15. Hohlfeld R, Conti-Tronconi B, Kalies I, Bertrams J , Toyka KV. Genetic restriction of autoreactive acetylcholine receptor-specific T l y m p h o c y t e s i n m y a s t h e n i a g- r a v i s . J I m m u n o l 1985;135:23932399. 16. Whiting P. Vincent A. Newsom-Davis J. Monoclonal antibodies to TorGdo acetylcholine receptor. Characterisation of antigenic determinants within the cholinergic binding site. Eur J Biochem 1985;150:533-539. 17. Beeson D, Brydson M, Wood H,Vincent A. Newsom-Davis J. Human muscle acetylcholine receptor: cloning and expression in Escherichia coli of cDNA for the alpha-subunit. Biochem SOC Trans 1989;17:219-220. 18. Beeson D, Moms A, Vincent A, Newsom-Davis J. The human muscle nicotinic acetylcholine receptor alpha-subunit exists as two isoforms: a novel exon. EMBO J 1990;92101-2106. 19.Sommer N, Willcox N, Harcourt GC, Newsom-Davis J. Myasthenic thymus and thymoma are selectively enriched in AChRreactive T cells. Ann Neurol 1990;28:312-319. 20. Noda M, Furutani Y, Takahashi H,et al. Cloning and sequence analysis of calf cDNA and human genomic DNA encoding alphasubunit precursor of muscle acetylcholine receptor. Nature 1983;305:818-824. 21. Melms A, Chrestel S, Schalke BCG, et al. Autoimmune T lymphocytes in myasthenia gravis. Determination of target epitopes using T lines and recombinant products of the mouse nicotinic acetylcholine receptor gene. J Clin Invest 1989;83:785-790.
22. Jermy AC, Fisher CA, Vincent AC, Willcox NA, Newsom-Davis J. Experimental autoimmune myasthenia gravis induced in mice without adjuvant: genetic susceptibility and adoptive transfer of weakness. J Autoimmun 1989;2:675-688. 23. Scadding GK, Calder L, Vincent A, Prior C, Wray D, NewsomDavis J. Anti-acetylcholine receptor antibodies induced in mice by syngeneic receptor without adjuvants. Immunology 1986;58:151-155. 24. Markmann J , Lo D, Naji A, Palmiter RD, Brinster RL, HeberKatz E. Antigen presenting function of class I1 MHC expressing pancreatic beta cells. Nature 1988;336:476-479. 25. Davies TF. Cocultures of human thyroid monolayer cells and autologous T cells: impact of HLA class I1 antigen expression. J Clin Endocrinol Metab 1985;61:418-422. 26. Pette M, Fujita K, Kitze B, et al. Cytotoxic CD4+ human MBP specific T-lymphocyte lines isolated from MS patients and healthy donors [Abstract]. Neurology 1989;39(suppl 1):173. 27. Ota K, Matsui M, Milford EL, Mackin GA, Weiner HL, Hafler DA. T-cell recognition of an immunodominant myelin basic protein epitope in multiple sclerosis. Nature 1990;346:183-187. 28. Lennon VA, McCormick DJ, Lambert EH, Griesmann GE, Atassi MZ. Region of peptide 125-147 of acetylcholine receptor alpha subunit is exposed at neuromuscular junction and induces experimental autoimmune mysthenia gravis, T cell immunity, and modu l a t i n g a u t o a n t i b o d i e s . P r o c N a t l Acad S c i USA 1985:82:8805-8809.
Reorganization of corticospinal pathways following spinal cord injury Helge Topka, MD; Leonardo G. Cohen, MD; Reginald A. Cole, MD; and Mark Hallett, MD
Article abstract-To
assess changes in the relationship between cortical motor representation areas and their target muscles following spinal cord lesions, we studied motor evoked potentials (MEPs) to transcranial magnetic stimulation in six patients with complete spinal cord injuries at low thoracic levels and eight healthy subjects. Magnetic stimulation a t rest activated a larger fraction of the motoneuron pool and evoked MEPs with shorter latencies from a larger number of scalp positions in muscles immediately rostral to the level of a spinal cord injury than in corresponding muscles in controls. The MEPs associated with maximal voluntary activation were not significantly different in the two groups. These results suggest enhanced excitability of motor pathways targeting muscles rostral to the level of a spinal transection, reflecting reorganization of motor pathways either within cortical motor representation areas or at the level of the spinal cord. The data do not allow the determination of the contribution of spinal or cortical mechanisms. However, they support the notion of a limited flexible relationship between primary motor cortex and its target muscles following alterations of normal input-output patterns. NEUROLOGY 1991;41:1276-1283
Chronic deafferentation induces reorganization within cortical somato~ensory~-~ and motoPp6 representation maps in mammals. Although there may be different patterns of reorganization depending on the developmental stage of the animal, the ability of the CNS to reorganize following peripheral or central lesions ap-
pears preserved to some extent in mature animal^.^^^ Recent studies in adult humans using cortical magnetic stimulation techniques and somatosensory evoked potentials suggest that the capacity to reorganize motor and somatosensory pathways following peripheraP9 or central'0-12lesions is partially preserved in humans.
From Human Cortical Physiology Unit, Human Motor Control Section, Medical Neurology Branch, National Institute of Neurological Disorders and Stroke, National Institutes of Health, Bethesda, MD.
Dr. Topka is supported by Deutsche Forschungsgemeinschaft,Federal Republic of Germany.
Presented in part at the 42nd annual meeting of the American Academy of Neurology, Miami Beach, FL, April 1990 and at the 1st International Congress on Movement Disorders, Washington, DC, April 1990. Received August 17, 1990. Accepted for publication in final form January 11,1991. Address correspondence and reprint requests to Dr. Leonardo Cohen, Building 10, Room 5N226, NINDS, NIH, Bethesda, MD 20892.
1278 NEUROLOGY 41 August 1991
Table 1. Patient characteristics Age at
Age
injury Nature of injury+
Sensory impairment?
Motor impairment
Pt
(yrs)
Sex
1
39
M
Compression fracture T-11
34
Below T- 11
Paraplegic;no lower extremity spasms
2
37
F
Compression fracture T-12
33
Below T-12
Paraplegic;no lower extremity spasms
3
21
M
Compression fracture T-10
16
Below T- 11
Paraplegic; moderate lower extremity spasms
4
40
M
Burst fracture T-11
38
Below T-11
Paraplegic;no lower extremity spasms
5
43
F
Compression fracture T-12
23
Below T-12
Paraplegic;moderate lower extremity spasms
6
47
F
Compression fracture T9-10
42
Below T-10
Paraplegic;moderate lower extremity spasms
(yrs)
* Initial surgical treatment included implantation of Harrington rods. t Total loss of sensation below the level indicated.
Table 2. Motor evoked potentials
A
B
Muscle segment
Controls 8)
Patients (n = 6)
P’
MEPamplitude (PV)
Rostralt Lower
241 t 51 351 t 78
376 f 82 719 k 298
0.1213 0.4386
MEPamplitude (% MNPool)
Rostralt Lower
13.0 ? 3.9 10.7 f 5.0
29.9 t 3.0 31.0 f 11.2
0.0933
MEP threshold (%output)
Rostralt Lower
71.4 f 4.9 71.4 f 4.3
65.8 3.8 64.2 f 11.6
(n
=
*
0.0098 0.5168 0.2478
Values are expressed as mean f standard error.
* Mann-Whitney U test.
Figure 1. (A and B) Schema of transcranial magnetic stimulation and EMG recording. (A) Position of recording electrodes over abdominal wall muscles. Surface electrodes (shaded rectangles) were taped to the skin overlying the muscular portion of rectus abdominis (upper electrode) and external and internal oblique muscles (middle and lower electrodes). Open rectangles are reference electrodes. (B) Top of the subject’s head. The grid is centered on Cz (international 10-20 electrode system). Gridpoints are I cm apart and represent the scalp positions stimulated with the eight-shaped coil. Approximate position of the central sulcus in relation to the coil position is indicated by the thin curved line.
This report describes the reorganization of h u m a n motor pathways following long-term thoracic spinal cord lesions and discusses some of the mechanisms possibly involved.
*
Methods. We studied six patients (mean age, 37 8 years) who had severe traumatic injuries of the spinal cord at low thoracic levels occurring 2 to 20 years previously. We also
t Mean response of muscles rostral to the lesion (upper right, middle right, and middle left).
studied eight normal controls (mean age, 35 -t 14 years). All subjects gave their written informed consent for participation in the study, which was approved by the institute’s clinical investigational review committee. All patients had a complete neurologic examination. Clinical findings were verified, including intraoperative and radiologic findings, with respect to the sensory and motor level of the spinal cord lesion. The patients’ clinical features and relevant medical history are summarized in table 1. None of the patients had severe head trauma or loss of consciousness during the accident. Experimental procedures. Transcranial magnetic stimulation was delivered to different scalp positions using a commercially available magnetic stimulator (Cadwell MES-lo), which generates a multiphasic magnetic stimulus with a peak magnetic flux density of 2.2 tesla in the center of a circular coil (focal-point coil) with an inner diameter of 52 mm and an outer diameter of 98 mm (seven turns of wire). EMG was recorded bilaterally using surface electrodes (DISA 13K60) placed on the skin overlying the abdominal wall muscles. August 1991 NEUROLOGY 41 1277
1000
90.
. .
0
0 00 0
.
80
-
70
-
II.
60-
I
rn
SO -
I SCI PATIENTS (n=6)
CONTROLS (n=8)
CONTROLS
SCI PATIENTS
A
B
SCI PATIENTS
CONTROLS
C
Figure 2. (A through C) Motor evoked potentials (MEPs) from all recording sites i n spinal cord injury (SCZ) patients (n = 6) and controls ( n = 8). ( A ) Amplitudes of M E P s to transcranial magnetic stimulation (absolute amplitudes i n pV). (B) Amplitudes of M E P s expressed as a percentage of the largest M response obtained by stimulation of the supplying ventral roots. (C) Stimulus intensity (in percent output of stimulating unit) required to yield M E P s i n at least three of five trials at a gain of 50 pV. M E P s recorded f r o m lower segments of abdominal wall muscles i n the patients are indicated by open circles. Recording sites in each subject were defined choosing three levels on a line connecting the xyphoid and the right anterior superior iliac spine and one on the left (figure 1A). Given the electrode positions, it is most likely that the uppermost electrodes were recording EMG activity from rectus abdominis and external oblique muscles, whereas the electrodes in the middle and lower positions were recording predominantly from the external oblique with some contribution from the internal oblique. Abdominal wall muscles are supplied by spinal segments between T-6 and T-12.I3 EMG activity was amplified using a bandpass filter of 100 Hz to 2 kHz and recorded with a gain of 50 to 200 pV/div using a conventional EMG device (DISA 1500). The subjects lay supine on a bed in a quiet room during the experiments. Cortical magnetic stimuli were delivered a t rest approximately a t the end of the expiratory phase of a breathing cycle. Baseline and prestimulus muscle activity was monitored by a loudspeaker connected to the recording unit. Subjects were considered to be a t rest when no spontaneous EMG activity was detected in the monitored muscle groups with the recording unit set to a gain of 50 pV/div. Traces were printed during the experiment for later off-line analysis. Amplitudes of 10 to 15 motor evoked potentials (MEPs) to transcranial magnetic stimulation delivered a t maximal stimulus intensity (100%) were measured peak-to-peak and expressed as a percentage of the largest M response obtained by direct stimulation of the ventral roots over the spine (T-5 to T-12).This was done to estimate the percentage of the alpha motoneuron pool activity (% MNP) recruited by each cortical stimulation. Ventral roots were stimulated electrically or magnetically. Supramaximal electrical stimulation was delivered to the spine at T-3 to T-6, T-6 to T-9, and T-9 to T-12 (cathode-anode,respectively) using a Digitimer D180 stimulation device. Electrical stimuli were applied with increasing stimulus intensity until no further increase in the M response of abdominal wall muscles was observed. Magnetic stimulation was delivered to various spinal positions between T-5 and T-12 by a circular coil at maximal stimulus intensity. Electrical or magnetic stimulation over the spine is likely to excite 1278 NEUROLOGY 41 August 1991
ventral roots a t their exit from the spinal ~ a n a l . ' ~To J ~avoid possible displacements of metallic implants in the patients, electrical stimulation generally was used. Motor thresholds for cortical MEPs were determined by delivering transcranial magnetic stimuli from a circular coil at various stimulus intensities with the center of the coil positioned over Cz. Moving the coil 2 cm in any direction from Cz did not change the threshold for M E P activation. The MEPs were amplified and recorded a t a gain of 50 pV/div, and threshold was separately defined for each muscle level as the lowest stimulus intensity that evoked MEPs in at least three of five trials. This intensity was found by reducing or increasing the stimulus intensity in 5-percent steps. The effect of voluntary activation of target muscles on amplitude and latency of MEPs was then assessed. Cortical stimuli were delivered by a circular coil a t maximal stimulus intensity with the center of the coil positioned over Cz during rest or maximal voluntary activation of target muscles. Muscle activity was monitored with a loudspeaker connected to the recording unit but not quantified. In abdominal wall muscles, shortest latencies are obtained with background activation of 10% and largest amplitudes with 60% of maximal voluntary contraction.'6 The cortical representation areas of abdominal muscles were mapped by delivering focal magnetic stimulation to scalp positions 1 cm apart with an eight-shaped coil (Cadwell Corticoil). Technical aspects of this coil design have been discussed elsewhere.I7The induced electric field reaches its peak in the area under the junction of both wings of the coil, which was positioned over the target locations. Two to three stimuli were delivered to each scalp position with the subject at rest. As before, MEP amplitudes were averaged and expressed relative to the largest peripheral M response obtained by stimulation of the ventral roots. Maps of the cortical motor representations were obtained by calculating the mean amplitude of MEPs evoked by magnetic stimulation for each scalp position. Statistical analysis. All results are expressed as mean standard error. The Mann-Whitney U test and the Wilcoxon signed rank test were used for statistical comparisons.
*
CONTROLS
PATIENTS
I *
*
500 pV
L 10 msec
Results. In controls, amplitudes of peripheral M responses recorded from the upper and middle segments of the abdominal wall muscles were typically maximal when the ventral roots in the vicinity of vertebrae T-6 to T-9 were stimulated, and responses in the lower segments were maximal with stimulation over T-9 to T-12. Amplitudes of peripheral M responses recorded from the lower segment, but not from the upper and middle segments, were significantly reduced in the patients (2,087 f 161pV)compared with normal subjects (5,631 f 1,340 pV; p < 0.02), indicating the involvement of low thoracic spinal segments in the spinal injury. Consequently, only M E P recordings obtained from the upper and middle segments of the abdominal wall muscles were considered to represent preserved spinal segments rostra1 to the level of the lesion, and recordings from the lower segments were analyzed separately. Cortical stimuli of equal intensity evoked proportionately larger MEPs in muscles proximal to the level of the lesion (upper and middle segments) in spinal cord injury patients (29.9 3.0 % MNP) than in corre-
Figure 3. M responses to electrical or magnetic stimulation over the spine (a) and MEPs to transcranial magnetic stimulation (b) in five patients and five controls. Recordings shown were made f r o m the middle segment of the abdominal muscles on the right side of the body.
sponding muscles in normal subjects (13.0 f 3.9 ?6 MNP; p = 0.0098) (table 2; figures 2 and 3). Magnetic stimulation delivered at rest evoked MEPs in preserved muscle segments at shorter latencies in patients (16.4 f 0.71 msec) than in controls (18.6 & 0.55 msec; p = 0.039). Proportional amplitudes of MEPs recorded from the lowest muscle segment were more variable from patient to patient and were not consistently larger in the patient group (31.0 -+ 11.2 ?6 M N P in the patients, 10.7 f 5.0 in controls; p = 0.093) (table 2; figures 2 and 3). The muscle level from which the largest relative M E P could be evoked differed from patient to patient with respect to the level of the spinal lesion. With injury a t T-10 (n = 2), MEPs were largest in the middle segment of the abdominal wall muscles, and with injury a t T-11 (n = 2), they were largest in the lower segment. With injury at T-12 (n = 2), there was no clear preferential activation of one muscle segment. MEPs could be evoked at slightly smaller stimulus intensities in patients (65.8 & 3.8% output) than in controls (71.4 & 4.9% output), but the observed difAugust 1991 NEUROLOGY 41 1279
LEFT HEMISPHERE
I
RIGHT HEMISPHERE\
Figure 4 . Map of cortical representation of abdominal wall muscles in a representative SCI patient (patient 4). Recordings were made from the upper segment of the abdominal wall muscles on the right side of the body. Three EMG recordings are superimposed for each scalp position.
ference was not statistically significant for either muscle segment (table 2). In the controls, MEP amplitudes were larger in subjects with low MEP thresholds (r2 = 0.704, p 5 0.0001; linear regression), whereas the patients’ amplitudes correlated only weakly with motor thresholds (r2 = 0.259, p 5 0.031; linear regression). Patients tended to have large MEPs in muscles regardless of the motor threshold. Effects of voluntary activation of target muscles on MEP amplitude and latency. In both groups, voluntary activation of abdominal wall muscles led to a decrease in latency and an increase in amplitude of MEPs recorded from muscle segments proximal to the level of the spinal injury (p = 0.0022; Wilcoxon signed rank test). Mean latency shift was 2.1 0.68 msec in the controls and 1.6 -+ 0.47 msec in the patients. Voluntary activation facilitated the amplitude of cortical MEPs by 34.3 -+ 3.6 75 MNP in controls and by 23.8 -t 6.7 5% MNP in patients. The maximal amplitude obtained during voluntary activation of the target muscles, however, was not significantly different between groups. Effects of voluntary activation on amplitude facilitation tended to be smaller for MEPs recorded from the lower muscle segment in patients (11.95 r 4.23 % MNP) than in controls, but the difference was not significant (p = 0.297). 1280 NEUROLOGY 41 August 1991
Mapping of cortical representation areas. MEPs recorded from all segments of the abdominal wall musculature were evoked from a larger number of scalp positions in the patients (24.8 k 7.5 positions) than in the controls (5.2 ? 2.3 positions; p = 0.02).In both patients and controls, amplitudes of MEPs evoked by focal magnetic stimulation were slightly more variable than those evoked by stimulation with the circular coil. Motor representation areas in patients were expanded in both the anteroposterior axis and the mediolateral axis as compared with controls (figures 4 and 5, A and
B).
In two controls, no MEPs could be obtained by focal magnetic stimulation. In five patients and four controls, unilateral scalp stimulation evoked bilateral muscle responses. In two other controls, MEPs were elicited only with stimulation of the contralateral hemisphere. The largest cortical representation area of abdominal wall muscles determined in a control is shown in figure 5, B (subject 11).
Discussion. Transcranial magnetic stimulation delivered a t rest evokes proportionately larger MEPs from a larger number of scalp positions in muscles rostra1 to the level of the lesion in patients with spinal cord injury
A
UP
%MNP
mid R
R
25 15
cz
15
10
I
10
Ant.
Hemisphere
Post.
%MNP
25
20
B
Left
F
15
% M N P 25
UP
1
R
SUBJECT1
I
15 10
25
10
..
Ant.
Post.
%MNP
25
20 l5
mid R
5
5 Left
I
Ft Post.
Hemisphere
I
SUBJECT11
I
f:i 15
Left
Figure 5. (A and B) Size of cortical motor representation areas targeting abdominal muscles in two representative patients (A) and in two controls (B). M E P s for each scalp position were averaged and expressed as a percentage of the peripheral motoneuron pool activity. T h e X , Y, Z coordinates show the coil position i n the coronal axis ( X ) and i n the anteriorposterior axis ( Y ) and the recruitable fraction of the motoneuron pool activity (% M N P ) (2).U p R = upper segment of the abdominal muscles on the right side of the body; mid R = middle segment of the abdominal muscles o n the right. Note presence of ipsilateral responses in controls and patients. August 1991 NEUROLOGY 41 1281
than in corresponding muscles in normal subjects. Voluntary activation of target muscles facilitates MEPs in both groups and leads to maximal MEP amplitudes in spinal cord injury patients that are not different from those of normal subjects. These results are in accord with an earlier studylo that reported that magnetic stimulation a t rest elicited larger MEPs from a larger number of scalp positions in partially preserved muscles in a patient with complete spinal cord injury a t the C-5 level and in another patient who had partially recovered from a (Cl-C2) spinal cord injury as compared with three normal subjects. The authors inferred that the motor cortex projection system had been reorganized. Larger amplitudes, shorter latencies of magnetically evoked MEPs, and a n enlarged map of motor output a t rest reflect enhanced excitability of motor pathways targeting muscles rostral to the level of a spinal transection. While reorganization of cortical motor representation areas following a complete spinal cord injury is conceivable, it does not exclude rearrangement on the level of the spinal motoneuron pool, or a combination of both. In principle, changes in the excitability of descending pathways that occur on the level of the motor cortex or on the level of the spinal cord could have similar effects on MEPs t o transcranial magnetic stimulation. Thus, the paradigm used does not allow distinguishing the relative contribution of either one. Spinal segments immediately rostral to the lesion might have been reorganized, leading to increased excitability of the spinal alpha motoneuron pool at rest and requiring fewer and less frequent descending volleys to discharge. Sprouting of axon collaterals in spinal segments immediately rostral t o the level of a transection and the subsequent formation of new synapses over time occur in newbornla and adult rats.lg Regenerative sprouting of corticospinal axons rostral to the level of complete spinal cord injuries also occurs in humans.zo Alternatively, enhanced excitability of descending pathways to transcranial magnetic stimulation could reflect reorganization within cortical motor representation areas. Transcranial magnetic stimulation, as utilized in our study, is likely to activate corticospinal neurons transsynaptically rather than exciting directly descending axonsZ1and may activate a larger number of cortical interneurons that synapse onto the same corticospinal neuron following a spinal transection. Corticospinal neurons that previously had subthreshold effects on motoneurons or spinal interneurons might now have become capable of discharging motoneurons utilizing collaterals that connect to the motoneuron pools of several muscles in multiple spinal segments.22*23 Two different mechanisms are conceivable. First, the number of active, excitatory, intracortical interneurons that synapse onto preserved, intact corticospinal neurons might increase over time (sprouting). Second, the strength of existing, but previously inactive, intracortical or corticospinal connections might have been modified in response to alteration of the usual sensorimotor input-output pattern following the spinal cord transection (unmasking). Unlike the sprouting of new dendrites, the unmasking of preexisting synapses 1282 NEUROLOGY 41 August 1991
would not require structural changes and could account for the shifts in cortical motor representation areas that have been observed within hours following peripheral nerve lesions in adult rats.5 In addition to changes within the corticospinal projection system itself, it is possible that the gain-setting of descending brainstem projections organized parallel to corticospinal pathways has changed following the spinal transection. Motor facilitation under the influence of direct monosynaptic connections between caudal areas of the reticular formation and ventral horn cells occurs in mammal^.^^,^^ However, the existence of a similar parallel control system has not been established in humans. T h e techniques applied in our study do not allow the determination of where adaptive modifications of the descending projection system occur and which specific mechanisms are involved. However, the data imply that a flexible functional relationship between motor cortex and target muscles, suggested earlier,26.27 exists in adult humans.
Acknowledgments We thank J. Rothwell and Lisa McShane for helpful comments and B.J. Hessie for editorial assistance.
References 1. Franck J I . Functional reorganization of cat somatic sensorymotor cortex after selective dorsal root rhizotomy. Brain Res 1980;186:458-462. 2. Kalaska JF, Pomeranz B. Chronic paw denervation causes an agedependent appearance of novel responses from forearm in ‘paw cortex’ in kittens and adult cats. J Neurophysiol 1979;42:618-633. 3. Kelahan AM, Ray RH, Doetsch GS. Time dependent changes in the organization of somatosensory cerebral cortex following digit amputation in adult racoons. Somatosens Res 1984;2:49-81. 4. Merzenich MM, Kaas J H , Wall J T , Sur M, Nelson FLJ,Felleman DJ. Progression of change following median nerve section in the cortical representation of the hand areas 3b and 1 in adult owl and squirrel monkeys. Neuroscience 1983;10:639-665. 5. Sanes J N , Suner S, Lando J F , Donoghue J P . Rapid reorganization of adult rat motor cortex somatic representation patterns a f t e r m o t o r n e r v e i n j u r y . P r o c N a t l Acad S c i USA 1988;85:2003-2007. 6.Donoghue J P , Sanes J N . Organization of adult motor cortex representation patterns following neonatal forelimb nerve injury in rats. J Neurosci 1988;8:934-943. 7. McKinley PA, Jenkins WM, Smith JL. Merzenich MM. Agedependent capacity for somatosensory cortex reorganization in chronic spinal cats. Brain Res 1987;428:136-139. 8. Cohen LG, Bandinelli S, Findley TW, Hallett M. Motor reorganization after upper limb amputation in humans: a study with focal magnetic stimulation. Brain 1991;114:615-627. 9. Sica REP, Sanz OP, Cohen LG, Freyre JD, Panizza M. Changes in the N1-P1 component of the somatosensory cortical evoked response in patients with partial limb amputations. Electromyogr Clin Neurophysiol 1984;24:415-427. 10. Levy WJ, Amassian VE, Traad M, Cadwell J . Focal magnetic coil stimulation reveals motor cortical system reorganized in humans after traumatic quadriplegia. Brain Res 1990;510:130-134. 11. Topka HR, Cole R, Hallett M, Cohen I,. Reorganization in the map of outputs of human motor cortex in adults following low thoracic traumatic spinal cord injury [Abstract]. Neurology 1990;40(suppl 1):214. 12. Topka HR, Hallett M, Cohen LG. Reduced facilitation of motor evoked potentials during voluntary activation of target muscles
immediately proximal to the level of a spinal cord injury [Abstract]. Mov Disord 1990;5(suppl 1):33. 13. Williams PL, Warwick R, eds. Gray's Anatomy, 37th ed. New York: Churchill Livingstone, 1989. 14. Mills KR and Murray NMF. Electrical stimulation over the human vertebral column: which neural elements are excited? Electroencephalogr Clin Neurophysiol 1986;63:582-589. 15. Ugawa Y, Rothwell JC, Day BL, Thompson PD, Marsden CD. Magnetic stimulation over spinal enlargements. J Neurol Neurosurg Psychiatry 1989;52:1025-1032. 16. Plassman BL, Gandevia SC. Comparison of human motor cortical projections to abdominal muscles and intrinsic muscles of the hand. Exp Brain Res 1989;78:301-308. 17. Cohen LG, Roth B, Nilsson J, et al. Effects of coil design on delivery of focal magnetic stimulation: technical considerations. Electroencephalogr Clin Neurophysiol 1990;75:350-357. 18. Prendergast J, Stelzner DJ. Increases in collateral axonal growth rostra1 to a thoracic hemisection in neonatal and weanling rat. J Comp Neurol 1976;166:145-162. 19. Bernstein JJ, Bernstein ME. Axonal regeneration and formation of synapses proximal to the site of lesion following hemisection of the rat spinal cord. Exp Neurol 1971;30:336-351.
20. Wolman L. Post-traumatic regeneration of nerve fibres in the human spinal cord and its relation to intramedullary neuroma. J Pathol Bacteriol 1967;94:123-129. 21. Day BL, Thompson PD, Dick J P , Nakashima K, Marsden CD. Different sites of action of electrical and magnetic stimulation of the human brain. Neurosci Lett 1987;75:101-106. 22. Shinoda Y, Yamaguchi T, Futami T . Multiple axon collaterals of single corticospinal axons in cat spinal cord. J Neurophysiol 1986;55:425-448. 23. Shinoda Y, Yokota JI, Futami T. Divergent projections of individual corticospinal axons to motoneurons of multiple muscles in the monkey. Neurosci Lett 1981;23:7-12. 24. Holstege G, Kuypers HGJM, Boer RC. Anatomical evidence for direct brainstem projections to the somatic motoneuronal cell groups and autonomic preganglionic cell groups in the cat spinal cord. Brain Res 1979;171:329-333. 25. McCall RB, Aghajanian GK. Serotonergic facilitation of facial motoneuron excitation. Brain Res 1979;169:11-27. 26. Brown TG, Sherrington CS. On the instability of a cortical point. Proc R SOCLond [Biol] 1912;85:250-277. 27. Lashley KS. Temporal variation in the function of the gyms precentralis in primates. Am J Physiol 1923;65:585-602.
Leg paresthesias induced by magnetic brain stimulation in patients with thoracic spinal cord injury Leonardo G. Cohen, MD; Helge Topka, MD; Reginald A. Cole, MD; and Mark Hallett, MD
Article abstract-We studied the induction of leg paresthesias by magnetic stimulation of the brain in seven patients with thoracic T9- 12 spinal cord injury and in four normal volunteers by delivering transcranial magnetic stimulation over scalp positions 1cm apart with a Cadwell MES-10 magnetic stimulator and an 8-shaped magnetic coil a t 100% stimulus intensity. We asked subjects to report sensations felt after each stimulus. In all normal subjects, magnetic stimulation evoked sensations described as tingling or a wave descending along the leg, usually accompanied by EMG responses in leg muscles. In three of the seven patients, stimulation evoked sensations of tingling, numbness, touch, or a wave descending along the leg, lasting up to 10 seconds and referred to different parts of the legs and toes. In the patients, sensations were felt more distally the closer the site of stimulation was to the midline. Patients with leg paresthesias had less motor reorganization in abdominal muscles than those without paresthesias. These findings suggest that portions of the cortical representation areas for body parts deafferented by a complete spinal cord injury can remain related to those body parts for up to several years. A central origin of these paresthesias is probable. NEUROLOGY 1991;41:1283-1288
Following median nerve transection in the owl and squirrel monkey, large cortical regions in area 3b, which previously received inputs from the deafferented body parts, become responsive to inputs from bordering glabrous skin regions that expanded into the former median nerve representational zone.' These changes start almost immediately following the lesion and be-
come more complex and organized with time. Similar changes occur in cats following thoracic spinal cord injury.*v3 Recent studies using noninvasive techniques suggest that reorganization of sensory and motor pathways may take place in humans following a m p ~ t a t i o n .Reorga~.~ nization also occurs in human motor systems after spi-
~~
From the Human Cortical Physiology Unit, Human Motor Control Section, Medical Neurology Branch, National Institute of Neurological Disorders and Stroke, National Institutes of Health, Bethesda, MD. Dr. Topka is supported by Deutsche Forschungsgemeinschaft, Federal Republic of Germany. Presented in part at the 42nd annual meeting of the American Academy of Neurology, Miami Beach, FL, April 1990. Received August 20, 1990. Accepted for publication in final form January 11, 1991. Address correspondence and reprint requests to Dr. Leonardo G. Cohen, Building 10, Room 5N226, NINDS, NIH. Bethesda, MD 20892.
August 1 9 9 1 NEUROLOGY 41 1 2 8 3
Reorganization of corticospinal pathways following spinal cord injury Helge Topka, Leonardo G. Cohen, Reginald A. Cole, et al. Neurology 1991;41;1276 DOI 10.1212/WNL.41.8.1276 This information is current as of August 1, 1991 Updated Information & Services
including high resolution figures, can be found at: http://www.neurology.org/content/41/8/1276.full.html
Citations
This article has been cited by 28 HighWire-hosted articles: http://www.neurology.org/content/41/8/1276.full.html##otherarticles
Permissions & Licensing
Information about reproducing this article in parts (figures,tables) or in its entirety can be found online at: http://www.neurology.org/misc/about.xhtml#permissions
Reprints
Information about ordering reprints can be found online: http://www.neurology.org/misc/addir.xhtml#reprintsus
Neurology ® is the official journal of the American Academy of Neurology. Published continuously since 1951, it is now a weekly with 48 issues per year. Copyright © 1991 by Edgell Communications, Inc.. All rights reserved. Print ISSN: 0028-3878. Online ISSN: 1526-632X.
22. Jermy AC, Fisher CA, Vincent AC, Willcox NA, Newsom-Davis J. Experimental autoimmune myasthenia gravis induced in mice without adjuvant: genetic susceptibility and adoptive transfer of weakness. J Autoimmun 1989;2:675-688. 23. Scadding GK, Calder L, Vincent A, Prior C, Wray D, NewsomDavis J. Anti-acetylcholine receptor antibodies induced in mice by syngeneic receptor without adjuvants. Immunology 1986;58:151-155. 24. Markmann J , Lo D, Naji A, Palmiter RD, Brinster RL, HeberKatz E. Antigen presenting function of class I1 MHC expressing pancreatic beta cells. Nature 1988;336:476-479. 25. Davies TF. Cocultures of human thyroid monolayer cells and autologous T cells: impact of HLA class I1 antigen expression. J Clin Endocrinol Metab 1985;61:418-422. 26. Pette M, Fujita K, Kitze B, et al. Cytotoxic CD4+ human MBP specific T-lymphocyte lines isolated from MS patients and healthy donors [Abstract]. Neurology 1989;39(suppl 1):173. 27. Ota K, Matsui M, Milford EL, Mackin GA, Weiner HL, Hafler DA. T-cell recognition of an immunodominant myelin basic protein epitope in multiple sclerosis. Nature 1990;346:183-187. 28. Lennon VA, McCormick DJ, Lambert EH, Griesmann GE, Atassi MZ. Region of peptide 125-147 of acetylcholine receptor alpha subunit is exposed at neuromuscular junction and induces experimental autoimmune mysthenia gravis, T cell immunity, and modu l a t i n g a u t o a n t i b o d i e s . P r o c N a t l Acad S c i USA 1985:82:8805-8809.
Reorganization of corticospinal pathways following spinal cord injury Helge Topka, MD; Leonardo G. Cohen, MD; Reginald A. Cole, MD; and Mark Hallett, MD
Article abstract-To
assess changes in the relationship between cortical motor representation areas and their target muscles following spinal cord lesions, we studied motor evoked potentials (MEPs) to transcranial magnetic stimulation in six patients with complete spinal cord injuries at low thoracic levels and eight healthy subjects. Magnetic stimulation a t rest activated a larger fraction of the motoneuron pool and evoked MEPs with shorter latencies from a larger number of scalp positions in muscles immediately rostral to the level of a spinal cord injury than in corresponding muscles in controls. The MEPs associated with maximal voluntary activation were not significantly different in the two groups. These results suggest enhanced excitability of motor pathways targeting muscles rostral to the level of a spinal transection, reflecting reorganization of motor pathways either within cortical motor representation areas or at the level of the spinal cord. The data do not allow the determination of the contribution of spinal or cortical mechanisms. However, they support the notion of a limited flexible relationship between primary motor cortex and its target muscles following alterations of normal input-output patterns. NEUROLOGY 1991;41:1276-1283
Chronic deafferentation induces reorganization within cortical somato~ensory~-~ and motoPp6 representation maps in mammals. Although there may be different patterns of reorganization depending on the developmental stage of the animal, the ability of the CNS to reorganize following peripheral or central lesions ap-
pears preserved to some extent in mature animal^.^^^ Recent studies in adult humans using cortical magnetic stimulation techniques and somatosensory evoked potentials suggest that the capacity to reorganize motor and somatosensory pathways following peripheraP9 or central'0-12lesions is partially preserved in humans.
From Human Cortical Physiology Unit, Human Motor Control Section, Medical Neurology Branch, National Institute of Neurological Disorders and Stroke, National Institutes of Health, Bethesda, MD.
Dr. Topka is supported by Deutsche Forschungsgemeinschaft,Federal Republic of Germany.
Presented in part at the 42nd annual meeting of the American Academy of Neurology, Miami Beach, FL, April 1990 and at the 1st International Congress on Movement Disorders, Washington, DC, April 1990. Received August 17, 1990. Accepted for publication in final form January 11,1991. Address correspondence and reprint requests to Dr. Leonardo Cohen, Building 10, Room 5N226, NINDS, NIH, Bethesda, MD 20892.
1278 NEUROLOGY 41 August 1991
Table 1. Patient characteristics Age at
Age
injury Nature of injury+
Sensory impairment?
Motor impairment
Pt
(yrs)
Sex
1
39
M
Compression fracture T-11
34
Below T- 11
Paraplegic;no lower extremity spasms
2
37
F
Compression fracture T-12
33
Below T-12
Paraplegic;no lower extremity spasms
3
21
M
Compression fracture T-10
16
Below T- 11
Paraplegic; moderate lower extremity spasms
4
40
M
Burst fracture T-11
38
Below T-11
Paraplegic;no lower extremity spasms
5
43
F
Compression fracture T-12
23
Below T-12
Paraplegic;moderate lower extremity spasms
6
47
F
Compression fracture T9-10
42
Below T-10
Paraplegic;moderate lower extremity spasms
(yrs)
* Initial surgical treatment included implantation of Harrington rods. t Total loss of sensation below the level indicated.
Table 2. Motor evoked potentials
A
B
Muscle segment
Controls 8)
Patients (n = 6)
P’
MEPamplitude (PV)
Rostralt Lower
241 t 51 351 t 78
376 f 82 719 k 298
0.1213 0.4386
MEPamplitude (% MNPool)
Rostralt Lower
13.0 ? 3.9 10.7 f 5.0
29.9 t 3.0 31.0 f 11.2
0.0933
MEP threshold (%output)
Rostralt Lower
71.4 f 4.9 71.4 f 4.3
65.8 3.8 64.2 f 11.6
(n
=
*
0.0098 0.5168 0.2478
Values are expressed as mean f standard error.
* Mann-Whitney U test.
Figure 1. (A and B) Schema of transcranial magnetic stimulation and EMG recording. (A) Position of recording electrodes over abdominal wall muscles. Surface electrodes (shaded rectangles) were taped to the skin overlying the muscular portion of rectus abdominis (upper electrode) and external and internal oblique muscles (middle and lower electrodes). Open rectangles are reference electrodes. (B) Top of the subject’s head. The grid is centered on Cz (international 10-20 electrode system). Gridpoints are I cm apart and represent the scalp positions stimulated with the eight-shaped coil. Approximate position of the central sulcus in relation to the coil position is indicated by the thin curved line.
This report describes the reorganization of h u m a n motor pathways following long-term thoracic spinal cord lesions and discusses some of the mechanisms possibly involved.
*
Methods. We studied six patients (mean age, 37 8 years) who had severe traumatic injuries of the spinal cord at low thoracic levels occurring 2 to 20 years previously. We also
t Mean response of muscles rostral to the lesion (upper right, middle right, and middle left).
studied eight normal controls (mean age, 35 -t 14 years). All subjects gave their written informed consent for participation in the study, which was approved by the institute’s clinical investigational review committee. All patients had a complete neurologic examination. Clinical findings were verified, including intraoperative and radiologic findings, with respect to the sensory and motor level of the spinal cord lesion. The patients’ clinical features and relevant medical history are summarized in table 1. None of the patients had severe head trauma or loss of consciousness during the accident. Experimental procedures. Transcranial magnetic stimulation was delivered to different scalp positions using a commercially available magnetic stimulator (Cadwell MES-lo), which generates a multiphasic magnetic stimulus with a peak magnetic flux density of 2.2 tesla in the center of a circular coil (focal-point coil) with an inner diameter of 52 mm and an outer diameter of 98 mm (seven turns of wire). EMG was recorded bilaterally using surface electrodes (DISA 13K60) placed on the skin overlying the abdominal wall muscles. August 1991 NEUROLOGY 41 1277
1000
90.
. .
0
0 00 0
.
80
-
70
-
II.
60-
I
rn
SO -
I SCI PATIENTS (n=6)
CONTROLS (n=8)
CONTROLS
SCI PATIENTS
A
B
SCI PATIENTS
CONTROLS
C
Figure 2. (A through C) Motor evoked potentials (MEPs) from all recording sites i n spinal cord injury (SCZ) patients (n = 6) and controls ( n = 8). ( A ) Amplitudes of M E P s to transcranial magnetic stimulation (absolute amplitudes i n pV). (B) Amplitudes of M E P s expressed as a percentage of the largest M response obtained by stimulation of the supplying ventral roots. (C) Stimulus intensity (in percent output of stimulating unit) required to yield M E P s i n at least three of five trials at a gain of 50 pV. M E P s recorded f r o m lower segments of abdominal wall muscles i n the patients are indicated by open circles. Recording sites in each subject were defined choosing three levels on a line connecting the xyphoid and the right anterior superior iliac spine and one on the left (figure 1A). Given the electrode positions, it is most likely that the uppermost electrodes were recording EMG activity from rectus abdominis and external oblique muscles, whereas the electrodes in the middle and lower positions were recording predominantly from the external oblique with some contribution from the internal oblique. Abdominal wall muscles are supplied by spinal segments between T-6 and T-12.I3 EMG activity was amplified using a bandpass filter of 100 Hz to 2 kHz and recorded with a gain of 50 to 200 pV/div using a conventional EMG device (DISA 1500). The subjects lay supine on a bed in a quiet room during the experiments. Cortical magnetic stimuli were delivered a t rest approximately a t the end of the expiratory phase of a breathing cycle. Baseline and prestimulus muscle activity was monitored by a loudspeaker connected to the recording unit. Subjects were considered to be a t rest when no spontaneous EMG activity was detected in the monitored muscle groups with the recording unit set to a gain of 50 pV/div. Traces were printed during the experiment for later off-line analysis. Amplitudes of 10 to 15 motor evoked potentials (MEPs) to transcranial magnetic stimulation delivered a t maximal stimulus intensity (100%) were measured peak-to-peak and expressed as a percentage of the largest M response obtained by direct stimulation of the ventral roots over the spine (T-5 to T-12).This was done to estimate the percentage of the alpha motoneuron pool activity (% MNP) recruited by each cortical stimulation. Ventral roots were stimulated electrically or magnetically. Supramaximal electrical stimulation was delivered to the spine at T-3 to T-6, T-6 to T-9, and T-9 to T-12 (cathode-anode,respectively) using a Digitimer D180 stimulation device. Electrical stimuli were applied with increasing stimulus intensity until no further increase in the M response of abdominal wall muscles was observed. Magnetic stimulation was delivered to various spinal positions between T-5 and T-12 by a circular coil at maximal stimulus intensity. Electrical or magnetic stimulation over the spine is likely to excite 1278 NEUROLOGY 41 August 1991
ventral roots a t their exit from the spinal ~ a n a l . ' ~To J ~avoid possible displacements of metallic implants in the patients, electrical stimulation generally was used. Motor thresholds for cortical MEPs were determined by delivering transcranial magnetic stimuli from a circular coil at various stimulus intensities with the center of the coil positioned over Cz. Moving the coil 2 cm in any direction from Cz did not change the threshold for M E P activation. The MEPs were amplified and recorded a t a gain of 50 pV/div, and threshold was separately defined for each muscle level as the lowest stimulus intensity that evoked MEPs in at least three of five trials. This intensity was found by reducing or increasing the stimulus intensity in 5-percent steps. The effect of voluntary activation of target muscles on amplitude and latency of MEPs was then assessed. Cortical stimuli were delivered by a circular coil a t maximal stimulus intensity with the center of the coil positioned over Cz during rest or maximal voluntary activation of target muscles. Muscle activity was monitored with a loudspeaker connected to the recording unit but not quantified. In abdominal wall muscles, shortest latencies are obtained with background activation of 10% and largest amplitudes with 60% of maximal voluntary contraction.'6 The cortical representation areas of abdominal muscles were mapped by delivering focal magnetic stimulation to scalp positions 1 cm apart with an eight-shaped coil (Cadwell Corticoil). Technical aspects of this coil design have been discussed elsewhere.I7The induced electric field reaches its peak in the area under the junction of both wings of the coil, which was positioned over the target locations. Two to three stimuli were delivered to each scalp position with the subject at rest. As before, MEP amplitudes were averaged and expressed relative to the largest peripheral M response obtained by stimulation of the ventral roots. Maps of the cortical motor representations were obtained by calculating the mean amplitude of MEPs evoked by magnetic stimulation for each scalp position. Statistical analysis. All results are expressed as mean standard error. The Mann-Whitney U test and the Wilcoxon signed rank test were used for statistical comparisons.
*
CONTROLS
PATIENTS
I *
*
500 pV
L 10 msec
Results. In controls, amplitudes of peripheral M responses recorded from the upper and middle segments of the abdominal wall muscles were typically maximal when the ventral roots in the vicinity of vertebrae T-6 to T-9 were stimulated, and responses in the lower segments were maximal with stimulation over T-9 to T-12. Amplitudes of peripheral M responses recorded from the lower segment, but not from the upper and middle segments, were significantly reduced in the patients (2,087 f 161pV)compared with normal subjects (5,631 f 1,340 pV; p < 0.02), indicating the involvement of low thoracic spinal segments in the spinal injury. Consequently, only M E P recordings obtained from the upper and middle segments of the abdominal wall muscles were considered to represent preserved spinal segments rostra1 to the level of the lesion, and recordings from the lower segments were analyzed separately. Cortical stimuli of equal intensity evoked proportionately larger MEPs in muscles proximal to the level of the lesion (upper and middle segments) in spinal cord injury patients (29.9 3.0 % MNP) than in corre-
Figure 3. M responses to electrical or magnetic stimulation over the spine (a) and MEPs to transcranial magnetic stimulation (b) in five patients and five controls. Recordings shown were made f r o m the middle segment of the abdominal muscles on the right side of the body.
sponding muscles in normal subjects (13.0 f 3.9 ?6 MNP; p = 0.0098) (table 2; figures 2 and 3). Magnetic stimulation delivered at rest evoked MEPs in preserved muscle segments at shorter latencies in patients (16.4 f 0.71 msec) than in controls (18.6 & 0.55 msec; p = 0.039). Proportional amplitudes of MEPs recorded from the lowest muscle segment were more variable from patient to patient and were not consistently larger in the patient group (31.0 -+ 11.2 ?6 M N P in the patients, 10.7 f 5.0 in controls; p = 0.093) (table 2; figures 2 and 3). The muscle level from which the largest relative M E P could be evoked differed from patient to patient with respect to the level of the spinal lesion. With injury a t T-10 (n = 2), MEPs were largest in the middle segment of the abdominal wall muscles, and with injury a t T-11 (n = 2), they were largest in the lower segment. With injury at T-12 (n = 2), there was no clear preferential activation of one muscle segment. MEPs could be evoked at slightly smaller stimulus intensities in patients (65.8 & 3.8% output) than in controls (71.4 & 4.9% output), but the observed difAugust 1991 NEUROLOGY 41 1279
LEFT HEMISPHERE
I
RIGHT HEMISPHERE\
Figure 4 . Map of cortical representation of abdominal wall muscles in a representative SCI patient (patient 4). Recordings were made from the upper segment of the abdominal wall muscles on the right side of the body. Three EMG recordings are superimposed for each scalp position.
ference was not statistically significant for either muscle segment (table 2). In the controls, MEP amplitudes were larger in subjects with low MEP thresholds (r2 = 0.704, p 5 0.0001; linear regression), whereas the patients’ amplitudes correlated only weakly with motor thresholds (r2 = 0.259, p 5 0.031; linear regression). Patients tended to have large MEPs in muscles regardless of the motor threshold. Effects of voluntary activation of target muscles on MEP amplitude and latency. In both groups, voluntary activation of abdominal wall muscles led to a decrease in latency and an increase in amplitude of MEPs recorded from muscle segments proximal to the level of the spinal injury (p = 0.0022; Wilcoxon signed rank test). Mean latency shift was 2.1 0.68 msec in the controls and 1.6 -+ 0.47 msec in the patients. Voluntary activation facilitated the amplitude of cortical MEPs by 34.3 -+ 3.6 75 MNP in controls and by 23.8 -t 6.7 5% MNP in patients. The maximal amplitude obtained during voluntary activation of the target muscles, however, was not significantly different between groups. Effects of voluntary activation on amplitude facilitation tended to be smaller for MEPs recorded from the lower muscle segment in patients (11.95 r 4.23 % MNP) than in controls, but the difference was not significant (p = 0.297). 1280 NEUROLOGY 41 August 1991
Mapping of cortical representation areas. MEPs recorded from all segments of the abdominal wall musculature were evoked from a larger number of scalp positions in the patients (24.8 k 7.5 positions) than in the controls (5.2 ? 2.3 positions; p = 0.02).In both patients and controls, amplitudes of MEPs evoked by focal magnetic stimulation were slightly more variable than those evoked by stimulation with the circular coil. Motor representation areas in patients were expanded in both the anteroposterior axis and the mediolateral axis as compared with controls (figures 4 and 5, A and
B).
In two controls, no MEPs could be obtained by focal magnetic stimulation. In five patients and four controls, unilateral scalp stimulation evoked bilateral muscle responses. In two other controls, MEPs were elicited only with stimulation of the contralateral hemisphere. The largest cortical representation area of abdominal wall muscles determined in a control is shown in figure 5, B (subject 11).
Discussion. Transcranial magnetic stimulation delivered a t rest evokes proportionately larger MEPs from a larger number of scalp positions in muscles rostra1 to the level of the lesion in patients with spinal cord injury
A
UP
%MNP
mid R
R
25 15
cz
15
10
I
10
Ant.
Hemisphere
Post.
%MNP
25
20
B
Left
F
15
% M N P 25
UP
1
R
SUBJECT1
I
15 10
25
10
..
Ant.
Post.
%MNP
25
20 l5
mid R
5
5 Left
I
Ft Post.
Hemisphere
I
SUBJECT11
I
f:i 15
Left
Figure 5. (A and B) Size of cortical motor representation areas targeting abdominal muscles in two representative patients (A) and in two controls (B). M E P s for each scalp position were averaged and expressed as a percentage of the peripheral motoneuron pool activity. T h e X , Y, Z coordinates show the coil position i n the coronal axis ( X ) and i n the anteriorposterior axis ( Y ) and the recruitable fraction of the motoneuron pool activity (% M N P ) (2).U p R = upper segment of the abdominal muscles on the right side of the body; mid R = middle segment of the abdominal muscles o n the right. Note presence of ipsilateral responses in controls and patients. August 1991 NEUROLOGY 41 1281
than in corresponding muscles in normal subjects. Voluntary activation of target muscles facilitates MEPs in both groups and leads to maximal MEP amplitudes in spinal cord injury patients that are not different from those of normal subjects. These results are in accord with an earlier studylo that reported that magnetic stimulation a t rest elicited larger MEPs from a larger number of scalp positions in partially preserved muscles in a patient with complete spinal cord injury a t the C-5 level and in another patient who had partially recovered from a (Cl-C2) spinal cord injury as compared with three normal subjects. The authors inferred that the motor cortex projection system had been reorganized. Larger amplitudes, shorter latencies of magnetically evoked MEPs, and a n enlarged map of motor output a t rest reflect enhanced excitability of motor pathways targeting muscles rostral to the level of a spinal transection. While reorganization of cortical motor representation areas following a complete spinal cord injury is conceivable, it does not exclude rearrangement on the level of the spinal motoneuron pool, or a combination of both. In principle, changes in the excitability of descending pathways that occur on the level of the motor cortex or on the level of the spinal cord could have similar effects on MEPs t o transcranial magnetic stimulation. Thus, the paradigm used does not allow distinguishing the relative contribution of either one. Spinal segments immediately rostral to the lesion might have been reorganized, leading to increased excitability of the spinal alpha motoneuron pool at rest and requiring fewer and less frequent descending volleys to discharge. Sprouting of axon collaterals in spinal segments immediately rostral t o the level of a transection and the subsequent formation of new synapses over time occur in newbornla and adult rats.lg Regenerative sprouting of corticospinal axons rostral to the level of complete spinal cord injuries also occurs in humans.zo Alternatively, enhanced excitability of descending pathways to transcranial magnetic stimulation could reflect reorganization within cortical motor representation areas. Transcranial magnetic stimulation, as utilized in our study, is likely to activate corticospinal neurons transsynaptically rather than exciting directly descending axonsZ1and may activate a larger number of cortical interneurons that synapse onto the same corticospinal neuron following a spinal transection. Corticospinal neurons that previously had subthreshold effects on motoneurons or spinal interneurons might now have become capable of discharging motoneurons utilizing collaterals that connect to the motoneuron pools of several muscles in multiple spinal segments.22*23 Two different mechanisms are conceivable. First, the number of active, excitatory, intracortical interneurons that synapse onto preserved, intact corticospinal neurons might increase over time (sprouting). Second, the strength of existing, but previously inactive, intracortical or corticospinal connections might have been modified in response to alteration of the usual sensorimotor input-output pattern following the spinal cord transection (unmasking). Unlike the sprouting of new dendrites, the unmasking of preexisting synapses 1282 NEUROLOGY 41 August 1991
would not require structural changes and could account for the shifts in cortical motor representation areas that have been observed within hours following peripheral nerve lesions in adult rats.5 In addition to changes within the corticospinal projection system itself, it is possible that the gain-setting of descending brainstem projections organized parallel to corticospinal pathways has changed following the spinal transection. Motor facilitation under the influence of direct monosynaptic connections between caudal areas of the reticular formation and ventral horn cells occurs in mammal^.^^,^^ However, the existence of a similar parallel control system has not been established in humans. T h e techniques applied in our study do not allow the determination of where adaptive modifications of the descending projection system occur and which specific mechanisms are involved. However, the data imply that a flexible functional relationship between motor cortex and target muscles, suggested earlier,26.27 exists in adult humans.
Acknowledgments We thank J. Rothwell and Lisa McShane for helpful comments and B.J. Hessie for editorial assistance.
References 1. Franck J I . Functional reorganization of cat somatic sensorymotor cortex after selective dorsal root rhizotomy. Brain Res 1980;186:458-462. 2. Kalaska JF, Pomeranz B. Chronic paw denervation causes an agedependent appearance of novel responses from forearm in ‘paw cortex’ in kittens and adult cats. J Neurophysiol 1979;42:618-633. 3. Kelahan AM, Ray RH, Doetsch GS. Time dependent changes in the organization of somatosensory cerebral cortex following digit amputation in adult racoons. Somatosens Res 1984;2:49-81. 4. Merzenich MM, Kaas J H , Wall J T , Sur M, Nelson FLJ,Felleman DJ. Progression of change following median nerve section in the cortical representation of the hand areas 3b and 1 in adult owl and squirrel monkeys. Neuroscience 1983;10:639-665. 5. Sanes J N , Suner S, Lando J F , Donoghue J P . Rapid reorganization of adult rat motor cortex somatic representation patterns a f t e r m o t o r n e r v e i n j u r y . P r o c N a t l Acad S c i USA 1988;85:2003-2007. 6.Donoghue J P , Sanes J N . Organization of adult motor cortex representation patterns following neonatal forelimb nerve injury in rats. J Neurosci 1988;8:934-943. 7. McKinley PA, Jenkins WM, Smith JL. Merzenich MM. Agedependent capacity for somatosensory cortex reorganization in chronic spinal cats. Brain Res 1987;428:136-139. 8. Cohen LG, Bandinelli S, Findley TW, Hallett M. Motor reorganization after upper limb amputation in humans: a study with focal magnetic stimulation. Brain 1991;114:615-627. 9. Sica REP, Sanz OP, Cohen LG, Freyre JD, Panizza M. Changes in the N1-P1 component of the somatosensory cortical evoked response in patients with partial limb amputations. Electromyogr Clin Neurophysiol 1984;24:415-427. 10. Levy WJ, Amassian VE, Traad M, Cadwell J . Focal magnetic coil stimulation reveals motor cortical system reorganized in humans after traumatic quadriplegia. Brain Res 1990;510:130-134. 11. Topka HR, Cole R, Hallett M, Cohen I,. Reorganization in the map of outputs of human motor cortex in adults following low thoracic traumatic spinal cord injury [Abstract]. Neurology 1990;40(suppl 1):214. 12. Topka HR, Hallett M, Cohen LG. Reduced facilitation of motor evoked potentials during voluntary activation of target muscles
immediately proximal to the level of a spinal cord injury [Abstract]. Mov Disord 1990;5(suppl 1):33. 13. Williams PL, Warwick R, eds. Gray's Anatomy, 37th ed. New York: Churchill Livingstone, 1989. 14. Mills KR and Murray NMF. Electrical stimulation over the human vertebral column: which neural elements are excited? Electroencephalogr Clin Neurophysiol 1986;63:582-589. 15. Ugawa Y, Rothwell JC, Day BL, Thompson PD, Marsden CD. Magnetic stimulation over spinal enlargements. J Neurol Neurosurg Psychiatry 1989;52:1025-1032. 16. Plassman BL, Gandevia SC. Comparison of human motor cortical projections to abdominal muscles and intrinsic muscles of the hand. Exp Brain Res 1989;78:301-308. 17. Cohen LG, Roth B, Nilsson J, et al. Effects of coil design on delivery of focal magnetic stimulation: technical considerations. Electroencephalogr Clin Neurophysiol 1990;75:350-357. 18. Prendergast J, Stelzner DJ. Increases in collateral axonal growth rostra1 to a thoracic hemisection in neonatal and weanling rat. J Comp Neurol 1976;166:145-162. 19. Bernstein JJ, Bernstein ME. Axonal regeneration and formation of synapses proximal to the site of lesion following hemisection of the rat spinal cord. Exp Neurol 1971;30:336-351.
20. Wolman L. Post-traumatic regeneration of nerve fibres in the human spinal cord and its relation to intramedullary neuroma. J Pathol Bacteriol 1967;94:123-129. 21. Day BL, Thompson PD, Dick J P , Nakashima K, Marsden CD. Different sites of action of electrical and magnetic stimulation of the human brain. Neurosci Lett 1987;75:101-106. 22. Shinoda Y, Yamaguchi T, Futami T . Multiple axon collaterals of single corticospinal axons in cat spinal cord. J Neurophysiol 1986;55:425-448. 23. Shinoda Y, Yokota JI, Futami T. Divergent projections of individual corticospinal axons to motoneurons of multiple muscles in the monkey. Neurosci Lett 1981;23:7-12. 24. Holstege G, Kuypers HGJM, Boer RC. Anatomical evidence for direct brainstem projections to the somatic motoneuronal cell groups and autonomic preganglionic cell groups in the cat spinal cord. Brain Res 1979;171:329-333. 25. McCall RB, Aghajanian GK. Serotonergic facilitation of facial motoneuron excitation. Brain Res 1979;169:11-27. 26. Brown TG, Sherrington CS. On the instability of a cortical point. Proc R SOCLond [Biol] 1912;85:250-277. 27. Lashley KS. Temporal variation in the function of the gyms precentralis in primates. Am J Physiol 1923;65:585-602.
Leg paresthesias induced by magnetic brain stimulation in patients with thoracic spinal cord injury Leonardo G. Cohen, MD; Helge Topka, MD; Reginald A. Cole, MD; and Mark Hallett, MD
Article abstract-We studied the induction of leg paresthesias by magnetic stimulation of the brain in seven patients with thoracic T9- 12 spinal cord injury and in four normal volunteers by delivering transcranial magnetic stimulation over scalp positions 1cm apart with a Cadwell MES-10 magnetic stimulator and an 8-shaped magnetic coil a t 100% stimulus intensity. We asked subjects to report sensations felt after each stimulus. In all normal subjects, magnetic stimulation evoked sensations described as tingling or a wave descending along the leg, usually accompanied by EMG responses in leg muscles. In three of the seven patients, stimulation evoked sensations of tingling, numbness, touch, or a wave descending along the leg, lasting up to 10 seconds and referred to different parts of the legs and toes. In the patients, sensations were felt more distally the closer the site of stimulation was to the midline. Patients with leg paresthesias had less motor reorganization in abdominal muscles than those without paresthesias. These findings suggest that portions of the cortical representation areas for body parts deafferented by a complete spinal cord injury can remain related to those body parts for up to several years. A central origin of these paresthesias is probable. NEUROLOGY 1991;41:1283-1288
Following median nerve transection in the owl and squirrel monkey, large cortical regions in area 3b, which previously received inputs from the deafferented body parts, become responsive to inputs from bordering glabrous skin regions that expanded into the former median nerve representational zone.' These changes start almost immediately following the lesion and be-
come more complex and organized with time. Similar changes occur in cats following thoracic spinal cord injury.*v3 Recent studies using noninvasive techniques suggest that reorganization of sensory and motor pathways may take place in humans following a m p ~ t a t i o n .Reorga~.~ nization also occurs in human motor systems after spi-
~~
From the Human Cortical Physiology Unit, Human Motor Control Section, Medical Neurology Branch, National Institute of Neurological Disorders and Stroke, National Institutes of Health, Bethesda, MD. Dr. Topka is supported by Deutsche Forschungsgemeinschaft, Federal Republic of Germany. Presented in part at the 42nd annual meeting of the American Academy of Neurology, Miami Beach, FL, April 1990. Received August 20, 1990. Accepted for publication in final form January 11, 1991. Address correspondence and reprint requests to Dr. Leonardo G. Cohen, Building 10, Room 5N226, NINDS, NIH. Bethesda, MD 20892.
August 1 9 9 1 NEUROLOGY 41 1 2 8 3
Reorganization of corticospinal pathways following spinal cord injury Helge Topka, Leonardo G. Cohen, Reginald A. Cole, et al. Neurology 1991;41;1276 DOI 10.1212/WNL.41.8.1276 This information is current as of August 1, 1991 Updated Information & Services
including high resolution figures, can be found at: http://www.neurology.org/content/41/8/1276.full.html
Citations
This article has been cited by 28 HighWire-hosted articles: http://www.neurology.org/content/41/8/1276.full.html##otherarticles
Permissions & Licensing
Information about reproducing this article in parts (figures,tables) or in its entirety can be found online at: http://www.neurology.org/misc/about.xhtml#permissions
Reprints
Information about ordering reprints can be found online: http://www.neurology.org/misc/addir.xhtml#reprintsus
Neurology ® is the official journal of the American Academy of Neurology. Published continuously since 1951, it is now a weekly with 48 issues per year. Copyright © 1991 by Edgell Communications, Inc.. All rights reserved. Print ISSN: 0028-3878. Online ISSN: 1526-632X.
