Constraint-induced movement therapy (CIMT) has become a well-established treatment for patients with hemiplegia, not only after stroke, but also in unilateral spastic cerebral palsy (USCP). In fact, most centers treating children and adolescents with this disorder offer CIMT. Numerous studies showed a good effect of CIMT in various cohorts of patients with USCP, but with considerable variability. Therefore, one would wish to be able to predict the response to CIMT in individual patients, especially since other treatment modalities like intensive bimanual training programs are available and have been demonstrated to be similarly effective.
1Five years ago, we published our observation that, on a behavioral level, patients with ipsilateral cortico-spinal projections responded differently to CIMT when compared with patients with preserved contralateral projections 2 : while both groups improved in the quality of movements with the paretic hand, only the contralateral group became faster, while the ipsilateral group became slower. Recently, we reported that this difference was not only present on a behavioral level, but that these two groups of patients also showed diverging changes of cortical excitability and taskrelated activation of the motor cortex controlling the paretic hand: 3 in patients with contralateral projections, transcranial magnetic stimulation of the motor cortex controlling the paretic hand elicited motor evoked potentials with a higher amplitude than before CIMT, while in patients with ipsilateral projections, the motor evoked potential amplitudes decreased. Likewise, after CIMT, functional magnetic resonance imaging (MRI) during active movements of the paretic hand demonstrated a stronger activation in the (contralateral) sensorimotor cortex of patients with contralateral projections, but a weaker activation in the (ipsilateral) motor cortex of patients with ipsilateral projections. Therefore, we are now more confident of our finding that these two groups of patients indeed respond differently to CIMT, not only on a behavioral level, but also with different mechanisms of cortical neuromodulation induced by therapy.Islam et al. 4 challenge the findings of our study. 2 In contrast to us, they did not study prototypical samples of patients with distinct types of cerebral lesions and corticospinal organization, but investigated a typical clinical sample of 16 children and adolescents chosen by their eligibility for CIMT according to clinical characteristics; both transcranial magnetic stimulation and MRI data were available for 11 participants. This sample included patients with contralateral, ipsilateral, and 'mixed type' cortico-spinal projections; underlying lesions comprised maldevelopments, white matter lesions and ischemic cortico-subcortical infarcts. Not surprisingly, in this heterogeneous sample, they could not identify any factors influencing the outcome of CIMT in individual participants. Most of them ameliorated in at least one of the hand function tests used in the study (Jebson-Taylor...