Focal Capsular Vascular Lesions Can
Selectively Deafferent the Prerolandic or the
Parietal Cortex: Somatosensory Evoked
Potentials Evidence
Frangois Mauguiere, MD," and John E. Desmedt, MDI-

Four patients with a unilateral focal vascular accident involving the internal capsule (but not the cortex) were studied
electrophysiologically. Averaged somatosensory evoked potentials (SEPs) to electrical stimulation of the median nerve
on the left or the right side were analyzed. In the 3 patients with hemiparesis and normal somatic sensation, the
precentral P22 and N30 SEP components were lost, whereas the parietal components were preserved. In another
patient with clinical somatosensory loss unaccompanied by any central motor impairment, the precentral SEP components were preserved, whereas the parietal SEP components were lost. Thus, a small capsular lesion can eliminate
distinct cortical SEP components by selectively involving either the axons of the thalamic VPLc nucleus going to
parietal receiving cortex or the axons of thalamic VPLo going to motor area 4. These findings extend to subcortical
lesions the diagnostic value of SEPS in patients with dissociated clinical motor and sensory signs.
Mauguicre F, Desmedt JE. Focal capsular vascular lesions can selectively deafferent the prerolandic or the
parietal cortex: somatosensory evoked potentials evidence. Ann Neurol 1991;30:71-75

scalp somatosensory evoked potentials
(SEPs) to median nerve stimulation reveal compound
profiles El-S] and their distinct short-latency compoAveraged

nents reflect evoked responses of either precentral
(components P22, N30) or postcentral (N20, P27) cortical areas respectively because they undergo selective
alterations or loss in patients with focal cortical lesions
19-12]. Persistence of the precentral P22-N30 SEP
components after destruction of parietal cortex implies that these components can be activated directly
through distinct projections {9] from the thalamic
ventro-postero-lateralis pars oralis nucleus to motor
area 4 {13-161 in the absence of parietofrontal connections. There is no evidence as to whether the
thalamocortical fibers projecting respectively to either
precentral or postcentral cortex are anatomically segregated in the internal capsule. In this report, we describe
4 patients in whom a small focal capsular lesion selectively involved one of these two sets of thalamocortical
fibers. These results reveal (1) an anatomical segregation of thalamocortical somatosensory axons on their
way to precentral or postcentral cortex and (2) the diag-

From the *Department of Electroencephalographyand Epileptology,
Faculty of Medicine Lyon-Nord, Hopital Neurologique, Lyon,
France, and the tBram Research Unit, University of Brussels Faculty
of Medicine, Brussels, Belgium.

nostic value of SEPs in patients with subcortical lesions
and dissociated clinical motor and sensory signs.

Methods
The extent of the focal lesion was documented by computed
tomography (CT), nuclear magnetic resonance (NMR), or
both. Clinical somatosensory deficits were carefully assessed
at the time of SEP recording (see 181). SEPs were evoked
by 0.2-msec electrical pulses delivered to the right or to the
left median nerve at the wrist (cathode located proximally,
intensity adjusted to elicit a slight thumb twitch). Intervals
between stimuli varied randomly between 500 and 600 msec.
The proximal arm was grounded. Eight recording electrodes
(impedance less than 5 kbl) were placed bilaterally over the
scalp. The afferent volley was monitored over the brachial
plexus at Erb's point. The recording channels were referred
to the earlobe. A system bandpass of 1.6 to 3,200 Hz and
136-ksec bins (computer sampling time) were used for averaging. Throughout the recording session, the electroencephalogram (EEG) was monitored, and EEG samples with excess
interference from muscles or eye blinks were automatically
edited out from the average. SEP components were labeled
for polarity (N for negative, P for positive) and modal peak

Received Aug 20, 1990, and in revised form Dec 3 1. Accepted for
publication Jan 7, 1991.

Address correspondence to Dr Desmedt, Brain Research Unit, 115
Boulevard de Waterloo, Brussels 1000, Belgium.

Copyright 0 1991 by the American Neurological Association

71

Fig 2. Patient 2; small hematoma in left internal capsule with
no involvement of cortex. Somatosensory evoked potentials t o stimulation of right (A) or ldt (B) median nerve at the wrist. Same
presentation as Figure I . No cortical parietal response to stimulation on the affected side (A, lower trace). (C)Drawings of
computed tomographic scans.
Fig 1. Patient 1; small hematoma in le& internal capsule with
no tnvolvement of cortex. Somatosensory evoked potentials (SEPs)
to stimulation of ldt (A)or nght (B) median nerve at the
wrist. For each frame, the upper traces present SEPs recorded at
contralateralfmntal scalp and the lower trace, SEPs at contralateral parzetal scab. SEP components are labeled. No cortical
frontal response to stimulation on the afsected side (A, upper
trace). (C) Computed tomographic scans.
latency in normal adults of standard body size. Two averages
of 2,000 trials each were compared for stimulation on either
the left or the right side. Ethics committees approved the
study and subjects gave informed consent.

Patient Reports
Palient 1
This 58-year-old right-handed woman presented with a sudden left hemiplegia with rotation of eyes and head to the
right. Touch, joint, vibration, temperature sensations, as well
as stereognosis were preserved on the left side. There was
left tactile extinction on bilateral stimulation of the hands and

72 Annals of Neurology

Vol 30 No 1 July 1991

left visual and auditory neglect. CT scans showed the integrity of the cerebral cortex including the rolandic regions, and
a right capsular ischemic lesion in the anterior choroid artery
territory (Fig 1C). SEPs were recorded 14 days after the
vascular accident when the right hemiplegia was still severe.
SEPs to median nerve stimulation on the normal (right)
side disclosed a P14 farfield and contralateral parietal N20,
P27, P45, and N6O components, as well as precentral P22
and N30 components (Fig 1B).SEPs to median nerve stimulation on the affected (left) side showed normal P14 farfield,
loss of precentral responses, and fairly preserved parietal
N20, P27, P45, and NGO.

Patient 2
This 5 7-year-old right-handed man suffered from untreated
hypertension and presented with a sudden right hemiplegia
with aphasia. CT scans showed a left capsular hematoma (Fig
2C). Clinical examination 1 month later disclosed a right
brachiofacid anesthesia with loss of touch and joint sensations, astereognosis, and reduced vibration sensations. Motor
aphasia and homonymous hemianopsia were present.
SEPs were recorded 2 years later when the blood pressure

had been stabilized through antihypertensive therapy. The
neurological condition and verbal performance had improved
to a small but significant extent, and there had been no clinical evidence for any progression of the cerebrovascular disease that has actually remained stable up to now. The right
hemiplegia was still severe, and there was a right hemihypoesthesia for touch, joint, vibration, pain, and temperature.
The right hand had a severe astereognosis as well as ataxia
in the absence of visual control (“main instable ataxique” of
Alajouanine [17}). Stimulation of the median nerve on the
normal (left) side elicited normal SEPs with all classical components (Fig 2B). SEPs to stimulation of the median nerve
o n the affected (right) side showed loss of parietal responses,
but persistence of a slowed and somewhat delayed precentral
positivity, which was identified as P22. This was followed by
a negativity identified as a frontal N30 (Fig 2A).

Patient 3
This 49-year-old right-handed woman woke up in the morning with a right hemiparesia, which regressed in about 3
months. Five years later, she presented after another accident
of sudden onset with right hemiparesia and severe left cephalalgia. CT scans showed a left capsular hematoma (Fig 3C).
Arteriography further disclosed a left caudate angioma. SEPs
were recorded 1 month after the second episode at which
time she had a right hemiparesis with hemihypoesthesia.
Touch, joint, vibration, temperature, and pain sensations
were clearly reduced but not lost on the right side. The right
hand had astereognosis and impaired graphesthesia. SEPs to
stimulation of the median nerve on the affected (right) side
showed loss of precentral components and fairly preserved
parietal responses (Fig 3A).
Patient 4
This 76-year-old right-handed man suddenly developed a
right ataxic syndrome with unsteadiness of the right foot and
clumsiness of the right hand (main instable ataxique 1171).
At the time of SEP recording 9 days later, there was no
hemiparesis nor any cerebellar sign. The tendon reflexes
were symmetrical. All somatic sensations were normal. CT
scans showed a small left capsulothalamichematoma (Fig 4C)
and NMR, 4 weeks later, disclosed residual hyposignal (Fig
4D). SEPs to stimulation of the median nerve on the affected
(right) side showed loss of frontal responses and fairly preserved parietal response with slowed time course (Fig 4 A ) .

Discussion
Distinct parietal and frontal short-latency SEP components have been disclosed in scalp topographical mapping C18-221, and critical evidence for either precentral or postcentral location of their neural generators
relies essentially on their selective loss in neurological
patients with well-defined focal lesions [9, 101. T h e
combined evidence suggests parietal generators located
in area 3b for component N20 and in area 1 for P27,
both components being lost in patients with a parietal
lesion. In front of the central fissure, component P22
manifests a motor area 4 generator, whereas N30 appears to involve several generators that can presumably

Fig 3.Patient 3; small hematom in le$t internal capsule with
no i m o l v m n t of cortex. Same presentation as Figure 1. No cortical precentral response t o stimulation on the affected side (A,
upper truce). (Ci Computed tomographic scans.

be located in premotor and supplementary motor areas
because N30 persists after frontal lobectomy or postcentral lesions [9]. Motor area excitation plays a role
in the activation of N30 through known corticocortical
connections C23, 241. W h e n assessing N30 in patients,
it should be remembered that this frontal component
is vulnerable to short interstimulus intervals and also
deteriorates rather early in chronological aging IS].
Therefore, it is desirable to compare the N30 elicited
by right or left stimulation in patients with a unilateral
lesion.
T h e present results help assess the clinical significance of the known corticocortical connections from
parietal areas 1, 2 , 5 , and 7 to precentral areas 4 and

Mauguikre and Desmedt: Focal Capsular Vascular Lesions

73

the (preserved) parietofrontal connections indeed appear unable to elicit the short-latency precentral SEPs.
We think that the function of these corticocortical connections is rather to mediate longer latency interactions
such as manifested by Pl00 cognitive components

wi.

Fig 4. Patient 4; small hemutom in lejit internal capsule with
no involvement of cortex. Sam presentation as Figure 1. No cortical precentral response to stimulation on the ufiected side (A,
upper trare). (C) Computed tomographic scan. (D) Nuclear
magnetic resonance, 4 weeks later.

The results also document an issue that had not been
clarified in anatomical studies on primates, namely,
whether the thalamocortical fibers projecting respectively to precenual or postcentral cortex might present
any degree of anatomical segregation in their course
through the internal capsule. We found that either of
these two sets of fibers could be fairly selectively involved by a small subcortical vascular lesion. SEP evidence in our patients indicates that one set of fibers
was blocked or lesioned resulting in the loss of the
corresponding cortical SEP responses, whereas the
other set of fibers still evoked fairly preserved cortical
responses. In all patients except Patient 2, the anatomical integrity of the cortex could be documented by
neuroimaging methods at the time of SEP recording.
In Patient 2, the arterial pressure was stabilized by
therapy and there had been no clinical evidence for
any progression of the cerebrovascular disease during
the delay of 2 years between CT scan and SEP recording. The clinical status of this patient has recently
been found unchanged, 11 years after the vascular accident. Therefore, the loss of parietal SEP response and
clinical hypoesthesia in Patient 2 can be assumed to
reflect cortical deafferentation through a single subcortical lesion (capsular hematoma) inflicted during the
acute episode.
Although the diagnostic value of SEPs in focal cortical lesions seems well established, their usefulness in
capsular lesions had hitherto not been documented.
Indeed, posterior thalamic or capsular lesions frequently result in the loss of all cortical SEPs {S, 251.
Our present data reveal the diagnostic value of SEPs
for documenting dissociated clinical motor and sensory
signs in patients with capsulothalamic lesions.
Ths research was supported by grants from the Faculty of Medicine
Lyon-Nord of the University Claude Bernard and from the Belgian
Fund for Medical Scientific Research.

References

6 [15, 16, 23, 241. In patients with a complete parietal
lesion, these connections are lost, but the precentral
P22 and N30 are still recorded, which suggests that
these short-latency SEP responses are evoked, not by
corticocortical connections from parietal cortex, but directly through thalamic projections from VPLo to motor cortex 197. The present results validate this interpretation by showing that precentral P22 and N30 can
be lost in patients with a capsular lesion but intact precentral and postcentral cortex (Figs 1, 3, 4), in whom

74 Annals of Neurology Vol 30 No 1 July 1991

1. Noel P, Desmedt JE. Cerebral SEPs after vascular lesions of the
brainstem and diencephalon. Brain 1975;98:113-128
2. Green JB, McLeod S. Short-latency SEPs in patients with neurological lesions. Arch Neurol 1979;36:846-85 1
3. Chiappa KH, Choi SK, Young RR. Short-latency SEPs following median nerve stimulation in patients with neurological lesions. In: Desmedt JE, ed. Progress in clinical neurophysiology,
vol7. Base1 Karger, 1980:264-281
4. Anziska B, Cracco RQ: Short-latency SEPs to median neme
stimulation. Electroenceph Clin Neurophysiol 1981;52:53 1-

537
5. Desmedt JE, Cheron G. Noncephalic reference recording of

early SEPs to finger stimulation in adult or aging man: differentiation of widespread N18 and contralareral N20 from prerolandic P22 and N30. Electroencephalogr Clin Neurophysiol
1981;52:553-570
6. Yamada T, Kimura J, Wilkinson JT,et al. Short- and longlatency median nerve SEPs. Arch Neurol 1983;40:215-220
7. Liiders H, Lesser R, Hahn J, et al. Subcortical SEPs to median
nerve stimulation. Brain 1983;106:341-372
8. MauguiPre F, Desmedt JE. Thalamic pain syndrome of
Dejgrine-Roussy. Arch Neurol 1988;45:1312-1320
9. Mauguiere F, Desmedt JE, Courjon J. Astereognosis and dissociated loss of frontal and parietal components of SEPs in hemispheric lesions. Brain 1983;106:27 1-3 11
10. Slimp JC, Tamas LB, Stolov WC, et al. SEPs after removal of
somatosensory cortex in man. Electroencephalogr Clin Neurophysiol 1986;65:111-117
11. Tsuji S, Murai Y , Kadoya C. Topography of SEPs to median
nerve stimulation in patients with cerebral lesions. Electroencephalogr Clin Neurophysiol 1988;71:280-288
12. Ebner A, Deuschl G. Frontal and parietal components of enhanced SEPs. Electroencephalogr Clin Neurophysiol 1988;71:
170-179
13. Lemon RN, van der Burg J. Shon-latency peripheral inputs to
thalamic neurones projecting to the motor cortex in the monkey. Exp Brain Res 1979;36:445-462
14. Tanji J, Wise SP. Submodality distribution in the sensorimotor
cortex of the unanesthetized monkey. J Neurophysiol 1981;
45:467-481
15. Jones EG. The nature of afferent pathways conveying shortlatency inputs to the primate motor cortex. In: Desmedt JE, ed.
Motor control in health and disease. New York Raven Press,
19831263-285

16. Mountcastle VB. Central nervous system mechanisms in mechanoreceptive sensibhty. In: Handbook of physiology, sect 1. The
nervous system, vol 3. Bethesda: American Physiological Society, 1984:789-878
17. Alajouanine T, Thurel R, Ombredane A. Somarognosie er apraxie du membre sup6rieur gauche. Rev Neurol (Paris) 1934;
41:695-703
18. Desmedt JE, Bourguet M. Color imaging of scalp topography
of parietal and frontal components of SEPs to stimulation of
median or posterior tibial nerve in man. Electroencephalogr Clin
Neurophysiol 1985;62:1-17
19. Deiber MP, Giard MH, MauguiPre F. Separate generators with
distinct orientations for N20 and P22 SEPs to finger stimulation.
Electroencephalogr Clin Neurophysiol 1986;65:321-334
20. Desmedt JE, Nguyen T H , Bourguet M. Color imaging of human evoked potentials with reference to the N20, P22, P27 and
N30 somatosensory responses. Electroencephalogr Clin Neurophysiol 198?;68:1-22
21. MauguiGre F, Desmedt JE. Bilateral SEPs in 4 patients with
long-standing surgical hemispherectomy. Ann Neurol 1989;
26:724-73 1
22. DesmedtJE, Tomberg C. Mapping early SEPs in selective attention. Electroencephalogr Clin Neurophysiol 1989;74:321-346
23. Jones EG, Powell TP. An anatomical study of convergent sensory pathways within the cerebral cortex of the monkey. Brain
1970;93:793-820
24. Petrides M, Pandya DN. Projections to the frontal cortex from
the posterior parietal region in the rhesus monkey. J Comp
Neurol 1984;228:105-116
25. MauguiGre F, DesmedtJE, Courjon J. Neural generators of N18
and P14 farfield SEPs. Electroencephalogr Clin Neurophysiol
I983 ;56:283-292

MauguGre and Desmedt: Focal Capsular Vascular Lesions 75