Correlation between impaired dexterity and corticospinal tract dysgenesis in congenital hemiplegia

DOI: 10.1093/brain/awg069 Brain (2003), 126, 732±747 Correlation between impaired dexterity and corticospinal tract dysgenesis in congenital hemiplegia Julie Duque,1 Jean-Louis Thonnard,2 Yves Vandermeeren,1 Guillaume SeÂbire,3 Guy Cosnard4 and Etienne Olivier1 1Laboratory of Neurophysiology, 2Physical Medicine and Rehabilitation Unit, 3Department of Neuropediatrics and 4Neuroradiology Department, School of Medicine, Universite catholique de Louvain, Brussels, Belgium Summary One of the most devastating consequences of early corticospinal lesions is the impaired dexterity that results in a noticeable de®cit while manipulating small objects. One purpose of the present study was to investigate the extent to which a de®cit in the coordination of ®ngertip forces when grasping and lifting an object between the thumb and index ®nger could account for the impaired dexterity in patients with congenital hemiplegia (CH). A second objective was to examine whether, in these patients, de®cits in skilled hand movements are correlated with the importance of structural damage to the corticospinal tract. The scaling and coordination of ®ngertip forces during precision grip was investigated in 16 CH patients (aged 8±19 years) and 16 age- and sexmatched control subjects. Proprioception, stereognosis, pressure sensitivity and motor upper limb function (including digital and manual dexterity) were also assessed quantitatively. The structural damage of the corticospinal tract was estimated by measuring the cross-sectional area of cerebral peduncles with MRI and by calculating an index of symmetry between the two peduncles. In CH patients, a large number of parameters measured during the grip-lift task were signi®- Correspondence to: Professor Etienne Olivier, Laboratory of Neurophysiology, School of Medicine, Universite catholique de Louvain, 54 avenue Hippocrate, B-1200 Brussels, Belgium E-mail: cantly different when compared with those found in control subjects. Among those, the duration of the preloading and loading phases was signi®cantly longer in CH patients. In addition, both the dissimilarity and time-shift between the pro®les of the grip and load force rates, quanti®ed with the cross-correlation method, were also signi®cantly larger in CH patients; the time-shift was strongly correlated with impaired dexterity. These ®ndings suggest that impaired dextrous ®nger movements in CH patients may speci®cally result from their inability to ensure a precise synergy between ®ngertip forces while manipulating an object. Finally, the ®nding that the time-shift also correlated with the corticospinal tract dysgenesis, as estimated with the cerebral peduncle asymmetry, argues in favour of a critical role of the corticospinal system in the temporal coordination between different muscles involved in dextrous hand movements. Both digital and manual dexterity were also altered in the non-paretic hand of CH patients. This de®cit may reveal the contribution of the lesioned hemisphere to the control of ipsilateral skilled ®nger movements. Keywords: cerebral palsy; pyramidal tract; ®nger movement; grip-lift synergy Abbreviations: CH = congenital hemiplegia; GF = grip force; LF = loading force Introduction Manipulating small objects, and exploring their shape and texture, require highly skilled hand and ®nger movements. Although the biomechanical conditions for independent ®nger movements are present in many primates, the neural circuitry necessary for controlling these movements is most developed in higher primates, especially in man (Nakajima ã Guarantors of Brain 2003 et al., 2000). A substantial amount of anatomical, electrophysiological and clinical evidence suggests that the principal constituent of the motor system underlying the performance of highly skilled ®nger movements is the corticospinal pathway and, more speci®cally, its corticomotoneuronal component (see Porter and Lemon, 1993). The portion of Early corticospinal lesion and dexterity the corticospinal tract originating from the primary motor cortex is largely concerned with the control of muscles acting on the wrist and ®ngers, and earlier studies have suggested that the primary motor cortex is necessary, and largely suf®cient, to control skilled hand movements. However, recent functional imaging studies have shown that a large network including several other contralateral and ipsilateral cortical areas is also involved in the control of ®ne ®nger movements (Forssberg et al., 1999; Ehrsson et al., 2000, 2001). These observations suggest that, although the primary motor cortex is a prerequisite for the execution of skilled ®nger movements, their precise programming and control involve highly specialized brain structures. The manipulation of small objects between the tips of the thumb and index ®nger requires a precise coordination between the grip force (normal to the grip surface) and load force (tangential to the grip surface). This so-called `grip-lift synergy' is characterized by a smooth and parallel increase in both the grip and load forces, suggesting they are controlled in a predictive way (Westling and Johansson, 1984; Johansson and Westling, 1988). This anticipatory control was regarded as evidence for the existence of an internal representation of both the mechanical characteristics of the limbs and the object's physical properties in order to predict the consequences of voluntary movements; this process is thought to rely on an `internal forward model' (Flanagan and Wing, 1997; Witney et al., 2001a; Wolpert and Flanagan, 2001). After movement onset, the nervous system can optimally estimate the current state of the motor system by combining actual sensory feedback with the predictions of the forward model in order to update these internal representations (Wolpert et al., 2001). Therefore, even a simple movement such as grasping and loading an object involves subtle interplay between feed-forward and feedback mechanisms in order to generate a smooth vertical acceleration of the object (Witney et al., 2001b). The complexity of the grip-lift task was further revealed by developmental studies. Before the age of two, children use a pure feedback strategy while performing a grip-lift task, as shown by multiple successive increments in both grip and load force rates (Forssberg et al., 1992; Gordon et al., 1992). First signs of an anticipatory control of grip and load forces appear around the age of two and develop throughout childhood to reach adult performance at the age of eight years (Forssberg et al., 1991). This evolution of the grip-lift synergy parallels the development of manual skills (Forssberg et al., 1991) and may re¯ect the protracted maturation of the corticospinal tract in primates (Olivier et al., 1997; Paus et al., 1999) and the progressive implementation of the cortical circuitry responsible for the control of skilled ®nger movements (Forssberg et al., 1999; Ehrsson et al., 2000, 2001). Congenital h (...truncated)