Predicting developmental plasticity after perinatal stroke

Abstract

This commentary is on the original article by van der Aa et al on pages 707–712 of this issue. Understanding how the developing brain responds to early injury is both scientifically fascinating and clinically relevant. As a focal injury of defined timing in an otherwise healthy brain, perinatal stroke probably represents the ideal human model of developmental neuroplasticity. Recent advances in neuroimaging and non-invasive brain stimulation have facilitated exploration of neurophysiological processes occurring in affected children. In a significant addition to this progress, van der Aa et al. describe early imaging predictors of the ultimate balance between ipsi- and contra-lesional motor control, so essential to function in children with hemiparesis.1 Animal and human evidence has recently been combined to generate working models of functional motor organization following unilateral perinatal brain injury.2 Essentially, early unilateral motor system injury impairs hemispheric ability to innervate contralateral spinal lower motor neuron pools normally. The consequence is the relative, pathological success of congenitally present ipsilateral corticospinal projections to ‘take-over’ these connections, the degree of which correlates with worse motor function. While maximal in the first 2 years, this process probably evolves in a dynamic fashion throughout development. Such understanding has generated potential central therapeutic targets in perinatal stroke-induced cerebral palsy, where either enhancing contralateral or inhibiting ipsilateral corticomotor systems might facilitate motor learning and improved function in affected children.3 The ability of constraint-induced movement therapy to enhance motor function in children with hemiparesis may well depend on such relative shift of motor function toward the lesioned hemisphere. In different but comparable adult stroke models, therapeutic non-invasive brain stimulation including repetitive transcranial magnetic stimulation (rTMS) appears to best enhance motor function by inhibiting the non-lesioned motor cortex.4 Such strategies are now being considered in children and could be directly informed by the early predictive imaging markers outlined here. The authors found no clear effect of injury timing on subsequent motor organization patterns. Timing was accurate with participants drawn from neonatal intensive care unit populations and arterial lesions spanning both preterm and term children (venous being exclusively preterm). Such ascertainment precluded the large proportion of both perinatal stroke types presenting later in infancy where the ability of similar imaging biomarkers to predict organization would be interesting. Recent studies by this group and others suggest diffusion tensor imaging may carry such utility,5 though the simple visual approach described in the current study is appealing for practicality and ready clinical availability. Consistent with previous studies, early clinical neuroimaging was predictive of eventual hemiparesis and remains practical for surveillance and early counselling of parents. The value added is the correlation to long-term neurophysiology, rather than simple clinical outcome. Based on the models mentioned above, it is increasingly apparent that such neurophysiology, rather than just clinical hemiparesis, will dictate approaches and success (or failure) of specific rehabilitation strategies. Of course, not all hemiparetic cerebral palsy is created equally and there are likely to be many additional variables influencing developmental motor plasticity and recovery, particularly genetics and neurological comorbidities, to be considered. How will such advances be translated into new therapeutic interventions and better outcomes? Such early prediction and elucidation of perinatal stroke developmental plasticity models are directly complementary to ongoing efforts to do just that. Two clinical trials of non-invasive brain stimulation coupled with motor learning therapy in perinatal stroke are currently underway (clinicaltrials.gov/NCT01104064 and /NCT01189058). Both are targeting the non-lesioned primary motor cortex with inhibitory rTMS. Their main difference centers on the precise issue identified here with one trial excluding children without TMS evidence of contralateral pathways (i.e. strong ipsilateral) while the other does not. While it is reasonable to avoid inhibiting what may be prominent ipsilateral cortical control centres (particularly with limited pediatric brain stimulation trials to date), a recent interim safety analysis from the second trial suggests the approach is not only safe in children with strong ipsilateral control but may even be beneficial. As the results of these trials are awaited and new approaches are developed, attention to the early imaging biomarkers of the developmental plasticity that underlies them identified by van der Aa et al. promise to be invaluable.