for quality assurance and offline analysis. Motor outcomes are measured at baseline, and 1 week, 2 months and 6 months post-intervention.

A. Primary Objective Motor Outcome: Assisting Hand Assessment (AHA). This is the established standard for the objective quantification of bilateral hand function in children with hemiparetic CP. 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 this age range. Sensitivity to change and excellent clinimetric properties have been established in multiple pediatric hemiparetic CP clinical trials. Trained therapists have successfully executed >100 AHA measurements in previous trials with no limitations and robust data.

B. Primary Subjective Motor Outcome: Canadian Occupational Performance Measure (COPM). Individualized, family-centered tool identifying child and family-perceived difficulties in self-care, productivity (school), and activities. Such subjective measures are essential in hemiparetic CP trials. Validated for these ages and such trials, the COPM was a robust measure in previous perinatal stroke trials. The investigators have recently characterized how COPM goals are set in this population and their relationship to success (Haspels et al, unpublished).

C. Novel real-life motor outcome: Actigraphy. No existing motor outcome measure can quantify continuous use of the upper extremities during normal activities. The investigators propose to overcome this limitation using actigraphy. Lightweight wrist accelerometers can constantly measure and store subtle movements. Such systems can track movements in disabled persons including those with CP145. Participants will be fitted bilaterally with actiwatches (MotionWatch) to record mean movements every 2 seconds for 48 hour epochs of time (baseline, 1 week, 2 months, 6 months) as well as continuously during the 2 week intervention phase. A standard diary will record sleep/wake patterns and activities. Importantly, The investigators will do this bilaterally (2 watches) to generate the primary outcome of an actigraphic asymmetry index (AAI) calculated between the affected and unaffected limbs.

Secondary Clinical Outcomes summarized include: Mirror movements, Children's Hand-Use Experience Questionnaire (CHEQ), Jebsen Taylor Test of Hand Function, Box and Blocks Test (BBT), Stanford Expectations of Treatment Scale (SETS), Child and Adolescent Social Support Scale (CASSS), Loneliness and Social Dissatisfaction Questionnaire (LSDQ), PedsQL-CP, CNS Vital signs, Child and Adolescent Scale of Participation (CASP), APSP parental outcome measure (POM), and Health Utilities Index (HUI).

Safety and Tolerability. An experienced data safety and monitoring board will be established. Primary adverse outcome for tDCS is any decrease in function of either upper extremity, screened for by both uni- and bi-manual measures. An established tDCS Safety and Tolerability evaluation (TST) will be performed on days 1, 5, and 10 on all subjects. Only trained personnel will administer treatments under the approval and certification from experienced brain stimulation team members. All tDCS interventions will occur within a hospital setting with immediate access to medical care in the unlikely event care is required. Any potential adverse events are immediately reported to the site and principal investigators, DSMB safety monitor, and ethics boards. For any potentially serious adverse events, treatments will be suspended pending decisions of the above.

Neuroplasticity Outcomes. For each component below, change in the primary outcome is anticipated to favour increased motor control in the lesioned hemisphere associated with functional improvements. All sites will capture neuroimaging data according to matched protocols. Only Alberta sites will complete the TMS and KINARM robotic measures.

A. MRI: Connectivity and chemistry. Images will be acquired on dedicated 3T research scanners across centres using standardized protocols within established neuroimaging networks. Baseline imaging will be performed within 4 weeks of intervention. Post-intervention imaging will be performed 5-7 days after the last day of treatment and at 6 months following treatment. Scan time will be ~60 minutes. Anatomical imaging following established perinatal stroke protocols includes T1- and T2-weighted 3D BRAVO anatomical sequences. For task fMRI, subjects will perform contraction of affected hand and unaffected hand in an event-related design. In response to a cue, participants will squeeze an air-filled bulb attached to a pressure transducer that quantifies grip pressure. A static fixation cross will indicate to stop and rest for 14 seconds. Brain activity in response to each hand movement will be determined using SPM12 by employing a General Linear Model. Registration to each participant's anatomical images will be performed using reverse deformation fields from SPM segmentation. Outcomes include extent of blood-oxygen level dependent response (BOLD) activation, peak locations and M1 laterality index (LI). For resting state MRI acquisition, children will be instructed to lie still with eyes open. Resting-state fMRI BOLD signals will be extracted from each M1 and other areas of interest as defined above with temporal cross-correlation. Correlation coefficients will be Fisher transformed and entered into a random effects General Linear Model consistent with validated methods. Outcomes include intrahemispheric (M1-S1) and interhemispheric (M1-M1) connectivity. For white matter connectivity, diffusion tensor imaging will include 60 directions (b-values= 0,2000s/mm2). Voxel size is 2.5mm isotropic. Fractional anisotropy (FA), color coded-FA, ADC, and b0 maps will all be generated in MRTrix3. A multi-ROI approach will generate tracts of interest using probabilistic tractography. Outcomes will expand previous experience interrogating corticospinal, somatosensory, and transcallosal tracts. Metabolite concentrations in bilateral M1 will be measured via magnetic resonance spectroscopy (MRS), providing information about neuronal and cell membrane health, energy metabolism, health of glial cells and excitatory neurotransmitters. Preliminary data from the APSP has demonstrated the feasibility of these methods in children with perinatal stroke. Established neuroimaging experts will ensure optimal comparability between centres according to predefined protocols. The centralized imaging repository system of the Stroke Imaging Lab for Children (SILC) supported by Brain Canada will use established protocols and systems (BrainCode) for anonymized online transfer of images to and from site investigators.

B. TMS: Robotic motor mapping. TMS studies will occur at the ACH Pediatric Brain Stimulation Laboratory, a state-of-the-art, child-friendly facility that has completed >3 million stimulations in >280 children (Zewdie, unpublished) without complication. The investigators are the first pediatric centre in the world with the Axilum TMS Robot72 and have developed a rapid motor mapping technique (see Figure 2). Using each subject's anatomical MRI and neuronavigation (Brainsight2), a 10x10 grid with 7mm spacing will be placed over M1. A recently developed rapid mapping protocol (4 stimulations per site) will obtain mean MEP values for each active site. The robot allows precise targeting with real-time motion correction and rapid mapping. The primary TMS outcome will be area-under-the-curve of customized heat maps of contralesional motor cortex. Multiple upper extremity muscles will be simultaneously mapped from both hemispheres including dedicated ipsilateral projections from the contralesional hemisphere. Primary outcomes are motor map area, volume, and center of gravity. Additional TMS measures will include corticospinal tract arrangement, motor thresholds, MEP latency, and cortical silent period. The investigators have successfully demonstrated these methods in children with stroke.

C. KINARM Robot: Sensorimotor function. The KINARM exoskeleton (Kinesiological Instrument for Normal and Altered Reaching Movement) modified for children will assess limb movements at the shoulder and elbow. A standardized, validated assessment of sensorimotor function will include three tasks: position sense via a position matching task, kinesthesia, and visually guided reaching. The investigators have demonstrated the ability of these assessments to demonstrate specific elements of disordered sensorimotor function (50 perinatal stroke, 150 healthy children) including outcomes of position sense, kinesthesia, and motor dysfunction of the "unaffected" limb. Primary outcomes will be position sense during a position-matching task and bilateral motor impairment using a visually guided reaching task. The investigators have just installed a dedicated pediatric KINARM at the Alberta Children's Hospital where all Alberta subjects will attend.

Inclusion Criteria:

1. Clinical and MRI confirmed perinatal ischemic stroke (NAIS, APPIS, PVI)
2. Symptomatic hemiparetic CP including parent/child perceived limitations in function
3. Able to briefly lift light object off a surface (estimated House class 3-6).
4. Informed child consent/assent and parental/guardian consent

Exclusion Criteria:

1. Other neurological disorder not related to perinatal stroke
2. Multifocal stroke
3. Severe hemiparesis (no voluntary contraction, MACS V)
4. Sever spasticity (Modified Ashworth Scale >3)
5. Severe delay or inability to comply with protocol
6. Unstable epilepsy
7. TMS or MRI contraindication
8. Orthopedic surgery, constraint, brain stimulation or other modulatory therapy in past 6 months prior to camp day 1
9. Botulinum toxin injections in past 4 months prior to camp day 1