Bu makalede uzun dönemli cerebral palsi hastaları farklı yaş ve spastisite durumlarına göre manyetik rezonans görüntüleme yöntemleri ile EEG sonuçları eş zamanlı birleştirilerek harekete bağlı beyindeki değişiklikler ölçülmüş ve bunun sonucunda beynin yerine getiremediği motor aktiviteler için normal insan beyninden farklı olarak daha farklı yolları beyinde oluşturduğu ve vücut hareketlerini normale yakınlaştırmak için bir adaptasyon gösterdiği görülmektedir. Serebral palsi beyindeki farklı iletim yollarındaki hasara bağlı olarak gelişebilmektedir.
Verilen şekilde A grafiğinde normal bir insan beyninde nöral ileti yolları görülmektedir. Ard arda gelen diğer şekillerde ise serebral palsinin derecesine bağlı olarak olağan nöral yollardaki sorunlar kalın ve ince çizgilerle yada ileti yolunun hiç çizilmemesi şeklinde ifade edilmiştir. PVL periventriküler lökolomazi derecesi ise üste koyu ve grimsi çizgilerle ifade edilmiştir. Hekimler için serebral palsinin prognozunun kestiriminde cranial emar sonuçlarının istatistiksek olarak anlamlı bu verilerle kıyaslanarak yapılabilcek fizyoterapi programının yönlendirilmesi norolojik muayene ötesinde faydalı bir bilgi kaynağı oluşturmaktadır.
Pediatric Research:Volume 45(4, Part 1 of 2)April 1999pp 559-567
Central Motor Reorganization in Cerebral Palsy Patients with Bilateral Cerebral Lesions
[Regular Articles]
MAEGAKI, YOSHIHIRO; MAEOKA, YUKINORI; ISHII, SHOGO; EDA, ISEMATSU; OHTAGAKI, AYAMI; KITAHARA, TADASHI; SUZUKI, NORIKO; YOSHINO, KUNIO; IESHIMA, ATSUSHI; KOEDA, TATSUYA; TAKESHITA, KENZO
Division of Child Neurology, Institute of Neurologic Sciences, Faculty of Medicine, Tottori University, Yonaga 683-8504, Japan [Yo.M., Yu.M., Tat. K., K.T.]; Department of Pediatrics, East Shimane Rehabilitation Center, Yonago 690-0864, Japan [S.I.]; Department of Pediatrics, West Shimane Rehabilitation Center, Gotsu 695-0001, Japan [I.E., A.O.]; Kitakyushu City Sogo Ryoiku Center, Kitakyushu 802-0803, Japan [Tad. K.], Department of Pediatrics, National Nishitottori Hospital, Tottori 680-0901, Japan [N.S., K.Y.]; and Division of Child Neurology, Tottori Prefectural Kaike Rehabilitation Center for Disabled Children, Yonago 683-0004, Japan [A.I.]
Received April 7, 1998; accepted September 10, 1998.
Correspondence and reprint requests: Yoshihiro Maegaki, M.D., Division of Child Neurology, Institute of Neurologic Sciences, Faculty of Medicine, Tottori University, 36-1 Nishi-Machi, Yonago 683-8504, Japan.
Article Outline
ABSTRACT
METHODS
RESULTS
DISCUSSION
REFERENCES
Citing Articles Figures/Tables
Table 1
Table 2
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7
ABSTRACT TOP
Transcranial magnetic stimulation (TMS) has been used to describe cortical plasticity after unilateral cerebral lesions. The objective of this study was to find out whether cortical plasticity occurs after bilateral cerebral lesions. We investigated central motor reorganization for the arm and leg muscles in cerebral palsy (CP) patients with bilateral cerebral lesions using TMS. Seventeen patients (12 with spastic diplegia, 1 with spastic hemiplegia, and 4 with athetoid CP) and 10 normal subjects, were studied. On CT/MRI, bilateral periventricular leukomalacia was observed in all spastic patients with preterm birth. In two normal subjects, motor responses were induced in the ipsilateral tibialis anterior, but no responses were induced in any normal subject in the ipsilateral abductor pollicis brevis (APB) or biceps brachii (BB). Ipsilateral responses were more common among CP patients, especially in TMS of the less damaged hemisphere in patients with marked asymmetries in brain damage: in 3 abductor pollicis brevis, in 6 BBs, and in 15 tibialis anteriors. The cortical mapping of the sites of highest excitability demonstrated that the abductor pollicis brevis and BB sites in CP patients were nearly identical to those of the normal subjects. In patients with spastic CP born prematurely, a significant lateral shift was found for the excitability sites for the tibialis anterior. No similar lateral shift was observed in the other CP patients. These findings suggest that ipsilateral motor pathways are reinforced in both spastic and athetoid CP patients, and that a lateral shift of the motor cortical area for the leg muscle may occur in spastic CP patients with preterm birth.
Abbreviations APB, abductor pollicis brevis muscle; BB, biceps brachii muscle; CP, cerebral palsy; CST, corticospinal tract; MEP, motor evoked potential; PVL, periventricular leukomalacia; TA, tibialis anterior muscle; TMS, transcranial magnetic stimulation
Brain lesions early in fetal development, especially unilateral lesions, are less harmful than lesions acquired later, in both humans and animals, because the intact hemisphere can frequently compensate for the impairment. In animals, abundant ipsilateral corticospinal pathways develop after early hemispheric lesions, but are rare after lesions with adult onset (1,2). In humans, the relative contribution of the ipsilateral sensorimotor cortex to motor performance of the limbs has been shown by positron emission tomography (3) and transcranial Doppler ultrasonography (4).
Recently, the function of the motor system, including the corticospinal tract (CST), has been directly and noninvasively investigated by transcranial magnetic stimulation (TMS). An electrical current produced by discharging a bank of capacitors is passed through a wire coil placed over the scalp, producing a rapidly changing magnetic field that induces electrical currents in the brain (5). Transcranial magnetic stimulation has been applied in child subjects to investigate the maturation of the CST (6-8) and to examine central motor disturbances in patients with cerebral lesions (9,10). Ipsilateral hand motor responses to TMS are not usually elicited in normal subjects, but are frequently observed in patients with early brain lesions, especially congenital lesions (11-14). Ipsilateral motor responses in the proximal muscles of the arm are sometimes induced by TMS during muscle contraction, especially in young children (15-17). These responses may be facilitated by unilateral cerebral lesions in adults as well as in children (11,14,17-20).
However, there are few observations concerning reorganization or plasticity resulting from bilateral hemispheric lesions in either humans or animals. Therefore, we examined cerebral palsy (CP) patients with bilateral lesions for signs of central motor reorganization. We studied a group of patients with either spastic or athetoid CP whose symptoms and/or cranial images indicated bilateral hemispheric lesions, and compared these patients with a group of healthy volunteers. First, TMS was used to identify both contralateral and ipsilateral motor evoked potentials (MEPs) in three selected muscles. Second, the motor evoked potentials were used to map the cortical representations of the three muscles.
METHODS TOP
The study was approved by the Institutional Review Board. Informed consent was obtained from all adult subjects and from the parents of all children before the experiment.
Subjects. We studied 17 CP patients who have had no epileptic seizures (Table 1). The median age was 14 y (range: 10 to 49 y); 13 had spastic type disease and 4 had athetoid type. Of 13 spastic patients, 8 had diplegia with preterm birth, 1 had hemiplegia with preterm birth, and 4 had diplegia with full-term birth. Three of the 4 athetoid CP patients were born at full-term, the other prematurely. All 13 spastic patients showed milder motor impairments in the arms than the legs. Six of them had some lateralities in motor impairment. Six normal children (9 to 14 y old) and 4 normal adults (25 to 30 y old) were also studied as controls (median age, 14 y). Cranial computed tomography or magnetic resonance imaging was performed at 10 mm thickness in the axial section in all but one CP patient. Lesion severity of the periventricular white matter (periventricular leukomalacia, PVL) was assessed in each hemisphere using the scale from Okumura et al. (21): normal, mild, moderate, or severe. Periventricular leukomalacia was classified as mild when a volume loss or signal abnormality of the white matter was localized around the trigonal zone to the occipital horns; moderate when a white matter decrease extended to the bodies of lateral ventricles, and severe when a diffuse white matter decrease extended to the frontal horns. Periventricular leukomalacias were observed on both sides in all spastic patients born prematurely, including the hemiplegic patient, and in two of four spastic patients born at full-term (Table 1). Lesion severity was symmetrical in six patients and asymmetrical in five. Two diplegic patients with full-term birth showed mild cortical atrophies in the left fronto-parietal region, as well as bilateral PVLs. The other two diplegic patients with full-term birth had no abnormalities. There were no lesions in the three athetoid patients for whom images were available.
Table 1. Clinical and imaging characteristics of 17 patients with cerebral palsy evaluated for central motor reorganization
In this report, the "less damaged hemisphere" in spastic patients with motor laterality is the hemisphere contralateral to the limbs with the milder motor impairments, and the "more damaged hemisphere" is the one contralateral to the more impaired limbs. The presence of milder lesions in the less damaged hemisphere was confirmed with computed tomography or magnetic resonance images (Table 1).
Transcranial magnetic stimulation. Both normal subjects and CP patients tolerated TMS without any adverse effects during and after the procedure. Transcranial magnetic stimulation was carried out using a Magstrim 200 (Novametrix, Whitland, UK) with a figure 8-shaped magnetic coil (double 70 mm). Pulses had an approximate rise time of 100 µs and a duration of 1 ms. The current pulse through the coil was monophasic. The junction of the coil was held tangentially to the scalp with the loops of the coil equidistant from the scalp. The eddy current direction induced in the brain under the junction of the coil was anterior in the sagittal plane for the arm muscles, and lateral for the leg muscles (22,23). Motor evoked potentials were recorded using two surface electrodes positioned 3 cm apart over the muscle belly. Three muscles were studied; the tibialis anterior (TA), the biceps brachii (BB), and the abductor pollicis brevis (APB). Motor evoked potentials were recorded simultaneously on both sides of each muscle. Filters were set from 100 Hz to 5 kHz, and analysis time was 100 ms.
To identify the optimal stimulus intensity, consecutive stimulations were made with intensity increasing in increments of 5% of the maximum stimulator output. The resting threshold for each muscle was defined as the minimum intensity needed to produce two or more MEPs of at least 10 µV during four consecutive stimulations. Scalp landmarks were defined according to the 10-20 International System (24).
Identifying contralateral and ipsilateral motor responses. The center of the coil junction was positioned at the central region (C3 or C4) for the arm muscles because the most excitable areas for these muscles are near the central region (25). At this stimulus site, it is unlikely that the opposite motor cortex would be stimulated directly (13,17-20). The optimal site for the leg muscles, however, is near the vertex (23,26), meaning that the opposite hemisphere may be stimulated directly by the loops of the coil. Therefore, to study the presence of MEPs in the contralateral and ipsilateral TAs, the center of the coil junction was positioned at the scalp 4 to 5 cm lateral to the vertex on the imaginary line between the vertex and the auricle. The loops of the coil did not reach the opposite hemisphere at this coil position. Among healthy subjects, contralateral MEPs were only evoked with the coil in this position when they were relaxed (see "Results").
During TMS, each subject sat in a comfortable chair in a quiet room and was asked to relax the arms and legs. Using a stimulator intensity of 10% over the resting threshold for the ipsilateral muscles. MEPs were recorded in the contralateral and ipsilateral TAs, APBs, and BBs in all of the subjects. Measurements were made while the muscles were relaxed. When ipsilateral MEPs could not be evoked at maximum stimulator intensity, subjects were asked to gently contract the agonist muscles (about 50 to 100 µV of electromyographical activity) on both sides. The peak-to-peak amplitudes and the onset latencies were measured for every MEP. Transcranial magnetic stimulation was performed on both hemispheres of the CP patients and in the left hemisphere only of the healthy controls.
Mapping the cortical motor representations. A series of MEP recordings were made while moving the coil junction from the vertex in a line to the auricle in 1-cm increments, with the coil delivering a stimulus 10% higher than the resting threshold for the contralateral muscles. Contralateral and any ipsilateral MEPs were recorded simultaneously if they occurred. At each scalp position, four MEPs were recorded and the peak-to-peak amplitudes were averaged. The scalp sites at which the averaged amplitude was at least 50% of the maximum recorded amplitude for that muscle were considered to represent the excitable area. The motor cortical areas are considered to lie beneath the scalp positions where maximum MEPs could be evoked (27,28). Scalp positions were expressed as a percentage of the distance from the vertex to the auricle, so that the vertex was 0%, the region known under the 10-20 International System as C3 (4) was 40%, the region T3 (4) was 80%, and the auricle itself was 100%.
We examined the cortical motor representation sites for each of the three muscles in 10 normal subjects and in 11 CP patients (5 spastic patients with preterm birth, 4 spastic patients with full-term birth, and 2 athetoid patients). In spastic patients with motor laterality, TMS was applied to the less damaged hemisphere. In diplegic patients without motor laterality and in athetoid patients, TMS was carried out in the hemisphere with the lower resting threshold.
The Wilcoxon signed-rank test was used for examining differences between the cortical motor representation sites in normal subjects and in CP patients. Any difference was considered to be statistically significant if p ≤ 0.05 was found.
RESULTS TOP
Contralateral and ipsilateral motor responses. Transcranial magnetic stimulation could not be performed in two of four athetoid patients during a relaxed state because of their hypertonia. The presence of MEPs in the contralateral and ipsilateral arm and leg muscles are shown in Table 2. Contralateral MEPs were elicited in the relaxed APB and BB in all normal subjects. They were induced in the contracted APB and BB by TMS of each hemisphere in 16 of 17 CP patients. The hemiplegic patient had contralateral MEPs only after TMS of the less damaged hemisphere.
Table 2. Contralateral and ipsilateral motor evoked potentials after transcranial magnetic stimulation in 17 patients with cerebral palsy and in 10 normal subjects
Motor evoked potentials were never induced in the ipsilateral APB and BB, even during muscle contraction, in any normal subject or in any athetoid patient. In the ipsilateral APB, responses could be evoked in two spastic patients with motor laterality (patients 7 and 9) after TMS of the less damaged hemisphere and in one spastic patient without motor laterality (patient 2). Motor evoked potentials of the ipsilateral BB were evoked in three spastic patients with motor laterality (patients 6, 7, and 9) after TMS of the less damaged hemisphere. Three diplegic patients without laterality (patients 2, 3, and 13) also had ipsilateral MEPs. In these patients, the latencies of the ipsilateral MEPs were longer than those of the contralateral MEPs: the mean differences were 3.1 ms (SD = 2.6) in the APB and 2.5 ms (SD = 2.4) in the BB. The amplitudes were smaller in the ipsilateral MEPs than in the contralateral ones: the mean value of the amplitude ratios of the ipsilateral MEPs to the contralateral MEPs was 23.4% (SD = 31.0) in the APB and 40.7% (SD = 66.7) in the BB.
Motor evoked potentials were elicited in the contralateral TA during relaxation in all normal subjects. Motor evoked potentials of the contralateral TA could be evoked in five of the six spastic patients with motor laterality during relaxation and in all six during muscle contraction after TMS of the less damaged hemisphere, but only in one after TMS of the more damaged hemisphere, even during muscle contraction. In seven spastic patients without motor laterality, MEPs were elicited in the contralateral TA after TMS of the left hemisphere in three and the right hemisphere in four during relaxation. All seven patients had contralateral MEPs during contraction of the TA. Contralateral MEPs were evoked during TA contraction in three of the four athetoid patients.
In 10 healthy subjects, MEPs were never induced in the ipsilateral TA during relaxation, but they were elicited in two during muscle contraction (Fig. 1A and Table 2). In these two subjects, the latencies of the ipsilateral MEPs were similar to those of the contralateral ones: the differences were only 1.6 ms and 1.9 ms. The amplitudes of the ipsilateral MEPs were smaller than those of the contralateral ones: the ratios of the amplitude of the ipsilateral MEPs to the amplitude of the contralateral ones were 10.4% and 23.0%. Among the 17 CP patients, ipsilateral MEPs from the TA could be seen in 15. Ipsilateral MEPs after TMS of the less damaged hemisphere were evoked during relaxation in four of six spastic patients with motor laterality (patients 6, 7, 10, and 11) and only during muscle contraction in the other two (patients 8 and 9) (Fig. 1B). In the same patients, ipsilateral MEPs were never induced by TMS of the more damaged hemisphere during relaxation and were induced in only one patient (patient 10) during muscle contraction. Of seven spastic patients without motor laterality, ipsilateral MEPs could be elicited by TMS of the left hemisphere in two (patients 1 and 13) and by TMS of the right hemisphere in three (patients 1, 2, and 13) during relaxation. They could be induced in six patients during muscle contraction in both sides (Fig. 1C and Table 2). In the four athetoid patients, ipsilateral MEPs could be induced from the left hemisphere in three and from the right hemisphere in one during muscle contraction (Fig. 1D). There were no differences in latency between the ipsilateral and the contralateral MEPs in any of the CP patients except one (patient 13, 9.3 and 17.0 ms); the mean difference between the ipsilateral and contralateral MEP latencies was 0.95 ms (SD = 4.2) (23 values from 15 patients). The ratios of the amplitude of the ipsilateral MEPs to the amplitude of the contralateral ones were higher in CP patients than in controls; the mean value was 74.9% (SD = 42.2).
Figure 1. MEPs from the tibialis anterior after TMS in a normal subject (A), in patients with spastic CP (B and C), and athetoid CP (D). (A) Ipsilateral MEPs could be seen only during muscle contraction and had similar latencies and smaller amplitude than the contralateral ones. (B) Bilateral MEPs could be evoked only after TMS of the less damaged left hemisphere in patient 11, who had motor laterality in her arm and leg functions. (C) Ipsilateral MEPs were elicited after TMS of each hemisphere in patient 4 who had no motor laterality. (D) In patient 15, bilateral MEPs were evoked after TMS of the left hemisphere, and only contralateral MEPs could be seen after TMS of the right hemisphere. Ipsilateral MEPs have latency and amplitude similar to contralateral ones in all of these CP patients (B, C, and D). L = left; r = right. Arrows mark the onset of the responses.
Cortical motor representation sites for each muscle. In healthy controls, the cortical representation sites for the APB were on the contralateral hemisphere 5 to 8 cm lateral to the cranial vertex (mean, 34.1%; SD = 5.6%). The sites for the BB were more medial, only 4 to 6 cm lateral to the vertex (mean, 26.3%; SD = 2.5%) (Fig 2). In the 11 CP patients who were included in this part of the study, the cortical representation sites for the APB were 5 to 8 cm lateral to the vertex (mean, 32.4%; SD = 6.6%) and the sites for the BB, 4 to 8 cm (mean, 28.0%; SD = 4.8%). There were no significant differences in the cortical representation sites for the APB and BB between normal subjects and CP patients, or between normal subjects and spastic patients or athetoid patients (Fig. 3).
Figure 2. Functional cortical mapping of the contralateral MEPs from the tibialis anterior (TA), biceps brachii (BB), and abductor pollicis brevis (APB) by TMS at 1-cm intervals between the vertex and auricle in 10 normal subjects and in 11 CP patients. Four trials of TMS were carried out at each scalp site with the subject relaxed, and the amplitudes were averaged. Solid boxes indicate scalp sites where maximum MEPs were induced. Dotted areas indicate scalp sites where MEPs with an amplitude above 50% of the maximum were evoked. For the TA, a marked lateral shift occurred among patients 1, 2, 6, 7, and 9, all of whom were spastic patients who had been born prematurely.
Figure 3. Cortical motor representation sites (mean ± SD) of maximum responses from the contralateral tibialis anterior, biceps brachii, and abductor pollicis brevis in controls (n = 10) and in CP patients (n = 11). The cortical site is expressed as the percentage of the distance between the vertex and the auricle. The cortical motor representation site for the TA was shifted laterally in CP patients, especially spastic patients with preterm birth, compared with controls. Conversely, in the BB and APB, there were no significant differences in the cortical site of maximum responses between controls and CP patients. *p ≤0.05; **p ≤0.005
Figure 4 shows typical examples of MEP maps of the TA in one normal subject, three diplegic patients, and one athetoid patient. Cortical motor representation sites for the TA were medial to those of the arm muscles and lay within 4 cm of the vertex in normal subjects (mean, 14.7%; SD = 4.0%) (Figs. 2, 3, and 4A). In five spastic patients born prematurely (patients 1, 2, 6, 7, and 9), the most excitable cortical sites for the TA were lateral to those obtained in normal subjects, which ranged from 5 to 7 cm (mean, 29.5%; SD = 3.5%) (Fig. 4B). The sites for the TA were within the sites for the APB and BB (Fig. 3). One of four spastic patients with full-term birth (patient 10) had two peaks in MEP amplitude (Figs. 2 and 4C). The other three of four spastic patients with full-term birth (patients 11, 12, and 13) and two athetoid patients (patients 14 and 17) had cortical motor representation sites for the TA that were similar to those of the normal subjects (Figs. 2, 3, and 4D and 4E).
Figure 4. Contralateral MEPs of the tibialis anterior in a normal subject (A), a diplegic patient with preterm birth (B), two diplegic patients with full-term birth (C and D), and an athetoid patient (E). The locations of the superimposed MEPs correspond to the scalp positions where they were evoked. (A) The maximum MEPs were induced at the scalp 3 cm lateral to the vertex. (B) In patient 1, the maximum MEPs were evoked at a site 5 cm lateral to the vertex. (C) Patient 10 had two peaks in amplitude; a medial peak at 2 cm and a lateral peak at 6 cm. (D) Patient 13 had maximum MEPs at a site 2 cm. (E) In patient 17, the maximum MEPs were evoked at a site 2 cm.
Motor evoked potentials were found in the ipsilateral TA in 2 of 10 healthy subjects while they were relaxed (Fig. 5). When ipsilateral MEPs induced, the coil was 1 or 2 cm lateral to the vertex. The latencies of the ipsilateral MEPs were similar to those of the contralateral MEPs. Ipsilateral MEPs were evoked in seven spastic patients: four of them were preterm birth patients (patients 1, 2, 6, and 7) and three were full-term birth patients (patients 10, 11, and 13). The threshold intensity for the ipsilateral TA was the same as for the contralateral TA in six of these patients; only one patient had a higher threshold for the ipsilateral TA than the contralateral one. Ipsilateral MEPs were elicited by TMS at lateral scalp sites along the vertex-auricle line in all patients with preterm birth and in one patient with full-term birth (patient 13) (Fig. 6). The scalp sites eliciting the highest ipsilateral MEPs were 5 to 7 cm lateral to the vertex. In these patients, TMS near the vertex did not induce any ipsilateral MEPs (Fig. 6). In two patients with full-term birth (patients 10 and 11), ipsilateral MEPs were present diffusely, between the vertex and 9 cm laterally, peaking at 1 or 2 cm lateral to the vertex. The most excitable cortical sites for the ipsilateral TA corresponded to those for the contralateral TA in three of seven patients (patients 1, 6, and 11) (Fig. 6A). The highest ipsilateral MEPs were elicited with the coil 1 cm away from the highest cortical MEP sites for the contralateral TA in two patients (patients 2 and 7), 4 cm apart in 1 (patient 10), and 5 cm apart in 1 (patient 13) (Fig. 6B).
Figure 5. Contralateral and ipsilateral MEPs from the tibialis anterior elicited by TMS on the left of the vertex in a normal subject. The locations of the superimposed MEPs correspond to the scalp positions where they were evoked. TMS was performed while muscles were relaxed. Both ipsilateral and contralateral MEPs were present at the sites 1 and 2 cm lateral to the vertex, but only contralateral MEPs were induced at the more lateral sites.
Figure 6. Ipsilateral MEPs of the tibialis anterior in two diplegic patients, one with preterm birth (A) and one with full-term birth (B). The locations of the superimposed MEPs correspond to the scalp positions where they were evoked. The maximum MEPs were induced at a site 5 cm in patient 1 (A) and at a site 7 cm in patient 13 (B). The solid circles mark the location of the maximum contralateral MEPs in each patient. (see Figs. 4B and 4D).
DISCUSSION TOP
Ipsilateral MEPs of the TA were found in most of the CP patients. For five reasons, we believe that these were true readings, not artifacts from inadvertently stimulating the wrong brain hemisphere.
1) The coil was placed 4 to 5 cm lateral to the vertex, a position that in our experience and that of other investigators (13,17-20) is not likely to result in stimulation of the opposite hemisphere.
2) In four of the six spastic patients with motor laterality, ipsilateral MEPs were evoked by stimulation of the less damaged hemisphere during relaxation, but contralateral responses could not be induced from the more damaged hemisphere in any of them.
3) Cortical mapping in five of the seven spastic patients showed that the sites eliciting the strongest MEPs in the ipsilateral TA were at least 5 cm from the vertex. In these patients, ipsilateral MEPs were not evoked with the coil placed near the vertex.
4) One patient (patient 13) had ipsilateral MEP latencies that were moderately longer than the contralateral ones, suggesting that the ipsilateral pathway was truly different from the contralateral one.
5) The TMS stimulus would require about 10 ms to pass through the corpus callosum to the opposite hemisphere (29), but in almost all of the subjects (both normal and CP patients), the ipsilateral MEP latencies were only a few milliseconds, at most, longer than the contralateral ones.
Motor-evoked potentials from the ipsilateral TA were common among both the spastic and the athetoid patients and both the preterm birth and full-term birth patients. In all spastic CP patients with motor laterality, they could be evoked by TMS of the less damaged hemisphere. Ipsilateral MEP amplitudes were larger in CP patients than in normal subjects. Therefore, we suggest that ipsilateral motor pathways might develop after early brain lesions and that in patients with bilateral lesions, these alternate pathways might develop preferentially in the less damaged hemisphere (Fig. 7B-7E). When the brain damage is slight in both hemispheres, ipsilateral motor pathways might be facilitated in both hemispheres (Fig. 7F-7I). Motor evoked potentials were induced more frequently in the ipsilateral TA than in the ipsilateral APB and BB in both CP patients and normal subjects, suggesting that ipsilateral motor pathways may be more abundant in the legs than in the arms.
Figure 7. Possible mechanisms underlying ipsilateral MEPs in the tibialis anterior after bilateral cerebral lesions. (A) In normal humans, proximal limb muscles are believed to receive inputs via large crossed and small uncrossed corticospinal fibers and via bilateral cortico-reticulospinal fibers (preferentially contralateral). (B-E) In patients with markedly asymmetrical brain damage, ipsilateral motor pathways may be facilitated in the less damaged hemisphere: ipsilateral corticospinal pathways (B), cortico-reticulospinal pathways (C), axon collaterals (D), or persisting fetal pathways (E). (F-I) In patients with mild cerebral lesions in both hemispheres, ipsilateral motor pathways may be facilitated on both sides: ipsilateral corticospinal pathways (F), cortico-reticulospinal pathways (G), axon collaterals (H), or persisting fetal pathways (I).
The mechanisms of the formation of an ipsilateral motor pathway are proposed as follows: that normally occurring ipsilateral projections (Fig. 7A), such as 1) corticospinal tracts (Figs. 7B and 7F) and 2) cortico-reticulospinal tracts (Figs. 7C and 7G), establish more extensive and enhanced connections; 3) that ipsilateral projections originate from an abnormal branching of the contralateral corticospinal axons (Figs. 7D and 7H); and 4) that transient fetal connections persist without normal regression (Figs. 7E and 7I). Motor evoked potentials in the ipsilateral TA with latencies similar to the contralateral MEPs may indicate fast-conducting ipsilateral CST rather than polysynaptic pathways (Fig. 7B and 7F) (11). Benecke et al. (11) reported that ipsilateral MEPs, including those in the arm and leg muscles, were evoked by TMS and had amplitudes and latencies similar to the contralateral MEPs in hemiplegic patients with early brain lesions, but that ipsilateral MEPs had considerably smaller amplitudes and delayed latencies in patients with later brain lesions. They suggested that the ipsilateral corticospinal pathway might be reinforced after early brain lesions and that the cortico-reticulospinal pathway might be facilitated after later brain lesions. In kittens, the terminal fields of motor cortical neurons are widely and bilaterally distributed in the spinal cord in the early postnatal days, but in adulthood, the majority are localized to the contralateral side (30). Early brain lesions may prevent such normal elimination of the ipsilateral motor pathways (30,31).
Another normally occurring ipsilateral pathways via the brain-stem, cortico-reticulospinal tracts, might be facilitated (Fig. 7C and 7G). MEP latencies through such pathways may be prolonged (11,14,16,18,19). The only spastic patient with full-term birth (patient 13) had ipsilateral MEPs with latencies that were moderately longer than the contralateral ones. Such cortico-reticulospinal pathways or slow-conducting ipsilateral corticospinal pathways might be facilitated in this patient. Because there were some differences in latency and amplitude between the ipsilateral and contralateral MEPs from the APB and the BB, such pathways might be also facilitated in the arms.
A third possibility is that ipsilateral MEPs might originate from abnormal branching of the contralateral corticospinal axons (Fig. 7D and 7H) (1,2). Bilateral hand motor responses with similar latencies and amplitudes after TMS of the intact hemisphere are elicited in hemiplegic patients with congenital or early postnatal lesions (11-14). Using TMS mapping in hemiplegics with congenital and early postnatal lesions, we recently found that the cortical motor representation sites for the paretic hand were close to or at the same scalp sites as the intact hand (14,32). In our study, cortical motor representation sites for the ipsilateral TA were the same as or in the vicinity of those for the contralateral TA in five of seven patients. Axon collaterals are one possible mechanism for the observed ipsilateral MEPs.
The most striking finding was a lateral shift of the cortical motor representation sites for the TA in spastic patients born prematurely. This is consistent with the bilateral hypoperfusion in the superior motor cortex observed in diplegic CP patients on single photon emission computed tomography (33). To our knowledge, this is the first observation in humans of a lateral shift of the cortical motor representation area after a cerebral lesion.
One possible mechanism is the persistence of transient fetal connections without normal elimination (Figs. 7E and 7I). In the undeveloped rat brain, many cortical neurons, including those in the occipital cortex, have corticospinal projections of transient axon collaterals and such collateral projections are eliminated during development (34-36). Also, abundant corticospinal axons of the neonatal rat brain decrease rapidly after early postnatal days (36,37). Contralateral hemisphere hypertrophy caused by a unilateral neonatal brain lesion has been reported in adult rats (31,38). Therefore, the CST axons from the cortical neurons other than the motor cortex for the legs would persist without normal elimination and innervate the spinal motoneurons because of the PVL, which is believed to affect descending fibers for the legs more than the laterally placed fibers for the arms (39). Interestingly, patient 10, a diplegic patient who was born at full-term, had two peaks in MEP amplitude on the vertex-auricle line. She might have had brain lesions during the fetal period because she had only slight neonatal asphyxia at birth. If CST damage to the legs was partial, both a medial cortical peak (the usual cortical representation site) and a compensatory lateral peak might have developed (Figs. 2 and 4C).
Alternately, the shift in the motor representation site may have resulted from unmasking preexisting synaptic connections. Corticocortical connections are abundant within the motor cortex and between the motor cortex and the other cortical areas (40,41). Enlargement of the motor representation area has been reported after limb amputation (42,43), spinal cord injury (44), and facial palsy (45). In a positron emission tomography study of patients with cerebral lesions, Weiller et al. (46) reported a large extension of the hand field in the damaged hemisphere after capsular infarction. Adjacent areas of the motor cortex may be activated after disruption of the efferent pyramidal pathway or the afferent pathway (deafferentation) by deprivation of intercortical inhibition (47). In our study, the excitable areas for the TA were not enlarged, but they moved laterally in five spastic patients with preterm birth. In addition, they did not include the usual cortical representation areas for the TA in four of these patients (Figs. 2 and 4B). Therefore, it seems unlikely that this phenomenon results from unmasking preexisting connections.
No lateral deviation of the motor cortical site for the TA was observed in three of four diplegic patients with full-term birth, indicating that there may be critical windows for cortical reorganization. Also, two athetoid patients, one born prematurely and the other at full-term, did not show the lateral deviation of the motor cortical area. Such cortical reorganization may be associated only with CST lesions.
We cannot definitively confirm the mechanisms of the cortical reorganization based on these limited data. One limitation is the small number of patients studied. A second limitation is that the precise timing of the occurrence of the brain lesions is unknown. Further studies are necessary to confirm the mechanisms of the cortical plasticity after bilateral cerebral lesions both in humans and animals.
In conclusion, our study shows evidence of two types of plasticity after early bilateral cerebral lesions: facilitation of ipsilateral motor pathways and lateral deviation of the motor cortical area for the legs. Ipsilateral motor pathways were reinforced both in spastic and athetoid CP patients, with preterm and full-term births. The cortical motor representation area for the TA moved laterally toward the area for the arm muscles in spastic patients with preterm birth, but not in spastic patients with full-term birth and athetoid patients.
Verilen şekilde A grafiğinde normal bir insan beyninde nöral ileti yolları görülmektedir. Ard arda gelen diğer şekillerde ise serebral palsinin derecesine bağlı olarak olağan nöral yollardaki sorunlar kalın ve ince çizgilerle yada ileti yolunun hiç çizilmemesi şeklinde ifade edilmiştir. PVL periventriküler lökolomazi derecesi ise üste koyu ve grimsi çizgilerle ifade edilmiştir. Hekimler için serebral palsinin prognozunun kestiriminde cranial emar sonuçlarının istatistiksek olarak anlamlı bu verilerle kıyaslanarak yapılabilcek fizyoterapi programının yönlendirilmesi norolojik muayene ötesinde faydalı bir bilgi kaynağı oluşturmaktadır.
Pediatric Research:Volume 45(4, Part 1 of 2)April 1999pp 559-567
Central Motor Reorganization in Cerebral Palsy Patients with Bilateral Cerebral Lesions
[Regular Articles]
MAEGAKI, YOSHIHIRO; MAEOKA, YUKINORI; ISHII, SHOGO; EDA, ISEMATSU; OHTAGAKI, AYAMI; KITAHARA, TADASHI; SUZUKI, NORIKO; YOSHINO, KUNIO; IESHIMA, ATSUSHI; KOEDA, TATSUYA; TAKESHITA, KENZO
Division of Child Neurology, Institute of Neurologic Sciences, Faculty of Medicine, Tottori University, Yonaga 683-8504, Japan [Yo.M., Yu.M., Tat. K., K.T.]; Department of Pediatrics, East Shimane Rehabilitation Center, Yonago 690-0864, Japan [S.I.]; Department of Pediatrics, West Shimane Rehabilitation Center, Gotsu 695-0001, Japan [I.E., A.O.]; Kitakyushu City Sogo Ryoiku Center, Kitakyushu 802-0803, Japan [Tad. K.], Department of Pediatrics, National Nishitottori Hospital, Tottori 680-0901, Japan [N.S., K.Y.]; and Division of Child Neurology, Tottori Prefectural Kaike Rehabilitation Center for Disabled Children, Yonago 683-0004, Japan [A.I.]
Received April 7, 1998; accepted September 10, 1998.
Correspondence and reprint requests: Yoshihiro Maegaki, M.D., Division of Child Neurology, Institute of Neurologic Sciences, Faculty of Medicine, Tottori University, 36-1 Nishi-Machi, Yonago 683-8504, Japan.
Article Outline
ABSTRACT
METHODS
RESULTS
DISCUSSION
REFERENCES
Citing Articles Figures/Tables
Table 1
Table 2
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7
ABSTRACT TOP
Transcranial magnetic stimulation (TMS) has been used to describe cortical plasticity after unilateral cerebral lesions. The objective of this study was to find out whether cortical plasticity occurs after bilateral cerebral lesions. We investigated central motor reorganization for the arm and leg muscles in cerebral palsy (CP) patients with bilateral cerebral lesions using TMS. Seventeen patients (12 with spastic diplegia, 1 with spastic hemiplegia, and 4 with athetoid CP) and 10 normal subjects, were studied. On CT/MRI, bilateral periventricular leukomalacia was observed in all spastic patients with preterm birth. In two normal subjects, motor responses were induced in the ipsilateral tibialis anterior, but no responses were induced in any normal subject in the ipsilateral abductor pollicis brevis (APB) or biceps brachii (BB). Ipsilateral responses were more common among CP patients, especially in TMS of the less damaged hemisphere in patients with marked asymmetries in brain damage: in 3 abductor pollicis brevis, in 6 BBs, and in 15 tibialis anteriors. The cortical mapping of the sites of highest excitability demonstrated that the abductor pollicis brevis and BB sites in CP patients were nearly identical to those of the normal subjects. In patients with spastic CP born prematurely, a significant lateral shift was found for the excitability sites for the tibialis anterior. No similar lateral shift was observed in the other CP patients. These findings suggest that ipsilateral motor pathways are reinforced in both spastic and athetoid CP patients, and that a lateral shift of the motor cortical area for the leg muscle may occur in spastic CP patients with preterm birth.
Abbreviations APB, abductor pollicis brevis muscle; BB, biceps brachii muscle; CP, cerebral palsy; CST, corticospinal tract; MEP, motor evoked potential; PVL, periventricular leukomalacia; TA, tibialis anterior muscle; TMS, transcranial magnetic stimulation
Brain lesions early in fetal development, especially unilateral lesions, are less harmful than lesions acquired later, in both humans and animals, because the intact hemisphere can frequently compensate for the impairment. In animals, abundant ipsilateral corticospinal pathways develop after early hemispheric lesions, but are rare after lesions with adult onset (1,2). In humans, the relative contribution of the ipsilateral sensorimotor cortex to motor performance of the limbs has been shown by positron emission tomography (3) and transcranial Doppler ultrasonography (4).
Recently, the function of the motor system, including the corticospinal tract (CST), has been directly and noninvasively investigated by transcranial magnetic stimulation (TMS). An electrical current produced by discharging a bank of capacitors is passed through a wire coil placed over the scalp, producing a rapidly changing magnetic field that induces electrical currents in the brain (5). Transcranial magnetic stimulation has been applied in child subjects to investigate the maturation of the CST (6-8) and to examine central motor disturbances in patients with cerebral lesions (9,10). Ipsilateral hand motor responses to TMS are not usually elicited in normal subjects, but are frequently observed in patients with early brain lesions, especially congenital lesions (11-14). Ipsilateral motor responses in the proximal muscles of the arm are sometimes induced by TMS during muscle contraction, especially in young children (15-17). These responses may be facilitated by unilateral cerebral lesions in adults as well as in children (11,14,17-20).
However, there are few observations concerning reorganization or plasticity resulting from bilateral hemispheric lesions in either humans or animals. Therefore, we examined cerebral palsy (CP) patients with bilateral lesions for signs of central motor reorganization. We studied a group of patients with either spastic or athetoid CP whose symptoms and/or cranial images indicated bilateral hemispheric lesions, and compared these patients with a group of healthy volunteers. First, TMS was used to identify both contralateral and ipsilateral motor evoked potentials (MEPs) in three selected muscles. Second, the motor evoked potentials were used to map the cortical representations of the three muscles.
METHODS TOP
The study was approved by the Institutional Review Board. Informed consent was obtained from all adult subjects and from the parents of all children before the experiment.
Subjects. We studied 17 CP patients who have had no epileptic seizures (Table 1). The median age was 14 y (range: 10 to 49 y); 13 had spastic type disease and 4 had athetoid type. Of 13 spastic patients, 8 had diplegia with preterm birth, 1 had hemiplegia with preterm birth, and 4 had diplegia with full-term birth. Three of the 4 athetoid CP patients were born at full-term, the other prematurely. All 13 spastic patients showed milder motor impairments in the arms than the legs. Six of them had some lateralities in motor impairment. Six normal children (9 to 14 y old) and 4 normal adults (25 to 30 y old) were also studied as controls (median age, 14 y). Cranial computed tomography or magnetic resonance imaging was performed at 10 mm thickness in the axial section in all but one CP patient. Lesion severity of the periventricular white matter (periventricular leukomalacia, PVL) was assessed in each hemisphere using the scale from Okumura et al. (21): normal, mild, moderate, or severe. Periventricular leukomalacia was classified as mild when a volume loss or signal abnormality of the white matter was localized around the trigonal zone to the occipital horns; moderate when a white matter decrease extended to the bodies of lateral ventricles, and severe when a diffuse white matter decrease extended to the frontal horns. Periventricular leukomalacias were observed on both sides in all spastic patients born prematurely, including the hemiplegic patient, and in two of four spastic patients born at full-term (Table 1). Lesion severity was symmetrical in six patients and asymmetrical in five. Two diplegic patients with full-term birth showed mild cortical atrophies in the left fronto-parietal region, as well as bilateral PVLs. The other two diplegic patients with full-term birth had no abnormalities. There were no lesions in the three athetoid patients for whom images were available.
Table 1. Clinical and imaging characteristics of 17 patients with cerebral palsy evaluated for central motor reorganization
In this report, the "less damaged hemisphere" in spastic patients with motor laterality is the hemisphere contralateral to the limbs with the milder motor impairments, and the "more damaged hemisphere" is the one contralateral to the more impaired limbs. The presence of milder lesions in the less damaged hemisphere was confirmed with computed tomography or magnetic resonance images (Table 1).
Transcranial magnetic stimulation. Both normal subjects and CP patients tolerated TMS without any adverse effects during and after the procedure. Transcranial magnetic stimulation was carried out using a Magstrim 200 (Novametrix, Whitland, UK) with a figure 8-shaped magnetic coil (double 70 mm). Pulses had an approximate rise time of 100 µs and a duration of 1 ms. The current pulse through the coil was monophasic. The junction of the coil was held tangentially to the scalp with the loops of the coil equidistant from the scalp. The eddy current direction induced in the brain under the junction of the coil was anterior in the sagittal plane for the arm muscles, and lateral for the leg muscles (22,23). Motor evoked potentials were recorded using two surface electrodes positioned 3 cm apart over the muscle belly. Three muscles were studied; the tibialis anterior (TA), the biceps brachii (BB), and the abductor pollicis brevis (APB). Motor evoked potentials were recorded simultaneously on both sides of each muscle. Filters were set from 100 Hz to 5 kHz, and analysis time was 100 ms.
To identify the optimal stimulus intensity, consecutive stimulations were made with intensity increasing in increments of 5% of the maximum stimulator output. The resting threshold for each muscle was defined as the minimum intensity needed to produce two or more MEPs of at least 10 µV during four consecutive stimulations. Scalp landmarks were defined according to the 10-20 International System (24).
Identifying contralateral and ipsilateral motor responses. The center of the coil junction was positioned at the central region (C3 or C4) for the arm muscles because the most excitable areas for these muscles are near the central region (25). At this stimulus site, it is unlikely that the opposite motor cortex would be stimulated directly (13,17-20). The optimal site for the leg muscles, however, is near the vertex (23,26), meaning that the opposite hemisphere may be stimulated directly by the loops of the coil. Therefore, to study the presence of MEPs in the contralateral and ipsilateral TAs, the center of the coil junction was positioned at the scalp 4 to 5 cm lateral to the vertex on the imaginary line between the vertex and the auricle. The loops of the coil did not reach the opposite hemisphere at this coil position. Among healthy subjects, contralateral MEPs were only evoked with the coil in this position when they were relaxed (see "Results").
During TMS, each subject sat in a comfortable chair in a quiet room and was asked to relax the arms and legs. Using a stimulator intensity of 10% over the resting threshold for the ipsilateral muscles. MEPs were recorded in the contralateral and ipsilateral TAs, APBs, and BBs in all of the subjects. Measurements were made while the muscles were relaxed. When ipsilateral MEPs could not be evoked at maximum stimulator intensity, subjects were asked to gently contract the agonist muscles (about 50 to 100 µV of electromyographical activity) on both sides. The peak-to-peak amplitudes and the onset latencies were measured for every MEP. Transcranial magnetic stimulation was performed on both hemispheres of the CP patients and in the left hemisphere only of the healthy controls.
Mapping the cortical motor representations. A series of MEP recordings were made while moving the coil junction from the vertex in a line to the auricle in 1-cm increments, with the coil delivering a stimulus 10% higher than the resting threshold for the contralateral muscles. Contralateral and any ipsilateral MEPs were recorded simultaneously if they occurred. At each scalp position, four MEPs were recorded and the peak-to-peak amplitudes were averaged. The scalp sites at which the averaged amplitude was at least 50% of the maximum recorded amplitude for that muscle were considered to represent the excitable area. The motor cortical areas are considered to lie beneath the scalp positions where maximum MEPs could be evoked (27,28). Scalp positions were expressed as a percentage of the distance from the vertex to the auricle, so that the vertex was 0%, the region known under the 10-20 International System as C3 (4) was 40%, the region T3 (4) was 80%, and the auricle itself was 100%.
We examined the cortical motor representation sites for each of the three muscles in 10 normal subjects and in 11 CP patients (5 spastic patients with preterm birth, 4 spastic patients with full-term birth, and 2 athetoid patients). In spastic patients with motor laterality, TMS was applied to the less damaged hemisphere. In diplegic patients without motor laterality and in athetoid patients, TMS was carried out in the hemisphere with the lower resting threshold.
The Wilcoxon signed-rank test was used for examining differences between the cortical motor representation sites in normal subjects and in CP patients. Any difference was considered to be statistically significant if p ≤ 0.05 was found.
RESULTS TOP
Contralateral and ipsilateral motor responses. Transcranial magnetic stimulation could not be performed in two of four athetoid patients during a relaxed state because of their hypertonia. The presence of MEPs in the contralateral and ipsilateral arm and leg muscles are shown in Table 2. Contralateral MEPs were elicited in the relaxed APB and BB in all normal subjects. They were induced in the contracted APB and BB by TMS of each hemisphere in 16 of 17 CP patients. The hemiplegic patient had contralateral MEPs only after TMS of the less damaged hemisphere.
Table 2. Contralateral and ipsilateral motor evoked potentials after transcranial magnetic stimulation in 17 patients with cerebral palsy and in 10 normal subjects
Motor evoked potentials were never induced in the ipsilateral APB and BB, even during muscle contraction, in any normal subject or in any athetoid patient. In the ipsilateral APB, responses could be evoked in two spastic patients with motor laterality (patients 7 and 9) after TMS of the less damaged hemisphere and in one spastic patient without motor laterality (patient 2). Motor evoked potentials of the ipsilateral BB were evoked in three spastic patients with motor laterality (patients 6, 7, and 9) after TMS of the less damaged hemisphere. Three diplegic patients without laterality (patients 2, 3, and 13) also had ipsilateral MEPs. In these patients, the latencies of the ipsilateral MEPs were longer than those of the contralateral MEPs: the mean differences were 3.1 ms (SD = 2.6) in the APB and 2.5 ms (SD = 2.4) in the BB. The amplitudes were smaller in the ipsilateral MEPs than in the contralateral ones: the mean value of the amplitude ratios of the ipsilateral MEPs to the contralateral MEPs was 23.4% (SD = 31.0) in the APB and 40.7% (SD = 66.7) in the BB.
Motor evoked potentials were elicited in the contralateral TA during relaxation in all normal subjects. Motor evoked potentials of the contralateral TA could be evoked in five of the six spastic patients with motor laterality during relaxation and in all six during muscle contraction after TMS of the less damaged hemisphere, but only in one after TMS of the more damaged hemisphere, even during muscle contraction. In seven spastic patients without motor laterality, MEPs were elicited in the contralateral TA after TMS of the left hemisphere in three and the right hemisphere in four during relaxation. All seven patients had contralateral MEPs during contraction of the TA. Contralateral MEPs were evoked during TA contraction in three of the four athetoid patients.
In 10 healthy subjects, MEPs were never induced in the ipsilateral TA during relaxation, but they were elicited in two during muscle contraction (Fig. 1A and Table 2). In these two subjects, the latencies of the ipsilateral MEPs were similar to those of the contralateral ones: the differences were only 1.6 ms and 1.9 ms. The amplitudes of the ipsilateral MEPs were smaller than those of the contralateral ones: the ratios of the amplitude of the ipsilateral MEPs to the amplitude of the contralateral ones were 10.4% and 23.0%. Among the 17 CP patients, ipsilateral MEPs from the TA could be seen in 15. Ipsilateral MEPs after TMS of the less damaged hemisphere were evoked during relaxation in four of six spastic patients with motor laterality (patients 6, 7, 10, and 11) and only during muscle contraction in the other two (patients 8 and 9) (Fig. 1B). In the same patients, ipsilateral MEPs were never induced by TMS of the more damaged hemisphere during relaxation and were induced in only one patient (patient 10) during muscle contraction. Of seven spastic patients without motor laterality, ipsilateral MEPs could be elicited by TMS of the left hemisphere in two (patients 1 and 13) and by TMS of the right hemisphere in three (patients 1, 2, and 13) during relaxation. They could be induced in six patients during muscle contraction in both sides (Fig. 1C and Table 2). In the four athetoid patients, ipsilateral MEPs could be induced from the left hemisphere in three and from the right hemisphere in one during muscle contraction (Fig. 1D). There were no differences in latency between the ipsilateral and the contralateral MEPs in any of the CP patients except one (patient 13, 9.3 and 17.0 ms); the mean difference between the ipsilateral and contralateral MEP latencies was 0.95 ms (SD = 4.2) (23 values from 15 patients). The ratios of the amplitude of the ipsilateral MEPs to the amplitude of the contralateral ones were higher in CP patients than in controls; the mean value was 74.9% (SD = 42.2).
Figure 1. MEPs from the tibialis anterior after TMS in a normal subject (A), in patients with spastic CP (B and C), and athetoid CP (D). (A) Ipsilateral MEPs could be seen only during muscle contraction and had similar latencies and smaller amplitude than the contralateral ones. (B) Bilateral MEPs could be evoked only after TMS of the less damaged left hemisphere in patient 11, who had motor laterality in her arm and leg functions. (C) Ipsilateral MEPs were elicited after TMS of each hemisphere in patient 4 who had no motor laterality. (D) In patient 15, bilateral MEPs were evoked after TMS of the left hemisphere, and only contralateral MEPs could be seen after TMS of the right hemisphere. Ipsilateral MEPs have latency and amplitude similar to contralateral ones in all of these CP patients (B, C, and D). L = left; r = right. Arrows mark the onset of the responses.
Cortical motor representation sites for each muscle. In healthy controls, the cortical representation sites for the APB were on the contralateral hemisphere 5 to 8 cm lateral to the cranial vertex (mean, 34.1%; SD = 5.6%). The sites for the BB were more medial, only 4 to 6 cm lateral to the vertex (mean, 26.3%; SD = 2.5%) (Fig 2). In the 11 CP patients who were included in this part of the study, the cortical representation sites for the APB were 5 to 8 cm lateral to the vertex (mean, 32.4%; SD = 6.6%) and the sites for the BB, 4 to 8 cm (mean, 28.0%; SD = 4.8%). There were no significant differences in the cortical representation sites for the APB and BB between normal subjects and CP patients, or between normal subjects and spastic patients or athetoid patients (Fig. 3).
Figure 2. Functional cortical mapping of the contralateral MEPs from the tibialis anterior (TA), biceps brachii (BB), and abductor pollicis brevis (APB) by TMS at 1-cm intervals between the vertex and auricle in 10 normal subjects and in 11 CP patients. Four trials of TMS were carried out at each scalp site with the subject relaxed, and the amplitudes were averaged. Solid boxes indicate scalp sites where maximum MEPs were induced. Dotted areas indicate scalp sites where MEPs with an amplitude above 50% of the maximum were evoked. For the TA, a marked lateral shift occurred among patients 1, 2, 6, 7, and 9, all of whom were spastic patients who had been born prematurely.
Figure 3. Cortical motor representation sites (mean ± SD) of maximum responses from the contralateral tibialis anterior, biceps brachii, and abductor pollicis brevis in controls (n = 10) and in CP patients (n = 11). The cortical site is expressed as the percentage of the distance between the vertex and the auricle. The cortical motor representation site for the TA was shifted laterally in CP patients, especially spastic patients with preterm birth, compared with controls. Conversely, in the BB and APB, there were no significant differences in the cortical site of maximum responses between controls and CP patients. *p ≤0.05; **p ≤0.005
Figure 4 shows typical examples of MEP maps of the TA in one normal subject, three diplegic patients, and one athetoid patient. Cortical motor representation sites for the TA were medial to those of the arm muscles and lay within 4 cm of the vertex in normal subjects (mean, 14.7%; SD = 4.0%) (Figs. 2, 3, and 4A). In five spastic patients born prematurely (patients 1, 2, 6, 7, and 9), the most excitable cortical sites for the TA were lateral to those obtained in normal subjects, which ranged from 5 to 7 cm (mean, 29.5%; SD = 3.5%) (Fig. 4B). The sites for the TA were within the sites for the APB and BB (Fig. 3). One of four spastic patients with full-term birth (patient 10) had two peaks in MEP amplitude (Figs. 2 and 4C). The other three of four spastic patients with full-term birth (patients 11, 12, and 13) and two athetoid patients (patients 14 and 17) had cortical motor representation sites for the TA that were similar to those of the normal subjects (Figs. 2, 3, and 4D and 4E).
Figure 4. Contralateral MEPs of the tibialis anterior in a normal subject (A), a diplegic patient with preterm birth (B), two diplegic patients with full-term birth (C and D), and an athetoid patient (E). The locations of the superimposed MEPs correspond to the scalp positions where they were evoked. (A) The maximum MEPs were induced at the scalp 3 cm lateral to the vertex. (B) In patient 1, the maximum MEPs were evoked at a site 5 cm lateral to the vertex. (C) Patient 10 had two peaks in amplitude; a medial peak at 2 cm and a lateral peak at 6 cm. (D) Patient 13 had maximum MEPs at a site 2 cm. (E) In patient 17, the maximum MEPs were evoked at a site 2 cm.
Motor evoked potentials were found in the ipsilateral TA in 2 of 10 healthy subjects while they were relaxed (Fig. 5). When ipsilateral MEPs induced, the coil was 1 or 2 cm lateral to the vertex. The latencies of the ipsilateral MEPs were similar to those of the contralateral MEPs. Ipsilateral MEPs were evoked in seven spastic patients: four of them were preterm birth patients (patients 1, 2, 6, and 7) and three were full-term birth patients (patients 10, 11, and 13). The threshold intensity for the ipsilateral TA was the same as for the contralateral TA in six of these patients; only one patient had a higher threshold for the ipsilateral TA than the contralateral one. Ipsilateral MEPs were elicited by TMS at lateral scalp sites along the vertex-auricle line in all patients with preterm birth and in one patient with full-term birth (patient 13) (Fig. 6). The scalp sites eliciting the highest ipsilateral MEPs were 5 to 7 cm lateral to the vertex. In these patients, TMS near the vertex did not induce any ipsilateral MEPs (Fig. 6). In two patients with full-term birth (patients 10 and 11), ipsilateral MEPs were present diffusely, between the vertex and 9 cm laterally, peaking at 1 or 2 cm lateral to the vertex. The most excitable cortical sites for the ipsilateral TA corresponded to those for the contralateral TA in three of seven patients (patients 1, 6, and 11) (Fig. 6A). The highest ipsilateral MEPs were elicited with the coil 1 cm away from the highest cortical MEP sites for the contralateral TA in two patients (patients 2 and 7), 4 cm apart in 1 (patient 10), and 5 cm apart in 1 (patient 13) (Fig. 6B).
Figure 5. Contralateral and ipsilateral MEPs from the tibialis anterior elicited by TMS on the left of the vertex in a normal subject. The locations of the superimposed MEPs correspond to the scalp positions where they were evoked. TMS was performed while muscles were relaxed. Both ipsilateral and contralateral MEPs were present at the sites 1 and 2 cm lateral to the vertex, but only contralateral MEPs were induced at the more lateral sites.
Figure 6. Ipsilateral MEPs of the tibialis anterior in two diplegic patients, one with preterm birth (A) and one with full-term birth (B). The locations of the superimposed MEPs correspond to the scalp positions where they were evoked. The maximum MEPs were induced at a site 5 cm in patient 1 (A) and at a site 7 cm in patient 13 (B). The solid circles mark the location of the maximum contralateral MEPs in each patient. (see Figs. 4B and 4D).
DISCUSSION TOP
Ipsilateral MEPs of the TA were found in most of the CP patients. For five reasons, we believe that these were true readings, not artifacts from inadvertently stimulating the wrong brain hemisphere.
1) The coil was placed 4 to 5 cm lateral to the vertex, a position that in our experience and that of other investigators (13,17-20) is not likely to result in stimulation of the opposite hemisphere.
2) In four of the six spastic patients with motor laterality, ipsilateral MEPs were evoked by stimulation of the less damaged hemisphere during relaxation, but contralateral responses could not be induced from the more damaged hemisphere in any of them.
3) Cortical mapping in five of the seven spastic patients showed that the sites eliciting the strongest MEPs in the ipsilateral TA were at least 5 cm from the vertex. In these patients, ipsilateral MEPs were not evoked with the coil placed near the vertex.
4) One patient (patient 13) had ipsilateral MEP latencies that were moderately longer than the contralateral ones, suggesting that the ipsilateral pathway was truly different from the contralateral one.
5) The TMS stimulus would require about 10 ms to pass through the corpus callosum to the opposite hemisphere (29), but in almost all of the subjects (both normal and CP patients), the ipsilateral MEP latencies were only a few milliseconds, at most, longer than the contralateral ones.
Motor-evoked potentials from the ipsilateral TA were common among both the spastic and the athetoid patients and both the preterm birth and full-term birth patients. In all spastic CP patients with motor laterality, they could be evoked by TMS of the less damaged hemisphere. Ipsilateral MEP amplitudes were larger in CP patients than in normal subjects. Therefore, we suggest that ipsilateral motor pathways might develop after early brain lesions and that in patients with bilateral lesions, these alternate pathways might develop preferentially in the less damaged hemisphere (Fig. 7B-7E). When the brain damage is slight in both hemispheres, ipsilateral motor pathways might be facilitated in both hemispheres (Fig. 7F-7I). Motor evoked potentials were induced more frequently in the ipsilateral TA than in the ipsilateral APB and BB in both CP patients and normal subjects, suggesting that ipsilateral motor pathways may be more abundant in the legs than in the arms.
Figure 7. Possible mechanisms underlying ipsilateral MEPs in the tibialis anterior after bilateral cerebral lesions. (A) In normal humans, proximal limb muscles are believed to receive inputs via large crossed and small uncrossed corticospinal fibers and via bilateral cortico-reticulospinal fibers (preferentially contralateral). (B-E) In patients with markedly asymmetrical brain damage, ipsilateral motor pathways may be facilitated in the less damaged hemisphere: ipsilateral corticospinal pathways (B), cortico-reticulospinal pathways (C), axon collaterals (D), or persisting fetal pathways (E). (F-I) In patients with mild cerebral lesions in both hemispheres, ipsilateral motor pathways may be facilitated on both sides: ipsilateral corticospinal pathways (F), cortico-reticulospinal pathways (G), axon collaterals (H), or persisting fetal pathways (I).
The mechanisms of the formation of an ipsilateral motor pathway are proposed as follows: that normally occurring ipsilateral projections (Fig. 7A), such as 1) corticospinal tracts (Figs. 7B and 7F) and 2) cortico-reticulospinal tracts (Figs. 7C and 7G), establish more extensive and enhanced connections; 3) that ipsilateral projections originate from an abnormal branching of the contralateral corticospinal axons (Figs. 7D and 7H); and 4) that transient fetal connections persist without normal regression (Figs. 7E and 7I). Motor evoked potentials in the ipsilateral TA with latencies similar to the contralateral MEPs may indicate fast-conducting ipsilateral CST rather than polysynaptic pathways (Fig. 7B and 7F) (11). Benecke et al. (11) reported that ipsilateral MEPs, including those in the arm and leg muscles, were evoked by TMS and had amplitudes and latencies similar to the contralateral MEPs in hemiplegic patients with early brain lesions, but that ipsilateral MEPs had considerably smaller amplitudes and delayed latencies in patients with later brain lesions. They suggested that the ipsilateral corticospinal pathway might be reinforced after early brain lesions and that the cortico-reticulospinal pathway might be facilitated after later brain lesions. In kittens, the terminal fields of motor cortical neurons are widely and bilaterally distributed in the spinal cord in the early postnatal days, but in adulthood, the majority are localized to the contralateral side (30). Early brain lesions may prevent such normal elimination of the ipsilateral motor pathways (30,31).
Another normally occurring ipsilateral pathways via the brain-stem, cortico-reticulospinal tracts, might be facilitated (Fig. 7C and 7G). MEP latencies through such pathways may be prolonged (11,14,16,18,19). The only spastic patient with full-term birth (patient 13) had ipsilateral MEPs with latencies that were moderately longer than the contralateral ones. Such cortico-reticulospinal pathways or slow-conducting ipsilateral corticospinal pathways might be facilitated in this patient. Because there were some differences in latency and amplitude between the ipsilateral and contralateral MEPs from the APB and the BB, such pathways might be also facilitated in the arms.
A third possibility is that ipsilateral MEPs might originate from abnormal branching of the contralateral corticospinal axons (Fig. 7D and 7H) (1,2). Bilateral hand motor responses with similar latencies and amplitudes after TMS of the intact hemisphere are elicited in hemiplegic patients with congenital or early postnatal lesions (11-14). Using TMS mapping in hemiplegics with congenital and early postnatal lesions, we recently found that the cortical motor representation sites for the paretic hand were close to or at the same scalp sites as the intact hand (14,32). In our study, cortical motor representation sites for the ipsilateral TA were the same as or in the vicinity of those for the contralateral TA in five of seven patients. Axon collaterals are one possible mechanism for the observed ipsilateral MEPs.
The most striking finding was a lateral shift of the cortical motor representation sites for the TA in spastic patients born prematurely. This is consistent with the bilateral hypoperfusion in the superior motor cortex observed in diplegic CP patients on single photon emission computed tomography (33). To our knowledge, this is the first observation in humans of a lateral shift of the cortical motor representation area after a cerebral lesion.
One possible mechanism is the persistence of transient fetal connections without normal elimination (Figs. 7E and 7I). In the undeveloped rat brain, many cortical neurons, including those in the occipital cortex, have corticospinal projections of transient axon collaterals and such collateral projections are eliminated during development (34-36). Also, abundant corticospinal axons of the neonatal rat brain decrease rapidly after early postnatal days (36,37). Contralateral hemisphere hypertrophy caused by a unilateral neonatal brain lesion has been reported in adult rats (31,38). Therefore, the CST axons from the cortical neurons other than the motor cortex for the legs would persist without normal elimination and innervate the spinal motoneurons because of the PVL, which is believed to affect descending fibers for the legs more than the laterally placed fibers for the arms (39). Interestingly, patient 10, a diplegic patient who was born at full-term, had two peaks in MEP amplitude on the vertex-auricle line. She might have had brain lesions during the fetal period because she had only slight neonatal asphyxia at birth. If CST damage to the legs was partial, both a medial cortical peak (the usual cortical representation site) and a compensatory lateral peak might have developed (Figs. 2 and 4C).
Alternately, the shift in the motor representation site may have resulted from unmasking preexisting synaptic connections. Corticocortical connections are abundant within the motor cortex and between the motor cortex and the other cortical areas (40,41). Enlargement of the motor representation area has been reported after limb amputation (42,43), spinal cord injury (44), and facial palsy (45). In a positron emission tomography study of patients with cerebral lesions, Weiller et al. (46) reported a large extension of the hand field in the damaged hemisphere after capsular infarction. Adjacent areas of the motor cortex may be activated after disruption of the efferent pyramidal pathway or the afferent pathway (deafferentation) by deprivation of intercortical inhibition (47). In our study, the excitable areas for the TA were not enlarged, but they moved laterally in five spastic patients with preterm birth. In addition, they did not include the usual cortical representation areas for the TA in four of these patients (Figs. 2 and 4B). Therefore, it seems unlikely that this phenomenon results from unmasking preexisting connections.
No lateral deviation of the motor cortical site for the TA was observed in three of four diplegic patients with full-term birth, indicating that there may be critical windows for cortical reorganization. Also, two athetoid patients, one born prematurely and the other at full-term, did not show the lateral deviation of the motor cortical area. Such cortical reorganization may be associated only with CST lesions.
We cannot definitively confirm the mechanisms of the cortical reorganization based on these limited data. One limitation is the small number of patients studied. A second limitation is that the precise timing of the occurrence of the brain lesions is unknown. Further studies are necessary to confirm the mechanisms of the cortical plasticity after bilateral cerebral lesions both in humans and animals.
In conclusion, our study shows evidence of two types of plasticity after early bilateral cerebral lesions: facilitation of ipsilateral motor pathways and lateral deviation of the motor cortical area for the legs. Ipsilateral motor pathways were reinforced both in spastic and athetoid CP patients, with preterm and full-term births. The cortical motor representation area for the TA moved laterally toward the area for the arm muscles in spastic patients with preterm birth, but not in spastic patients with full-term birth and athetoid patients.