Unravelling upper limb function and treatment response in unilateral cerebral palsy: a behavioral and neurological perspective

The exploration of the environment mainly occurs through interaction with our hands: we touch, manipulate, and inspect objects to learn how to use them. In this process, both sensory and motor systems are coordinated to provide integrated information which contribute to an optimal learning and is crucial during the first period after birth. A disruption in the brain regions responsible for these interactions might have devastating consequences for the development. This is what happens to children suffering from unilateral Cerebral Palsy (CP), whose unilateral brain lesions causes motor and sensory disturbances, leading to immediate alterations of the required hand-environment interactions. Consequently, these deficits negatively influence the children’s independence in performing daily life activities as well as their quality of life.

The clinical presentation of the upper limb (hand and arm) problems that children with unilateral CP present with may depend on behavioral or neurological factors. Among the behavioral factors we can find the sensorimotor impairments, which have been shown to affect upper limb movement patterns. More specifically, these movement patterns can be affected by increased spasticity or muscle weakness. Among the neurological factors, we can include the characteristics of the brain lesion, the underlying wiring pattern of the main motor drive (the corticospinal tract) and the functional connectivity of the sensorimotor network. However, it seems that no single factor has enough power to explain the variability in upper limb function, and that the combination of these factors should be investigated in a larger cohort. In addition to understanding which are the behavioral and neurological factors that can explain the deficits in upper limb function, defining new therapy approaches is fundamental to help the child reach its maximum functional potential. A variety of upper limb function training approaches have been investigated, mostly all focusing on motor execution, and it seems that one size does not fit all. Other training approaches focusing as well on motor representation and action understanding, also known problems in children with unilateral CP, may be beneficial. The combination of different treatment approaches may result in larger gains and will contribute to the identification of predictors of treatment response.

The main scope of this doctoral thesis is threefold. First, we used a quantitative evaluation and analysis approach to better understand upper limb movement patterns in typically developing children and children with unilateral CP. We focused on determining the effect of age in typically developing children and ascertaining the impact of sensorimotor impairments on upper limb movement patterns in children with unilateral CP (Chapter 2). Second, we aimed to determine which neurological factors composed the best cocktail to understand the pathophysiology of upper limb function in children with unilateral CP (Chapter 3). Lastly, we developed a novel treatment approach combining modified constraint-induced movement therapy and action-observation training and investigated its effects on clinical and kinematic measures as well as the value of behavioral and neurological factors to predict treatment response (Chapter 4).

To achieve the first goal, we first evaluated whether upper limb movement patterns changed with age in typically developing children. Here, we found that upper limb motion reaches its maturation around the age of 11-12 years old, with movement duration, peak velocity, trajectory straightness, as well as joint kinematics reaching a plateau at this age (chapter 2.1). The reference values provided in this study helped to further optimize the interpretation of upper limb deficits in children with neurodevelopmental disorders. In chapter 2.2, we included both children with unilateral CP and typically developing children and identified pathological movement patterns in children with unilateral CP. Secondly, we mapped the negative impact of spasticity and muscle weakness on these movement patterns, providing useful insights that will contribute to treatment planning. A collateral result from this study was the identification of a subset of three relevant tasks for studying upper limb movements in children with unilateral CP, i.e. reaching upward, reach-to-grasp a vertically oriented cylinder, and hand-to-shoulder.  

As previous research had shed light onto the important role of brain lesion characteristics and the type of corticospinal tract wiring pattern for an adequate upper limb functioning, we aimed to further investigate these factors in chapter 3.1. Our first finding showed that a combination of lesion locations significantly contributed to differentiate between the corticospinal tract wiring groups, re-classifying the participants in their original group with 57% of accuracy. Secondly, motor function was predicted by the combination of the type of corticospinal tract wiring (more preserved in individuals with contralateral corticospinal tract wiring), lesion extent and damage to the basal ganglia and thalamus, whilst sensory deficits seemed to be best predicted by the combination of a large and later lesion, and an ipsilateral or bilateral corticospinal tract wiring. Lastly, we found that the underlying corticospinal tract wiring seemed to disrupt the association between sensory and motor function, pointing toward different mechanisms of sensorimotor integration in unilateral CP. Since we found large variability in upper limb functionality in the ipsilateral and bilateral corticospinal tract wiring groups, we conducted a second study evaluating the impact of functional connectivity of the sensorimotor network in participants with a periventricular lesion (without cortical damage) and further explored the additional role of the corticospinal tract wiring pattern to predict upper limb motor function (chapter 3.2). In this study, we found that aberrant sensorimotor functional connectivity seemed to be corticospinal tract-dependent rather than specific from all the unilateral CP population: in the dominant hemisphere, the contralateral corticospinal tract group showed increased connectivity between primary motor cortex and premotor cortices, whereas the bilateral corticospinal tract group showed higher connectivity between primary motor cortex and somatosensory association areas. Lastly, we found little impact of sensorimotor functional connectivity on upper limb motor function, suggesting that the corticospinal tract wiring pattern still is the main factor predicting upper limb function.

The last objective of this doctoral thesis was to develop and explore the effects of a novel intervention approach combining modified constraint-induced therapy (mCIMT) and action-observation training (AOT) on upper limb sensorimotor function. The rationale for this chapter was the necessity to develop new treatment strategies with a broaden focus on the upper limb sensorimotor deficits that children with unilateral CP typically present with. Effective treatment approaches, like mCIMT, are mainly targeting motor execution problems (i.e. movement quality or efficiency), whilst children with unilateral CP also have motor planning and motor representation deficits. These deficits could then be targeted with action-observation training (AOT), which is a novel neurophysiological-based treatment model that activates brain areas in charge of executing the movement that the participant is observing. The combination of mCIMT and AOT may provide new opportunities for enhanced motor learning. In chapter 4.1, we defined and described the protocol to be used in chapters 4.2 and 4.3 and stated our hypotheses. We developed an evaluator-blinded randomized controlled trial including 44 children aged between 6 and 12 years. Children were randomized according to their hand function, age and type of corticospinal tract wiring. They participated in a 2-week day-camp and received intensive mCIMT therapy for six hours a day on 9 out of 11 consecutive days (54h), including AOT or control condition (15h). Children who additionally received AOT watched goal-directed actions and executed the observed actions with the more impaired upper limb. The control group performed the same actions after watching human-motion-free computer games. Results did not show between-group differences after a two-week intervention, although both groups improved in all outcome measures and retained the gains at follow-up (chapter 4.2). Although the additional AOT to mCIMT did not seem to further improve upper limb function in the general population of children with unilateral CP, we found that it is beneficial in those children with an initially impaired motor function, as this subgroup showed higher gains in bimanual performance. As a response to mCIMT, with or without AOT, in this intensive camp model, we found that children with initially impaired sensory function showed higher gains in manual dexterity after the intervention. In addition, we found that all children with and without mirror movements could improve after a mCIMT intervention, although children who had stronger mirror movements between hands had difficulties in retaining the gains. Regarding neurological predictors, we found that all children improved after the intervention irrespective of the underlying type of the lesion or the corticospinal tract wiring pattern. However, the combination of these two factors could better predict treatment response. Lastly, chapter 4.3 documented the effects of this intervention (mCIMT+AOT) on upper limb kinematic measures. After the intervention, children receiving mCIMT+AOT became faster than the control group (mCIMT+placebo). The combined total group showed a smoother and more efficient execution of the tasks. Furthermore, we identified improvements in scapular and shoulder movements, which could not have been identified with other clinical assessments. The results of these studies highlighted interesting findings on response to this intensive therapy model and are therefore are a step forward toward individualized treatment planning in children with unilateral CP. The additional value of AOT requires further investigation when integrating it in an intensive training model. These insights are needed to transfer this training approach to the clinical practice.

In conclusion, this doctoral thesis increased our knowledge in the behavioral and neurological determinants of upper limb function in children with unilateral CP contributing to the understanding of the underlying pathophysiology of upper limb dysfunction. In addition, this thesis applied a novel treatment model and showed that its effects depend on behavioral and neurological characteristics of the children, providing crucial information in the clinical decision-making process.