To investigate the reorganization of somatosensory and motor cortex in congenital brain injury.
We recorded motor evoked potentials (MEPs) following transcranial magnetic stimulation (TMS) and somatosensory evoked potentials (SEPs) in a 41 year old man with severe congenital right hemiparesis but only mild proprioceptive impairment. Brain magnetic resonance imaging showed a large porencephalic cavitation in the left hemisphere mainly involving the frontal and parietal lobes.
TMS showed fast-conducting projections from the undamaged primary motor cortex to both hands, whereas MEPs were not elicited from the damaged hemisphere. Left median nerve stimulation evoked normal short-latency SEPs in the contralateral undamaged somatosensory cortex. Right median nerve stimulation did not evoke any SEP in the contralateral damaged hemisphere, but a middle-latency SEP (positive-negative-positive, 39-44-48 ms) in the ipsilateral undamaged hemisphere, with a fronto-central scalp distribution.
Our data show that somatosensory function of the affected arm is preserved, most likely through slow-conducting non-lemniscal connections between the affected arm and ipsilateral non-primary somatosensory cortex. In contrast, motor function was poor despite fast-conducting ipsilateral cortico-motoneuronal output from the primary motor cortex of the undamaged hemisphere to the affected arm. This suggests that different forms of reorganization operate in congenital brain injury and that fast-conducting connections between primary cortex areas and ipsilateral spinal cord are not sufficient for preservation or recovery of function.
"Thus, reduced and stereotypical pattern of spontaneous movements in patients with hemiplegic CP would result in abnormal sensory feedback and altered cortical reorganization, thus leading to asymmetric somatosensory processing deficits [25,26]. A case study by Ragazzoni and colleagues (2002)  has further showed that somatosensory function of the affected (right) arm was preserved, whereas motor function was poor despite fast-conducting ipsilateral cortico-motoneuronal output from primary motor cortex of the intact hemisphere to the affected arm. This finding seems to suggest that different forms of motor and somatosensory reorganization are involved in congenital brain injury, and that fast-conducting connections between primary cortex areas and ipsilateral spinal cord are not sufficient for preservation or recovery of function. "
[Show abstract][Hide abstract] ABSTRACT: Although cerebral palsy (CP) is usually defined as a group of permanent motor disorders due to non-progressive disturbances in the developing fetal or infant brain, recent research has shown that CP individuals are also characterized by altered somatosensory perception, increased pain and abnormal activation of cortical somatosensory areas. The present study was aimed to examine hemispheric differences on somatosensory brain processing in individuals with bilateral CP and lateralized motor impairments compared with healthy controls. Nine CP individuals with left-dominant motor impairments (LMI) (age range 5-28 yrs), nine CP individuals with right-dominant motor impairments (RMI) (age range 7-29 yrs), and 12 healthy controls (age range 5-30 yrs) participated in the study. Proprioception, touch and pain thresholds, as well as somatosensory evoked potentials (SEP) elicited by tactile stimulation of right and left lips and thumbs were compared.
Pain sensitivity was higher, and lip stimulation elicited greater beta power and more symmetrical SEP amplitudes in individuals with CP than in healthy controls. In addition, although there was no significant differences between individuals with RMI and LMI on pain or touch sensitivity, lip and thumb stimulation elicited smaller beta power and more symmetrical SEP amplitudes in individuals with LMI than with RMI.
Our data revealed that brain processing of somatosensory stimulation was abnormal in CP individuals. Moreover, this processing was different depending if they presented right- or left-dominant motor impairments, suggesting that different mechanisms of sensorimotor reorganization should be involved in CP depending on dominant side of motor impairment.
"Thirdly, in CH patients with variable degree of recovery and weak or absent MM, the presence of ipsilateral non-branched CS projections to the paretic hand's MNs has been suggested (Figure 1.2.C). The existence of ipsilateral CS projections is commonly admitted since ipsilateral MEPs (iMEPs) with short-onset latency compatible with a CS conduction velocity have been repeatedly reported in healthy neonates, CH or hemispherectomised patients (Benecke et al., 1991; Carr et al., 1993; Maegaki et al., 1997; McDonnell et al., 1999; Nirkko et al., 1997; Nezu et al., 1999; Eyre et al., 2001; Thickbroom et al., 2001; Ragazzoni et al., 2002; Staudt et al., 2002). These ipsilateral CS projections are thought to be unbranched from (i.e. "
[Show abstract][Hide abstract] ABSTRACT: The studies gathered in the present work were underlain by two main goals. Firstly, to
improve knowledge about the patterns of cortical areas and descending projection (re)organisation
after early brain injury resulting in CH, using recent and complementary electrophysiological
(TMS) and imaging methods (PET, fMRI and quantification of CS tract damage on MRI).
Secondly, to seek for correlations between these patterns of reorganisation and the residual function
of the paretic upper limb, especially fine dexterity, assessed by means of quantitative tools
(dexterity indexes, functional scales and evaluation of the grip-lift synergy).
The specific aim of each experimental part has been detailed in the last section of the
Introduction (Purposes of this work) and will only be briefly reminded hereafter. Firstly, does early
subcortical injury leading to CH induce diaschisis measurable by PET in the overlying cortical
areas, as typically observed in adult stroke patients? Secondly, what is the respective contribution of
the damaged and undamaged hemispheres to the control of the paretic (and non-paretic) finger
movements and in MM, as evaluated by means of fMRI in CH patients with a cortical lesion?
Thirdly, what is the functional relevance of the fMRI activation disclosed in the damaged and
undamaged hemisphere of one CH patient with MM? Fourthly, does TMS elicit long-latency MEPs
(as compared to short-latency MEPs) in CH patients and could such long-latency MEPs reveal
particular aspects of motor reorganisation or be correlated with residual hand function? Fifthly,
what is the precise cortical origin of the crossed and ipsilateral CS axons projecting onto the MNs
of the paretic hand in CH patients, as disclosed by TMS coregistered to MRI and what is the
respective importance of these contralateral and ipsilateral CS projections for the control of paretic
2. Summary of Chapters II - VI
The main results and conclusions gathered from the five experimental parts (Chapters II –
VI) of the present work could be summarised as follows. In the first experiment (Chapter II), the
resting brain metabolism estimated by means of regional uptake of FDG has been compared
between a group of 6 children with right CH of subcortical origin and a group of 6 healthy adults.
CH children had a relatively higher FDG uptake than control subjects in a region encompassing the
left M1S1 and IPL in the damaged hemisphere, and the right M1, callosomarginal sulcus and
cingulate gyrus in the undamaged hemisphere. This stands out in contrast with the cortical
diaschisis typically induced by a subcortical injury in adult stroke patients. The relatively higher
FDG uptake found in the damaged hemisphere likely reflects an increased synaptic density and/or
activity in the disconnected cortex of CH children, consistent with the imbalance between inhibitory
and excitatory synapses reported in the motor cortex of man and monkey after early brain injury
(Sloper et al., 1980; Marín-Padilla, 1997). Similarly, the increased FDG uptake in the undamaged
hemisphere may be due to a relative imbalance between transcallosal excitatory and inhibitory
connections. Finally, whereas CCD is a common finding in adult stroke patients, it has not been
observed in CH children. This observation confirmed that CCD is not the rule after an early brain
injury. It has been suggested that this was thanks to the adaptive development of the immature
corticopontocerebellar and cerebellothalamocortical projections (Kerrigan et al., 1991; Shamoto and
In the second experiment (Chapter III), the fMRI activation related to the performance of
unilateral sequential finger-to-thumb opposition has been compared between a group of 6 patients
with right CH of cortical origin and a group of 6 matched control subjects. In the control group,
movements of the right (dominant) or left hand asymmetrically activated both hemispheres, with a
predominant activation of the hemisphere contralateral to the moving hand. This contralateral
predominance was more marked with left fingers movements. In contrast, in the CH group, paretic
fingers movements activated both hemispheres but with a strong ipsilateral (right) predominance. A
second-order analysis revealed that the residual digital dexterity of the paretic hand (as evaluated by
the Purdue Pegboard score) correlated positively with the fMRI activation intensity in the ipsilateral
parietal and PM cortices and the contralateral SMA. This suggests that the undamaged hemisphere
(at least the intraparietal and PM cortices) was involved in the control of the ipsilateral paretic hand
in CH patients and could thus significantly contribute to functional recovery. As suggested by the
correlation between MM in the non-paretic passive hand and the M1S1 activation disclosed in the
undamaged hemisphere during paretic finger movements, a part of the undamaged M1S1 activation
may be related to the recruitment of crossed CS projections leading to MM and/or to sensory
feedback from the involuntary mirroring hand. During sequential finger opposition with the nonparetic
hand, the fMRI activation was restricted to the contralateral, undamaged hemisphere. Some
part of the subtle impairment observed in the non-paretic hand of CH may be related to this
excessive lateralisation of the non-paretic hand control to the contralateral, undamaged hemisphere.
Indeed, in healthy subjects, each cerebral hemisphere is at least partly involved in the
control of the ipsilateral upper limb (Kim et al., 1993a-b; Chen et al., 1997; Ehrsson et al., 2001;
Nirkko et al., 2001). These activation patterns disclosed for both the paretic and non-paretic hand of
CH suggested an extensive reorganisation of the cortical motor network, the undamaged
hemisphere being predominantly solicited during the performance of finger movements with either
the ipsilateral paretic or contralateral non-paretic hand than in age-matched controls.
However, the PET or fMRI activation disclosed in the ipsilateral undamaged hemisphere -
especially in M1S1- during paretic hand movement may also be at least partly related to MM or to
imbalance between transcallosal inhibitory connections (Heinen et al., 1999; Cramer and Bastings,
2000; Staudt et al., 2001b; Vandermeeren et al., 2003a). The reorganisation of both the CS tract and
cortico-subcortical networks involved in the performance of fine finger movements has been
respectively studied by means of TMS and fMRI in one CH child with a unilateral schizencephaly
involving S1M1 (Chapter IV). In this patient, the unilateral schizencephalic lesion led to a dramatic
functional reorganisation involving both hemispheres. During paretic finger movements, fMRI
activation was disclosed both in the dysgenic cortex bordering the schizencephalic cleft and in the
undamaged hemisphere. Movements of both the ipsilateral paretic and contralateral non-paretic
fingers induced a similar activation pattern in the undamaged hemisphere, which involved M1S1,
the premotor and parietal areas, as well as the thalamus. The finding that TMS of this area yielded
iMEPs in the paretic hand supported the functional relevance of the undamaged S1M1 activation.
cMEPs with an identical latency and morphology were simultaneously recorded from the nonparetic
hand, suggesting the presence of bilaterally branched CS axons or, at least, the existence of
ipsilateral CS axons with a conduction velocity equal to that of the crossed CS projections (Farmer
et al., 1991; Carr et al., 1993; Eyre et al., 2001). In contrast, the involvement of the dysgenic cortex
bordering the schizencephalic cleft in the direct control of paretic finger movements is questionable
since TMS of the damaged hemisphere failed to elicit cMEPs in the contralateral paretic hand.
Alternatively, the fMRI activation disclosed in this dysgenesic cortex may either reflect indirect
involvement in the control of paretic hand movements through transcallosal connections or be an
epiphenomenon such as a lack of transcallosal inhibition (Heinen et al., 1999) or an abnormal
metabolism in early injured/disconnected cortical area. This study further supports the importance
of using complementary methods in order to assess cerebral plasticity.
Whereas earlier studies have shown a correlation between the pattern of short-latency MEPs
recorded from the paretic hand in CH patients and functional outcome (Carr et al., 1993; Maegaki et
al., 1997), long-latency MEPs have barely been described in CH and it is not known whether longlatency
MEPs reflect specific aspects of motor reorganisation and correlate with residual motor
function. In the fourth experiment reported in this work (Chapter V), MEPs were recorded from
intrinsic muscles of both hands in 12 CH patients and 12 age-matched controls. Dexterity and upper
limb function were quantitatively assessed in both groups. In CH patients, in addition to shortlatency
MEPs (mean latency: 20 ms, MEP20), long-latency MEPs were frequently elicited in either
the paretic or non-paretic hand. Four clusters of long-latency MEPs were identified, each cluster
being characterised by its mean latency: MEP35, MEP85, MEP160 and MEP225. The long-latency
MEPs were elicited by TMS of the contralateral and/or ipsilateral hemisphere. In controls, only
MEP225 were scarcely observed. The dexterity of the paretic hand was positively correlated with the
presence of contralateral MEP20 and MEP225 and inversely correlated with the presence of
ipsilateral MEP20, MEP85 and MEP225. The pattern of MEPs found in CH patients differs
dramatically from that reported in adult stroke patients, suggesting that long-latency MEPs are a
rather distinctive consequence of early CS lesions. The presence of ipsilateral MEP20 reflected a
reorganisation of the CS projections and it appears unlikely that each long-latency MEPs cluster
corresponded to a specific descending pathway, except maybe for MEP35 which could reflect either
slowed CS projections or brainstem descending pathways (Benecke et al., 1991; Carr et al., 1993;
Turton et al., 1996; Müller K et al., 1997; Ziemann et al., 1999). Rather, long-latency MEPs could
be an exaggeration of the excitability changes observed in healthy subjects after a single-pulse
magnetic stimulation (Calancie et al., 1987; Boniface et al., 1991, 1994; Wassermann et al, 1994),
reflecting an exacerbation of the excitability in the spinal and/or cortical structures.
As already mentioned, in both CH and stroke patients performing movements with the
paretic hand, functional imaging studies have consistently shown an enhanced and more widespread
cortical activation than in controls (Chollet et al., 1991; Weiller et al., 1992, 1993; Cao et al., 1994,
1998; Staudt et al., 2000b, 2002; Johansen-Berg et al., 2002; Ward et al., 2003a). Since primate
studies have amply demonstrated the existence of redundant CM projections originating from
multiple (non-)primary motor areas of both hemisphere and converging onto the same MNs (Dum
and Strick, 1991; He et al, 1993, 1995; Galea and Darian-Smith, 1997a-b), it has been suggested
that the abnormal activations observed during paretic hand movements reflected the adaptive
recruitment of alternative CM projections. The fifth experiment (Chapter VI) was intended to settle
whether, after brain injury, the increased and widespread cortical activations could be interpreted in
terms of adaptive redistribution of the cortical drive through CM projections originating outside the
M1 hand knob, looking for the location of CS cells at the origin of these supposedly adaptive CM
connections. This question was addressed in 12 CH patients by mapping the cortical representation
of the paretic 1DI in both hemispheres by means of TMS co-registered into MRI. The centre of
gravity (CoG) of these representations was then computed, normalised into the stereotactic space
and compared with that found for the 1DI in 12 control subjects. When crossed CM projection from
the damaged hemisphere to paretic hand motoneurons were preserved, they invariably originated
from the hand knob of M1, exactly from the same location as in controls, and not from the
perilesional M1 zone or from the adjacent PM cortex. In 8/12 CH patients, TMS also disclosed
uncrossed CM projections to the paretic hand motoneurons originating from the undamaged
hemisphere, which also originated exclusively from the M1 hand knob. Therefore, despite a wide
range of functional recovery, lesion location and origin, the plasticity of the CS system is marginal
in terms of redistribution of CM cells. In spite of the large cortical zones and numerous cortical
areas that contain CS cells and that are consistently activated in functional imaging studies, only
cells in hand knob of M1 are able to establish significant and functional CM connections with the
hand's motoneurons of the paretic hand.
Institute of Neuroscience, Catholic University of Louvain, 12/2003, Degree: PhD, Supervisor: Pr Etienne Olivier
Data provided are for informational purposes only. Although carefully collected, accuracy cannot be guaranteed. The impact factor represents a rough estimation of the journal's impact factor and does not reflect the actual current impact factor. Publisher conditions are provided by RoMEO. Differing provisions from the publisher's actual policy or licence agreement may be applicable.