- Email: [email protected]
Spinal and early scalp-recorded components of the somatosensory evoked potential following stimulation of the posterior tibial nerve
320
Electroencephalography and clinical Neurophysiologv, 1983, 55 : 320 330 Elsevier Scientific Publishers Ireland, Ltd.
S P I N A L AND EARLY S C A L P - R E C O R D E D C O M P O N E N T S OF T H E S O M A T O S E N S O R Y EVOKED P O T E N T I A L F O L L O W I N G S T I M U L A T I O N OF T H E P O S T E R I O R TIBIAL NERVE M A S U D SEYAL J, R O N A L D G. EMERSON and T I M O T H Y A. PEDLEY
EEG Systems Laboratory for Special Studies, The Neurological Institute, Columbia University College of Pt~vsicians and Surgeons, New York, N. E 10032 (U.S.A.) (Accepted for publication: November 2, 1982)
Recent reports have characterized the human early somatosensory evoked potential (SEP) elicited by stimulation of the median nerve and have provided clues to its neural generators (Desmedt and Brunko 1980; Desmedt and Cheron 1980; 1981a, b; Noel and Desmedt 1980; Allison and Hume 1981). The early SEP following stimulation of peripheral nerves in the lower extremity has been less extensively studied and its neural generators are correspondingly less certain (Tsumoto et al. 1972; Lueders et al. 1981; Rossini et al. 1981; Vas et al. 1981; Cruse et al. 1982; Kakigi et al. 1982). Furthermore, there is no consensus to date about such fundamental characteristics of the lower limb SEP as number and polarity of the early components, scalp topography of the response, and preferred recording techniques. Indeed, it is likely that variations in recording methods such as use of different montages, filter settings, stimulus rates, and in one instance general anesthesia, have contributed substantially to lack of uniform results. We now present a detailed analysis of the early somatosensory evoked potential following stimulation of the posterior tibial nerve (PTN) in the normal adult. Recordings were made from multiple electrodes over the spine and scalp using both Correspondence to: Timothy A. Pedley, M.D., The Neurological Institute, 710 West 168th St., New York, N.Y. 10032, U.S.A. Presented in part at the annual meeting of the American EEG Society in Phoenix, Ariz., November 11-13, 1982. Present address: Dept. of Neurology, University of California Medical Center, 2315 Stockton Blvd, Sacramento, Calif. 95817, U.S.A.
cephalic and non-cephalic reference electrodes. Our results clearly delineate both localized and widespread components of the scalp-recorded potentials and permit conclusions to be drawn regarding their possible neural generators. Materials and Methods
SEPs were recorded following stimulation of 22 legs in 12 normal adult volunteers (11 men) ranging in age from 18 to 42 years. The study was approved by the Institutional Review Board of Columbia-Prebyterian Medical Center and informed consent was obtained from participants. Volunteers lay on a hospital bed in a quiet, darkened room and were given oral diazepam to promote muscle relaxation and sleep. A typical recording session lasted up to 8 h, and several persons were studied on 2 or 3 different occasions. Stimuli were 0.2 msec square wave electrical pulses applied through electrodes placed over the posterior tibial nerve just behind the medial malleolus. The cathode was located 2 cm proximal to the anode. Stimulus intensity was adjusted to elicit a consistent twitch of the great toe that was not uncomfortable to the subject. The stimulation rate was 4/sec. Four recording electrodes were placed over the thoracic and lumbar spine spaced at 10 cm intervals beginning at the first lumbar spinous process (L~). Electrodes were placed over the fifth and second cervical spines (SC5 and SC2). Scalp electrodes were positioned according to the International 10-20 system. Further resolution was obtained by using additional electrodes located mid-
0013-4649/83/0000-0000/$03.00 © 1983 Elsevier Scientific Publishers Ireland, Ltd.
POSTERIOR TIBIAL SEP
321
way between standard sagittal placements. These are identified by a (') following the name of the anterior most neighboring conventional electrode (Fig. 3A inset). For example, Cz' refers to an electrode located midway between Cz and Pz. Electrodes over the shoulder, elbow and SC5 were used as non-cephalic references. Simultaneous 8-channel recordings were obtained using amplifiers with a gain of 100,000 and half-frequency high and low pass filters set to 30 Hz and 3 kHz, respectively. Signal averaging was performed by a computer with an analog to digital converter operating in an 8-channel multiplexed mode obtaining 10,000 samples/sec for each channel. Signals were averaged for 50 msec following each stimulus, and an average of 1000-4000 samples comprised each trial. Samples with muscle artifact were automatically excluded. Signal averaging was repeated at least once to ensure stability and reproducibility of the response. No curve smoothing was used. Evoked potential wave forms were plotted on a hard-copy device and also stored by the computer for further analysis. The various waves of the scalp SEP were labeled according to their positive (P) and negative (N) polarities at the presumed 'active' electrode and l 1-L1+10
L1+10
Ll+20
L: .......
SC5
SC2
~'~
f
\
\ G1 NEG UP O 5~,V
J
2 ~,.,
Fig. l. Spinal volley following stimulation of the right PTN. Vertical spacing is proportional to the interelectrode distance. Arrows indicate the onset of negativitycorresponding to arrival of the depolarizing wave front at the lower electrode of each recording pair. Electrode positions over lumbar and thoracic spine are specified as distance in centimeters rostral to L~. Two separate trials (dotted lines) and the average (solid lines) are shown.
given nominal latencies reflecting the mean latency for each peak in our group of subjects. Since absolute latencies have been used, there is considerable variation in latency measurements reflecting primarily the different heights of our volunteers. Results
The spinal volley In 10 individuals, recordings were obtained between successive electrodes at L], at 10, 20 and 30 cm above L], over SC5 and over SC2. In one person, additional recordings were made using SC5 as a reference for each of the other spinal electrodes. Fig. 1 illustrates the typical response. Vertical spacing in the figure is proportional to the interelectrode distances on the back. The wave form of the spinal volley was usually triphasic with a small initial positive deflection followed by a major negative wave and a subsequent positivity as seen by the lower electrode in a bipolar chain. The amplitude of the spinal volley was maximal over the lower thoracic region and gradually decreased over the upper thoracic and cervical areas. This was observed both in bipolar and referential recordings. There was a gradual increase in the latency of the response rostral from L~. For each subject, the onset latency of the initial negativity (Fig. 1, arrows) was measured at the various spinal levels. This measurement corresponds to the arrival of the leading edge of the depolarizing wave front of the afferent volley at the distalmost electrode of the recording pair. These values are listed in Table I along with the results of least squares linear regression analysis of the arrival time of the negativity versus distance from L]. A linear relationship between these two variables is demonstrated. This is shown graphically in Fig. 2 in which distance from LI of the distalmost electrode in each recording pair is plotted against arrival time of the afferent volley for 3 representative subjects.
The scalp-recorded potential Three early components were present with widespread, although not entirely uniform, distribution over the scalp in both sagittal and coronal planes (Fig. 3). They consisted of two initial small
322
M. S E Y A L ET AL.
40 30 5
E .o
20
O o
I0
~5 0
,o ,', ,'2
,'5 16 ,'7
2'od,
Latency (msec) Fig. 2. Onset latency of the spinal volley is plotted against distance of the recording electrode from L 1 for 3 representative subjects (subjects 1, 2 and 7 in Table I). The close fit of the original data to the regression lines (solid lines) is apparent. For clarity of presentation, only 3 sets of data from the table are plotted.
positive waves designated P28 and P31 with peak latencies of 27.6 + 1.9 msec (1 S.D.) and 31.2 + 1.9 msec respectively followed by a larger amplitude n e g a t i v i t y , N34 (peak latency 33.9 + 1.9 msec). A t recording sites away from the vertex, N34 showed a
gradual return to the baseline approximately 40 msec after the stimulus. These early potentials were seen only using non-cephalic reference electrodes since in-phase cancellation eliminated them from scalp to scalp recordings. In two individuals, several additional smaller but highly reproducible responses were seen between the onset of P2s and the peak of N34. Because these deflections were not seen in most persons, they will not be discussed further. They may, however, correspond to the multiple early potentials described by Rossini et al. (1981). At Cz and adjacent areas ipsilateral to the stimulus, the negative deflection of N34 was terminated by the onset of a large positive wave, P3s. This was followed by a major negativity of variable latency peaking about 45 msec after the stimulus. Unlike P28, P~l and N 3 4 , P38 had a restricted field involving primarily the central parasagittal region. To study the localized components of the SEP further, scalp-to-scalp recordings were obtained. Fpz was used as a reference since non-cephalic recordings had demonstrated that it was virtually inactive with respect to potentials occurring after 34 msec (Fig. 3B). Fig. 4 shows the nearly total in-phase cancellation of the 3 widespread early
TABLE I
Spinal volley following stimulation of P T N . No.
PT~
Onset latency (msec)
Length
stimulated
1 2 3 4 5 6 7 8 9 10
L L L L R L R L R R
(cm) LI
L I + 10cm
L I + 20cm
LI + 30crn
SC 5
L I -SC~
17.5 18.8 18.8 17.6 17.4 17.2 14.8 17.5 18.4 21.8
19.6 20.2 20.5 19.2 19.1 19.0 15.8 19.2 20.0
20.2 22.0 23.0 20.6 20,6 21.3 17.8 20.1 21.3
21.7 23.6
24.2 25.0 24.7 22.8 24.0 24.1 20.6 25.4 24.0 27.8
45 45 45 44 43 44 44 38 46 43
* Slope of least squares linear regression line. ** Correlation coefficient. *** P r o b a b i l i t y ( P ) .
not recorded
23.0 23.2 19.5 23.0
Slope * (m/see)
r **
69 69 69 82 62 58 62 56 68
0.991 0.993 0.992 0.991 0.997 0.979 0.996 0.990 0.999
P ***
tess than
0.00 t 0.001 0.05 0.001 0.001 0.001 0.001 0.001 0.001
POSTERIOR TIBIAL SEP
323 F p z-SC 5
I
+
.
~.
"~-~.
.
Fz'-SC5
\, ~, '/ ~l' c, cz c z C4c6, ~4 I~z'r~c. z~c~ c~ c, c,c/c c' FZ
V7
T4-8C5
I~
C ~4 ~-~_ C2-8C5
S
C
5
~
/
Cl-SC5 C3-8C5
p
~ Pz'-SC5
T3-SC5 C1-SHOULDER C3-SHOULDER
•
pP~¢~.=,
' ~"
",-.
A
vt
Jl.v
~
2
mJ~c
Fig. 3. SEP to left PTN stimulation. A: the top channel, SC5 to shoulder reference, illustrates the arrival of the spinal volley at SC5 (arrow). The next 8 channels show the scalp-recorded SEP in a coronal plane through Cz using an SC5 reference. The last two channels show the activity at scalp electrodes C] and C 3 using a shoulder reference. Three peaks, P2~, P31 and N34, are present in all channels except the first indicating their widespread distribution over the scalp. Each of these peaks is also present in scalp leads connected to a shoulder reference (last two channels). The cephalic origin of these potentials is confirmed by their absence from the SC5-shoulder derivation. Note the different calibration in channel 1. B: scalp SEP following left PTN stimulation recorded in the midline sagittal plane using an SC5 reference. As in A, P28, P~l and N~4 are widely distributed. Fpz is virtually inactive for events occurring after 35 msec.
T4 Fpz C4-Fpz C2-Fpz
Cz-Fpz
~ .
~" ~"
CS-Fpz T3-Fpz
°5,7
p o t e n t i a l s that o c c u r s w i t h s c a l p - t o - s c a l p recordi n g s w i t h p r e s e r v i n g l o c a l i z e d a c t i v i t y near Cz. T h e first l o c a l i z e d w a v e o f the s c a l p - r e c o r d e d S E P is a p o s i t i v i t y w i t h a n o m i n a l p e a k l a t e n c y o f 38 m s e c , the P~8. T h i s is a s y m m e t r i c a l l y d i s t r i b u t e d a b o u t C z w i t h c o n s i s t e n t l y greater i n v o l v e m e n t o f sites i p s i l a t e r a l to the s t i m u l a t e d P T N . T h e t o p o g r a p h y o f P38 varied o n l y s l i g h t l y f r o m s u b j e c t to subject. T h e p o s i t i v i t y w a s m a x i m a l at or j u s t lateral to C z i p s i l a t e r a l to the s t i m u l a t e d leg. In the m i d s a g i t t a l
Fig. 4. Left PTN SEP in the coronal plane through Cz using Fpz as a reference. The 3 widespread early potentials have been eliminated by in-phase cancellation from Fpz. A well-defined positivity, P38 appears as a localized potential maximum at Cz.
324
M. SEYAL ET AL.
5
i 05-805
C5-Fpz
C 1 -Fp
c. ~- s.c2 Cz
905
.~-
C2,-SC5
.~
" "
_..--~-'-"
..
$05
~ "
c6-sc5
~#-
T4-SC5
~
"
2 ms~¢
@
Fig. 5. Left PTN SEP in the coronal plane through Cz using SC5 (A) and Fpz (B) references. In this subject, P38 is present in all ipsilateral (left) leads. A simultaneous negativity (N38) appears in the contralateral homologous electrodes• In this example, the N34 is clearly separated from N3s by an inflection (arrows). Notice the apparent slight asynchrony between the p~ and N38 peaks with P38 being ' delayed'.
A. Left PTN Stimulation
B. Right PTN Stimulation
1.0 .8
g
.6
4~
~"/
\
"
g
o
2,\
,
N
~
-.2
li ~.
\
--.4
-.'~,-.~'" .."°
......
//// -.8
-I.0
"~'" ""0"" "'' '~
Fig. 6. Voltage distribution of the P38/N38 peak is illustrated in the coronal plane through Cz using data obtained from stimulation of 4 left (A) and 4 right (B) legs. For purposes of comparison, plotted voltages have been normalized by assigning an arbitrary value of unity to peak absolute voltage of the P38 complex for each stimulation PTN. In each case, the positivity is maximal at or just ipsilateral to Cz, and a negativity of variable voltage is usually present contralateral to the stimulated leg.
POSTERIOR TIBIAL SEP
A
325
,
. . . . . . i uv
2 m**¢
Fig. 7. Derivation of a later, bilateral positivity. A: these two traces demonstrate the localized negativity (N38) and positivity (P38) at C~ and C~ in derivations using Fpz as a reference. The first dotted line indicates the onset of the P38/N38 complex. In B is shown the arithmetic sum of the two traces in A. For a period of 3.5 msec, there is no deviation from the baseline indicating equal activity of opposite polarity at the two recording sites. At the second dotted line (3.5 msec after onset of P38/N38), the onset of a later complex potential is seen which is initially positive at both C~ and C~.
the onset of a late potential recorded at both C~ and C,~ and use the following argument to support this hypothesis. Fig. 7A demonstrates the localized N38 and P38 potentials recorded at C~ and C'3. In this example, the peak positivity at C~ occurs 1.5 msec after the peak negativity at C~. It is apparent, however, that the onset of activity occurs simultaneously at both recording sites, and that the slopes of the deflections are symmetrical although of opposite polarity for the first 3.5 msec. In Fig. 7B is shown the arithmetic sum of activity at C'3 and C]. There is no deviation from the baseline during the first 3.5 msec after the onset of localized activity at C'3 and C~ confirming the symmetry of the potential at both electrodes during this time. Thereafter, a complex potential starts which appears to be initially positive at both C~ and C 4.
Discussion
The spinal volley plane, P38 was of greatest voltage at Cz or Cz'. In most subjects, a simultaneous negativity ( N 3 8 ) was recorded over the contralateral parasagittal scalp (Fig. 5). The amplitude of N38 relative to P38 was extremely variable not only from subject to subject, but also from leg to leg. For a given limb, however, the amplitude ratio of P38 t o N38 was constant from trial to trial. In two instances, the voltage of the N38 approached that of P38 2. The potential distribution of P38/N38 in the coronal plane through Cz is plotted in Fig. 6. In some subjects, the peak of P38 appeared slightly delayed relative t o N38 although the onset latency was identical (Fig. 5). We postulated that the asynchrony in the N 3 8 / P 3 8 peaks resulted from 2 Fpz to non-cephalic recordings were obtained in all subjects. Occasionally a small negativity occurred at Fpz concurrent with N3s. In no case. however, did it exceed 10-20% of the amplitude of P38 recorded at the ipsilateral C~ or C~ electrode. When Fpz is used as a reference, this activity, when present, results in a small apparent increase in the amplitude of P38 and a corresponding decrease in the amplitude of N38 but does not affect the analysis of the 'near-field' potentials in any meaningful way. Alternative references such as Oz or mastoids were unsuitable since the amplitude of the subcortical potentials (P2~, P31, and N34 ) decreased substantially over the more posterior and lateral scalp regions.
SEPs elicited by stimulation of peripheral nerves in the legs have been recorded from the spinal cord using electrodes on the skin surface (Cracco 1973; Delbeke et al. 1978; Jones and Small 1978) and by using intrathecal and epidural electrodes (Magladery et al. 1951; Ertekin 1976; Lueders et al. 1981 ; Jones et al. 1982). We were able to record consistent and highly reproducible responses with surface electrodes at positions spanning the entire extent of the spinal cord from the L~ spinous process to the second cervical vertebra. The onset of the negative deflection at the recording electrode represents the arrival of the afferent volley transmitted in the fastest conducting fibers (Lorente de N6 1947; Patton 1982). From Fig. 2 it is apparent that for the fastest recorded fibers, the afferent volley travels up the spinal cord at constant velocity. This was true in all subjects tested. Linear regression analysis demonstrates that the data (see Table I) closely approximate a straight line. Our data do not support the slowing of conduction velocity as the volley travels rostrally reported by Cracco (1973) and Lueders et al. (1981). Based on surface length measurements, the mean cord conduction velocity for our group was
326
66 m / s e c (S.D. 0.78 m/sec). Using intrathecal recording electrodes placed adjacent to the dorsal columns, Ertekin (1976) measured slower conduction velocities ranging from 35 to 50 m/sec. Studies in both animals (Loeb 1976) and humans (Desmedt and Cheron 1980) indicate that conduction velocity in the dorsal columns is slower than in the large myelinated fibers of peripheral nerves. Since the maximum sensory conduction velocity for the posterior tibial nerve is 56.5 m / s e c (Behse and Buchtal 1971), one would expect the cord conduction velocity to be less than this. Desmedt and Cheron (1980) showed that if surface measurements of interspinous distances are used for calculations rather than actual cord length, an erroneously high estimate of spinal cord conduction velocity is obtained. Using direct measurements of cervical cord length, they found that the conduction velocity of the afferent cord volley following median nerve stimulation was approximately 19% slower than the mean conduction velocity from wrist to spinal cord. Our surface measurements probably give an excessive estimate of actual spinal cord length. The average cord is 45 cm long in men (Truex and Carpenter 1964). From the C6/C v segment to the C 2 segment, the typical distance is 5.5 cm (Desmedt and Cheron 1980). The actual spinal cord length caudal to SC5 may therefore be estimated to be about 39.5 cm. The surface distance between L~ and SC5 in our subjects was about 44.5 cm which is 5 cm (12%) longer than the probable actual cord length. If our calculated velocity is proportionately reduced from 66 m / s e c to 58 m/sec, direct measurements are more closely approximated (Ertekin 1976).
M. SEYAL ET AL.
The onset of P28 occurs 2.9 msec (S.D. 0.7 msec) after the onset of the negativity at SC5 (Fig. 3A). Assuming that conduction velocity in the dorsal columns rostral to SC5 is about 58 m/sec, it is unlikely that P28 arises from the dorsal column nuclei since the afferent volley would be expected to arrive in the nucleus gracilis 1.2 msec after the SC5 potential (mean measured distance from C 6 / C 7 to the dorsal column nuclei = 70 mm; Desmedt and Cheron 1980). Even a synaptic delay across the n. gracilis of 0.5-0.8 msec would not account for this disparity. It is more likely that P2s results from current flow associated with activity in more rostral structures, possibly the medial lemniscus, and may therefore may be analogous to the FF4 or P~4 wave seen following median nerve stimulation (Desmedt and Cheron 1980). We were able to record simultaneously from electrodes on the scalp and electrodes stereotactically implanted in nucleus ventralis posterolateralis (VPL) and in nucleus centromedianum (CM) of the thalamus in a single patient being treated for intractable pain. Electrode position was determined by stereotactic coordinates and verified by transaxial and coronal computed tomography. From the VPL electrode (referred to a non-cephalic lead) we recorded a high voltage diphasic positive-negative potential with multiple overrid-
, 30~V I
VPL
SC6
, , ,
Subcortical potentials The earliest scalp potentials recorded using a non-cephalic reference are two positive waves, P28 and ~ , followed by a scalp negativity of higher a m p l i t u d e , N34. We were unable to identify P2s in 2 of 11 subjects, but P3~ and N34 w e r e invariably present. These early potentials have a widespread distribution over both hemispheres but are not seen over the cervical spine (SC5 or SC2) or in scalp to scalp recordings. They are, therefore, most probably 'far-field' potentials generated at subcortical levels.
Fig. 8. SEP following right PTN stimulation recorded simultaneously from the contralateral scalp and thalamus. The peak of the VPL positivity occurs simultaneously with the peak of the scalp-recorded P31. V P L nucleus ventralis posterolateralis; CM, nucleus centromedianum. Note that calibration for the VPL trace is different from the others.
POSTERIOR TIBIAL SEP ing spike-like deflections (Fig. 8). A similarly configured wave of much lower amplitude was recorded from the CM electrode. The onset of the VPL positivity in this patient occurred at 30.2 msec with the peak at 34.2 msec. The peak of the VPL positivity exactly coincided with the peak of the scalp-recorded Psi. The peak of the subsequent VPL negativity occurred 1.8 msec after the peak of the N34 recorded at the scalp. P3~ may, therefore, reflect initial activation of the thalamic ventrobasal nuclear complex. However, the scalp-recorded N34 and the principal thalamic negativity do not correspond closely enough to indicate that N34 is entirely the result of the thalamic negativity (Emerson et al. in preparation). Allison and Hume ( 1981) suggested that the initial positivity recorded from VPL reflects the lemniscal volley arriving at the thalamus with the aftergoing negativity representing summated postsynaptic potentials. Desmedt and Cheron (1981b) described a 'far-field' N~s following median nerve stimulation that they speculated might arise in thalamus or thalamocortical radiations. It is likely that our N34 is the PTN analog of this. Using a midfrontal reference, Jones and Small (1978) observed a 'negative' potential of fairly constant amplitude over the spine between T 3 and C 2 with no apparent shift in peak latency. This potential was not recorded using a non-cephalic reference and presumably represents P31 seen at Fz but recorded via the inverting input of the amplifier. Using an SC5-Cz montage, Lueders et al. (1981) described a series of waves widely distributed over the scalp which they attributed to 'far-field' subcortically generated events. Our subcortical components P3~ and N34 probably correspond to their P27 and N30 waves (compare their Fig. 1 with our Fig. 3). Lueders et al. (1981), however, described an N24 wave that appears similar to our P28 in both configuration and relationship to subsequent components. They suggested that this potential arises from the afferent volley in the cervical cord. This conclusion was based largely on a non-linear spinal cord conduction velocity. In our subjects, cord velocity remained constant. In addition, we demonstrated an earlier spinal potential localized to SC5 corresponding to the afferent sensory vol-
327 ley at that level (Fig. 3A). Furthermore, our Pz8 was seen equally well in scalp to elbow or shoulder derivations (Fig. 3A) and in scalp to SC5 recordings. It was not present on SC5 to shoulder derivations. Therefore, we view the P28 as a true positivity recorded from the scalp electrode rather than a negativity recorded at the SC5 electrode. Kakigi et al. (1982) recorded a P:5 wave using a non-cephalic (hand) reference which appears to be identical to our P28 and the N24 of Lueders et al. (1981).
The cortical potentials Following N34 , a large localized positivity, ~8, was consistently recorded from Cz and adjacent sites ipsilateral to the stimulated leg in all subjects. In most individuals, a localized negativity, N38, occurred synchronously over the homologous contralateral scalp. The amplitude of the contralateral N38 relative to the ipsilateral P~8 was variable from person to person and indeed varied in the same individual depending on which leg was stimulated. Cruse et al. (1982) demonstrated a similar potential distribution for the P2 wave (peak latency 38 msec) and a late N 2 (peak latency 55 msec). The P38/N38 probably corresponds to the N20 component of the median SEP. In some recordings using non-cephalic references (Fig. 5A, arrow), there was a clear inflection marking the onset of contralateral N38 negativity and distinguishing it from the preceding N34. This appears to be analogous to the inflection in the contralateral negativity which marks the onset of the N20 and differentiates it from the preceeding N~s following median nerve stimulation (Desmedt and Cheron 1981b). Initial activation of the somatosensory cortex in mammals including primates results in a stereotyped initial surface positivity followed by a surface-negative response (Perl and Whitlock 1955). Localized cellular events occurring in deeper cortical layers characteristically result in a surface potential of opposite polarity. This polarity inversion results from the vertical orientation of pyramidal cells and the consequent large extracellular current flows occurring perpendicular to the surface which behave as virtual dipoles. Depolarization of cell bodies in layer IV (site of termination of specific thalamocortical relay afferent fibers) creates an active sink in the cortical
328 depths and a corresponding passive current source at the cortical surface. This is seen by a superficial or scalp electrode as a localized positivity (Goff et al. 1978). Since the primary sensory area for the leg and foot is located on the mesial aspect of the postcentral gyrus within the interhemispheric fissure, the initially positive primary cortical response representing depolarization of pyramidal cell bodies would be seen principally by electrodes located ipsilateral to the stimulated limb (Fig. 9). Slight differences in the location of the leg area (Penfield and Rasmussen 1950) would result in considerable variation in field of the scalp-recorded potential. If the leg area is located at the superficial edge of the interhemispheric fissure, the dipole orientation would be vertical or oblique at the scalp (Fig. 9A). On the other hand, if the leg area lies deeper in the interhemispheric fissure, the dipole orientation would be more horizontal with the initial positivity again projecting ipsilateral to the side of stimulation, but with the negative end of the dipole projecting contralaterally (Fig. 9B). In our view, ~8 and N3s simply represent two ends of a single dipole with variable orientation in the vertical plane. In our group of subjects, we observed a wide range of variation in P3JN38 distribution that would be expected from differences in dipole orientation (see also Fig. 6). Similar results have been reported by Cruse et al. (1982). Unlike early components of the SEP, potentials occurring after P38 varied considerably from trial to trial even within the same subject (see for example response at Cz in Fig. 3A). This variation appeared to be related to the subject's level of alertness but this was not studied systematically. In some individuals the peaks of P38 and N38 did not occur simultaneously: the peak of P38 appeared to be slightly delayed relative to N38. We hypothesize that this apparent asynchrony is caused by an additional component which begins about 3.5 msec after the onset of P3s/N38 and is seen as an initial positivity at both C~ and C~. We propose that this later potential is generated by a different neuronal population, perhaps the result of intracortical spread to secondary cortical areas. Soon after the onset of the primary cortical potentials (p~8/N3s), other cortical areas may be
M. SEYAL ET AL.
÷
A
4"
B
Y Fig. 9. Hypothesizedvariations in the field of distribution of the P38/N3spotential based upon the known anatomic variability of the location of the leg area. See text for further details.
activated resulting in a complex summation of potentials at the scalp. The scalp distribution and nature of these later components are presently under investigation. We conclude that stable, reproducible potentials can be recorded over the spine and scalp following stimulation of the PTN. Clinical application of these SEPs requires knowledge of the normal topography and variability of these potentials, and an understanding of their probable neural generators. The following montage might be considered for routine clinical purposes using a 4-
POSTERIOR T1BIALSEP channel evoked potential system: L1 spine to an electrode 10 cm above Channel 2: Fpz referred to SC5 Channel 3: C 3 or C a (ipsilateral to stimulated leg) referred to Fpz Channel 4: Cz referred to Fpz. Channel 1 :
This montage allows identification of the peripheral volley as it enters the spinal cord (channel 1), demonstrates the subcortical potentials (channel 2) and isolates the localized cortical responses (channels 3 and 4). Both channels 3 and 4 are necessary given the normal range of variability in the potential field of the P38/N38 and later waves. Some degree of sedation is required in most individuals to reduce muscle activity sufficiently to permit interpretable recording of the subcortical potentials.
Summary Somatosensory evoked potentials (SEPs) were elicited by stimulation of the posterior tibial nerve (PTN) in 12 normal adults. Recording using both ceptialic and non-cephalic references were obtained from multiple electrodes placed over the spine and scalp. Following PTN stimulation, the fastest recorded potentials of the afferent sensory volley proceeds up the spinal cord at constant velocity. After arrival of the volley at cervical cord levels, 3 widely distributed waves, P28, Pal and N34, are recorded from scalp electrodes. These 'far-field' potentials are followed by a localized positivity (P38) which has a peak voltage either at the vertex or just laterally toward the side of stimulation. A contralateral negativity (N38) was present in most individuals. We propose that P28 arises from medial lemniscus; that P31 is generated by ventrobasal thalamus; and that N34 is probably the result of further activity in thalamus a n d / o r thalamocortical radiations. The P38/N38 complex represents the primary cortical response to PTN stimulation. Its most consistent characteristic is a positivity at the vertex or immediately adjacent scalp areas ipsilateral to the stimulated leg. The
329 topography of the P38/N38potential varies slightly from individual to individual in a manner consistent with a functional dipole situated in the leg and foot area on the mesial aspect of the postcentral gyrus, whose exact location and orientation changes in accordance with known variations in the location of the leg area. R6sum6
Potentiels dvoquks somatosensoriels a la stimulation du nerf tibial postdrieur." composantes spinales et composantes du scalp Des potentiels 6voqu6s somatosensoriels (PES) ont 6t6 produits par stimulation du nerf tibial post6rieur (NTP) chez 12 adultes normaux. Des enregistrements utilisant des r6f6rences h la fois c6phaliques et non c6phaliques ont 6t6 obtenus partir d'61ectrodes multiples plac6es sur la moelle 6pini6re et le scalp. Apr6s stimulation du NTP, le plus rapide des potentiels de la vol6e aff6rente sensorielle remonte la moelle +pini6re ~ vitesse constante. Apr6s l'arriv6e de la vol6e au niveau de la moelle cervicale, 3 ondes ~t large distribution, P28, P31 et N34 sont recueillies par les 61ectrodes de scalp. Ces potentiels de champs lointains sont suivis par une positivit6 P38 localis~e qui a un voltage maximal soit au vertex, soit juste lat6ralement du c6t6 de la stimulation. Une n6gativit6 contralat+rale N38 6tait pr6sente chez la plupart des individus. Nous proposons que P28 a son origine dans le lemnisque m6dian, que P31 provient du thalamus ventrobasal et que N34 r6sulte probablement d'une activit6 plus tardive du thalamus e t / o u des radiations thalamo-corticales. Le complexe P38/N38 repr6sente la r6ponse corticale primaire h la stimulation du NTP. Sa principale caract6ristique est une positivitb au vertex ou sur le scalp imm6diatement ipsilat6ral h la jambe stimul6e. La topographie du potentiel P38/N38 varie l+g~rement d'une individu ~ l'autre; ces r6sultats sont compatibles avec le mod61e d'un dip61e fonctionnel situ6 sur l'aire de la jambe et du pied de la face m+diane du gyrus post-central, et dont la localisation et l'orientation changerait comme on sait que varie la localisation de l'aire de la jambe.
330
References Allison, T. and Hume, A.L. A comparative analysis of shortlatency somatosensory evoked potentials in man, monkey, cat and rat. Exp. Neurol., 1981, 72:592-611. Arrezo, J.C., Vaughn, H.G. and Legatt, A.D. Topography and intracranial sources of somatosensory evoked potentials in the monkey. II. Cortical components. Electroenceph. clin. Neurophysiol., 1981, 51: 1-18. Behse, F. and Buchthal, F. Normal sensory conduction in the nerves of the leg in man. J. Neurol. Neurosurg. Psychiat., 1971, 34: 404-414. Cracco, R.Q. Spinal evoked response: peripheral nerve stimulation in man. Electroenceph. clin. Neurophysiol., 1973, 35: 379 386. Cruse. R., Klem, G., Lesser, R. and Lueders, H. Paradoxical lateralization of cortical potentials evoked by stimulation of posterior tibial nerve. Arch. Neurol. (Chic.), 1982, 39: 222-225. Delbeke, J., McComas, A.J. and Kopec, S.J. Analysis of evoked lumbosacral potentials in man. J. Neurol. Neurosurg. Psychiat., 1978, 41: 293-302. Desmedt, J.E. and Brunko, E. Functional organization of farfield and cortical components of somatosensory evoked potentials in normal adults. Prog. clin. Neurophysiol., 1980, 17: 27-50. Desmedt, J.E. and Cheron+ G. Central somatosensory conduction in man: neural generators and interpeak latencies of far-field components recorded from neck and right or left scalp and earlobes. Electroenceph. clin. Neurophysiol., 1980, 50: 382-390. Desmedt, J.E. and Cheron, G. Prevertebral (oesophageal) recording of subcortical somatosensory evoked potentials in man: the spinal PI3 component and the dual nature of spinal generators. Electroenceph. clin. Neurophysiol., 1981 a, 52: 257-275. Desmedt, J.E. and Cheron, G. Non-cephalic reference recording of early somatosensory potentials to finger stimulation in adult or aging normal man: differentiation of widespread Now and contralateral N20 from the pre-rolandic P22 and N3o components. Electroenceph. clin. Neurophysiol., 1981 b, 52: 553-570. Ertekin, C. Studies of the h u m a n evoked spinogram. II. The conduction velocity along the dorsal funiculus. Acta neurol. scand., 1976, 53: 21-38. Goff, W.R., Allison, T. and Vaughn, H.G. Functional neuroanatomy of event-related potentials. In: E. Callaway, P. Tueting and S.H. Koslow (Eds.), Event-related Potentials in Man. Academic Press, New York, 1978: 1-79. Jones, SJ. and Small, D.G. Spinal and subcortical evoked potentials following stimulation of the posterior tibial nerve in man. Electroenceph. clin. Neurophysiol., 1978, 44: 299-306.
M. SEYAL ET A U Jones, S.J., Edgar, M.A. and Ransford, A.O. Sensory nerve conduction in the h u m a n spinal cord: epidural recordings made during scoliosis surgery. J. Neurol. Neurosurg. Psychiat., 1982, 45: 446-451. Kakigi, R., Shibasaki, H., Hashizume, A. and Kuroiwa, Y. Short latency somatosensory evoked spinal and scalp-recorded potentials following posterior tibial nerve stimulation in man. Electroenceph. clin. Neurophysiol., 1982, 53: 602-611. Loeb, G.E. Decreased conduction in the proximal projections of myelinated dorsal root ganglion cells in the cat. Brain Res., 1976, 103: 381-385. Lorento de N6, R. A study of nerve physiology. XVI. Analysis of the distribution of the action currents of nerve in volume conductors. Stud. Rockefeller Inst. reed. Res., 1947, 132: 384-482. Lueders, H., Andrish, J., Gurd, A., Weiker, G. and Klem, G. Origin of far-field subcortical potentials evoked by stimulation of the posterior tibial nerve. Electroenceph. clin. Neurophysiol., 1981, 52: 336-344. Magladery, J.W., Porter, W.E., Park, A.M. and Teasdall, R.D. Electrophysiological studies of nerve and reflex activity in normal man. IV. The two-neurone reflex and identification of certain action potentials from spinal roots and cord. Bull. Johns Hopk. Hosp., 1951, 88: 499-519. Noel, P. and Desmedt, J.E. Cerebral and far-field somatosensory evoked potentials in neurological disorders involving the cervical spinal cord, brainstem, thalamus and cortex. Prog. clin. Neurophysiol., 1980, 7: 205-230. Patton, H.D. Special properties of nerve trunks and tracts. In: T.R. Ruch and H.D. Patton (Eds.), Physiology and Biophysics, Vol. IV. Saunders, Philadelphia, Pa., 1982: 104-124. Penfield, W. and Rasmussen, T. The Cerebral Cortex of Man. A Clinical Study of Localization of Function. Macmillan, New York, 1950. Perl, E.R. and Whitlock, D.G. Potentials evoked in cerebral somatosensory region. J. Neurophysiol., 1955, 18: 486-501. Rossini, M.P., Caracco, R.Q., Cracco, J.B. and House, W.J. Short latency somatosensory evoked potentials to peroneal nerve stimulation: scalp topography and the effect of different frequency filters. Electroenceph. olin. Neurophysiol., 1981, 52: 540-552. Truex, R.C. and Carpenter, M.B. Strong and Elwyn's H u m a n Neuroanatomy. Williams and Wilkins, Baltimore, Md., 1964. Tsumoto, T., Hirose, N., Nonaka, S. and Takahashi+ M. Analysis of somatosensory evoked potentials to lateral popliteal nerve stimulation in man. Electroenceph. clin. Neurophysiol., 1972, 33: 379-388. Vas, G.A.,Cracco, J.B. and Cracco, R.Q. Scalp recorded short latency cortical and subcortical somatosensory evoked potentials to peroneal nerve stimulation. Electroenceph. clin. Neurophysiol., 1981, 52: 1-8.
Electroencephalography and clinical Neurophysiologv, 1983, 55 : 320 330 Elsevier Scientific Publishers Ireland, Ltd.
S P I N A L AND EARLY S C A L P - R E C O R D E D C O M P O N E N T S OF T H E S O M A T O S E N S O R Y EVOKED P O T E N T I A L F O L L O W I N G S T I M U L A T I O N OF T H E P O S T E R I O R TIBIAL NERVE M A S U D SEYAL J, R O N A L D G. EMERSON and T I M O T H Y A. PEDLEY
EEG Systems Laboratory for Special Studies, The Neurological Institute, Columbia University College of Pt~vsicians and Surgeons, New York, N. E 10032 (U.S.A.) (Accepted for publication: November 2, 1982)
Recent reports have characterized the human early somatosensory evoked potential (SEP) elicited by stimulation of the median nerve and have provided clues to its neural generators (Desmedt and Brunko 1980; Desmedt and Cheron 1980; 1981a, b; Noel and Desmedt 1980; Allison and Hume 1981). The early SEP following stimulation of peripheral nerves in the lower extremity has been less extensively studied and its neural generators are correspondingly less certain (Tsumoto et al. 1972; Lueders et al. 1981; Rossini et al. 1981; Vas et al. 1981; Cruse et al. 1982; Kakigi et al. 1982). Furthermore, there is no consensus to date about such fundamental characteristics of the lower limb SEP as number and polarity of the early components, scalp topography of the response, and preferred recording techniques. Indeed, it is likely that variations in recording methods such as use of different montages, filter settings, stimulus rates, and in one instance general anesthesia, have contributed substantially to lack of uniform results. We now present a detailed analysis of the early somatosensory evoked potential following stimulation of the posterior tibial nerve (PTN) in the normal adult. Recordings were made from multiple electrodes over the spine and scalp using both Correspondence to: Timothy A. Pedley, M.D., The Neurological Institute, 710 West 168th St., New York, N.Y. 10032, U.S.A. Presented in part at the annual meeting of the American EEG Society in Phoenix, Ariz., November 11-13, 1982. Present address: Dept. of Neurology, University of California Medical Center, 2315 Stockton Blvd, Sacramento, Calif. 95817, U.S.A.
cephalic and non-cephalic reference electrodes. Our results clearly delineate both localized and widespread components of the scalp-recorded potentials and permit conclusions to be drawn regarding their possible neural generators. Materials and Methods
SEPs were recorded following stimulation of 22 legs in 12 normal adult volunteers (11 men) ranging in age from 18 to 42 years. The study was approved by the Institutional Review Board of Columbia-Prebyterian Medical Center and informed consent was obtained from participants. Volunteers lay on a hospital bed in a quiet, darkened room and were given oral diazepam to promote muscle relaxation and sleep. A typical recording session lasted up to 8 h, and several persons were studied on 2 or 3 different occasions. Stimuli were 0.2 msec square wave electrical pulses applied through electrodes placed over the posterior tibial nerve just behind the medial malleolus. The cathode was located 2 cm proximal to the anode. Stimulus intensity was adjusted to elicit a consistent twitch of the great toe that was not uncomfortable to the subject. The stimulation rate was 4/sec. Four recording electrodes were placed over the thoracic and lumbar spine spaced at 10 cm intervals beginning at the first lumbar spinous process (L~). Electrodes were placed over the fifth and second cervical spines (SC5 and SC2). Scalp electrodes were positioned according to the International 10-20 system. Further resolution was obtained by using additional electrodes located mid-
0013-4649/83/0000-0000/$03.00 © 1983 Elsevier Scientific Publishers Ireland, Ltd.
POSTERIOR TIBIAL SEP
321
way between standard sagittal placements. These are identified by a (') following the name of the anterior most neighboring conventional electrode (Fig. 3A inset). For example, Cz' refers to an electrode located midway between Cz and Pz. Electrodes over the shoulder, elbow and SC5 were used as non-cephalic references. Simultaneous 8-channel recordings were obtained using amplifiers with a gain of 100,000 and half-frequency high and low pass filters set to 30 Hz and 3 kHz, respectively. Signal averaging was performed by a computer with an analog to digital converter operating in an 8-channel multiplexed mode obtaining 10,000 samples/sec for each channel. Signals were averaged for 50 msec following each stimulus, and an average of 1000-4000 samples comprised each trial. Samples with muscle artifact were automatically excluded. Signal averaging was repeated at least once to ensure stability and reproducibility of the response. No curve smoothing was used. Evoked potential wave forms were plotted on a hard-copy device and also stored by the computer for further analysis. The various waves of the scalp SEP were labeled according to their positive (P) and negative (N) polarities at the presumed 'active' electrode and l 1-L1+10
L1+10
Ll+20
L: .......
SC5
SC2
~'~
f
\
\ G1 NEG UP O 5~,V
J
2 ~,.,
Fig. l. Spinal volley following stimulation of the right PTN. Vertical spacing is proportional to the interelectrode distance. Arrows indicate the onset of negativitycorresponding to arrival of the depolarizing wave front at the lower electrode of each recording pair. Electrode positions over lumbar and thoracic spine are specified as distance in centimeters rostral to L~. Two separate trials (dotted lines) and the average (solid lines) are shown.
given nominal latencies reflecting the mean latency for each peak in our group of subjects. Since absolute latencies have been used, there is considerable variation in latency measurements reflecting primarily the different heights of our volunteers. Results
The spinal volley In 10 individuals, recordings were obtained between successive electrodes at L], at 10, 20 and 30 cm above L], over SC5 and over SC2. In one person, additional recordings were made using SC5 as a reference for each of the other spinal electrodes. Fig. 1 illustrates the typical response. Vertical spacing in the figure is proportional to the interelectrode distances on the back. The wave form of the spinal volley was usually triphasic with a small initial positive deflection followed by a major negative wave and a subsequent positivity as seen by the lower electrode in a bipolar chain. The amplitude of the spinal volley was maximal over the lower thoracic region and gradually decreased over the upper thoracic and cervical areas. This was observed both in bipolar and referential recordings. There was a gradual increase in the latency of the response rostral from L~. For each subject, the onset latency of the initial negativity (Fig. 1, arrows) was measured at the various spinal levels. This measurement corresponds to the arrival of the leading edge of the depolarizing wave front of the afferent volley at the distalmost electrode of the recording pair. These values are listed in Table I along with the results of least squares linear regression analysis of the arrival time of the negativity versus distance from L]. A linear relationship between these two variables is demonstrated. This is shown graphically in Fig. 2 in which distance from LI of the distalmost electrode in each recording pair is plotted against arrival time of the afferent volley for 3 representative subjects.
The scalp-recorded potential Three early components were present with widespread, although not entirely uniform, distribution over the scalp in both sagittal and coronal planes (Fig. 3). They consisted of two initial small
322
M. S E Y A L ET AL.
40 30 5
E .o
20
O o
I0
~5 0
,o ,', ,'2
,'5 16 ,'7
2'od,
Latency (msec) Fig. 2. Onset latency of the spinal volley is plotted against distance of the recording electrode from L 1 for 3 representative subjects (subjects 1, 2 and 7 in Table I). The close fit of the original data to the regression lines (solid lines) is apparent. For clarity of presentation, only 3 sets of data from the table are plotted.
positive waves designated P28 and P31 with peak latencies of 27.6 + 1.9 msec (1 S.D.) and 31.2 + 1.9 msec respectively followed by a larger amplitude n e g a t i v i t y , N34 (peak latency 33.9 + 1.9 msec). A t recording sites away from the vertex, N34 showed a
gradual return to the baseline approximately 40 msec after the stimulus. These early potentials were seen only using non-cephalic reference electrodes since in-phase cancellation eliminated them from scalp to scalp recordings. In two individuals, several additional smaller but highly reproducible responses were seen between the onset of P2s and the peak of N34. Because these deflections were not seen in most persons, they will not be discussed further. They may, however, correspond to the multiple early potentials described by Rossini et al. (1981). At Cz and adjacent areas ipsilateral to the stimulus, the negative deflection of N34 was terminated by the onset of a large positive wave, P3s. This was followed by a major negativity of variable latency peaking about 45 msec after the stimulus. Unlike P28, P~l and N 3 4 , P38 had a restricted field involving primarily the central parasagittal region. To study the localized components of the SEP further, scalp-to-scalp recordings were obtained. Fpz was used as a reference since non-cephalic recordings had demonstrated that it was virtually inactive with respect to potentials occurring after 34 msec (Fig. 3B). Fig. 4 shows the nearly total in-phase cancellation of the 3 widespread early
TABLE I
Spinal volley following stimulation of P T N . No.
PT~
Onset latency (msec)
Length
stimulated
1 2 3 4 5 6 7 8 9 10
L L L L R L R L R R
(cm) LI
L I + 10cm
L I + 20cm
LI + 30crn
SC 5
L I -SC~
17.5 18.8 18.8 17.6 17.4 17.2 14.8 17.5 18.4 21.8
19.6 20.2 20.5 19.2 19.1 19.0 15.8 19.2 20.0
20.2 22.0 23.0 20.6 20,6 21.3 17.8 20.1 21.3
21.7 23.6
24.2 25.0 24.7 22.8 24.0 24.1 20.6 25.4 24.0 27.8
45 45 45 44 43 44 44 38 46 43
* Slope of least squares linear regression line. ** Correlation coefficient. *** P r o b a b i l i t y ( P ) .
not recorded
23.0 23.2 19.5 23.0
Slope * (m/see)
r **
69 69 69 82 62 58 62 56 68
0.991 0.993 0.992 0.991 0.997 0.979 0.996 0.990 0.999
P ***
tess than
0.00 t 0.001 0.05 0.001 0.001 0.001 0.001 0.001 0.001
POSTERIOR TIBIAL SEP
323 F p z-SC 5
I
+
.
~.
"~-~.
.
Fz'-SC5
\, ~, '/ ~l' c, cz c z C4c6, ~4 I~z'r~c. z~c~ c~ c, c,c/c c' FZ
V7
T4-8C5
I~
C ~4 ~-~_ C2-8C5
S
C
5
~
/
Cl-SC5 C3-8C5
p
~ Pz'-SC5
T3-SC5 C1-SHOULDER C3-SHOULDER
•
pP~¢~.=,
' ~"
",-.
A
vt
Jl.v
~
2
mJ~c
Fig. 3. SEP to left PTN stimulation. A: the top channel, SC5 to shoulder reference, illustrates the arrival of the spinal volley at SC5 (arrow). The next 8 channels show the scalp-recorded SEP in a coronal plane through Cz using an SC5 reference. The last two channels show the activity at scalp electrodes C] and C 3 using a shoulder reference. Three peaks, P2~, P31 and N34, are present in all channels except the first indicating their widespread distribution over the scalp. Each of these peaks is also present in scalp leads connected to a shoulder reference (last two channels). The cephalic origin of these potentials is confirmed by their absence from the SC5-shoulder derivation. Note the different calibration in channel 1. B: scalp SEP following left PTN stimulation recorded in the midline sagittal plane using an SC5 reference. As in A, P28, P~l and N~4 are widely distributed. Fpz is virtually inactive for events occurring after 35 msec.
T4 Fpz C4-Fpz C2-Fpz
Cz-Fpz
~ .
~" ~"
CS-Fpz T3-Fpz
°5,7
p o t e n t i a l s that o c c u r s w i t h s c a l p - t o - s c a l p recordi n g s w i t h p r e s e r v i n g l o c a l i z e d a c t i v i t y near Cz. T h e first l o c a l i z e d w a v e o f the s c a l p - r e c o r d e d S E P is a p o s i t i v i t y w i t h a n o m i n a l p e a k l a t e n c y o f 38 m s e c , the P~8. T h i s is a s y m m e t r i c a l l y d i s t r i b u t e d a b o u t C z w i t h c o n s i s t e n t l y greater i n v o l v e m e n t o f sites i p s i l a t e r a l to the s t i m u l a t e d P T N . T h e t o p o g r a p h y o f P38 varied o n l y s l i g h t l y f r o m s u b j e c t to subject. T h e p o s i t i v i t y w a s m a x i m a l at or j u s t lateral to C z i p s i l a t e r a l to the s t i m u l a t e d leg. In the m i d s a g i t t a l
Fig. 4. Left PTN SEP in the coronal plane through Cz using Fpz as a reference. The 3 widespread early potentials have been eliminated by in-phase cancellation from Fpz. A well-defined positivity, P38 appears as a localized potential maximum at Cz.
324
M. SEYAL ET AL.
5
i 05-805
C5-Fpz
C 1 -Fp
c. ~- s.c2 Cz
905
.~-
C2,-SC5
.~
" "
_..--~-'-"
..
$05
~ "
c6-sc5
~#-
T4-SC5
~
"
2 ms~¢
@
Fig. 5. Left PTN SEP in the coronal plane through Cz using SC5 (A) and Fpz (B) references. In this subject, P38 is present in all ipsilateral (left) leads. A simultaneous negativity (N38) appears in the contralateral homologous electrodes• In this example, the N34 is clearly separated from N3s by an inflection (arrows). Notice the apparent slight asynchrony between the p~ and N38 peaks with P38 being ' delayed'.
A. Left PTN Stimulation
B. Right PTN Stimulation
1.0 .8
g
.6
4~
~"/
\
"
g
o
2,\
,
N
~
-.2
li ~.
\
--.4
-.'~,-.~'" .."°
......
//// -.8
-I.0
"~'" ""0"" "'' '~
Fig. 6. Voltage distribution of the P38/N38 peak is illustrated in the coronal plane through Cz using data obtained from stimulation of 4 left (A) and 4 right (B) legs. For purposes of comparison, plotted voltages have been normalized by assigning an arbitrary value of unity to peak absolute voltage of the P38 complex for each stimulation PTN. In each case, the positivity is maximal at or just ipsilateral to Cz, and a negativity of variable voltage is usually present contralateral to the stimulated leg.
POSTERIOR TIBIAL SEP
A
325
,
. . . . . . i uv
2 m**¢
Fig. 7. Derivation of a later, bilateral positivity. A: these two traces demonstrate the localized negativity (N38) and positivity (P38) at C~ and C~ in derivations using Fpz as a reference. The first dotted line indicates the onset of the P38/N38 complex. In B is shown the arithmetic sum of the two traces in A. For a period of 3.5 msec, there is no deviation from the baseline indicating equal activity of opposite polarity at the two recording sites. At the second dotted line (3.5 msec after onset of P38/N38), the onset of a later complex potential is seen which is initially positive at both C~ and C~.
the onset of a late potential recorded at both C~ and C,~ and use the following argument to support this hypothesis. Fig. 7A demonstrates the localized N38 and P38 potentials recorded at C~ and C'3. In this example, the peak positivity at C~ occurs 1.5 msec after the peak negativity at C~. It is apparent, however, that the onset of activity occurs simultaneously at both recording sites, and that the slopes of the deflections are symmetrical although of opposite polarity for the first 3.5 msec. In Fig. 7B is shown the arithmetic sum of activity at C'3 and C]. There is no deviation from the baseline during the first 3.5 msec after the onset of localized activity at C'3 and C~ confirming the symmetry of the potential at both electrodes during this time. Thereafter, a complex potential starts which appears to be initially positive at both C~ and C 4.
Discussion
The spinal volley plane, P38 was of greatest voltage at Cz or Cz'. In most subjects, a simultaneous negativity ( N 3 8 ) was recorded over the contralateral parasagittal scalp (Fig. 5). The amplitude of N38 relative to P38 was extremely variable not only from subject to subject, but also from leg to leg. For a given limb, however, the amplitude ratio of P38 t o N38 was constant from trial to trial. In two instances, the voltage of the N38 approached that of P38 2. The potential distribution of P38/N38 in the coronal plane through Cz is plotted in Fig. 6. In some subjects, the peak of P38 appeared slightly delayed relative t o N38 although the onset latency was identical (Fig. 5). We postulated that the asynchrony in the N 3 8 / P 3 8 peaks resulted from 2 Fpz to non-cephalic recordings were obtained in all subjects. Occasionally a small negativity occurred at Fpz concurrent with N3s. In no case. however, did it exceed 10-20% of the amplitude of P38 recorded at the ipsilateral C~ or C~ electrode. When Fpz is used as a reference, this activity, when present, results in a small apparent increase in the amplitude of P38 and a corresponding decrease in the amplitude of N38 but does not affect the analysis of the 'near-field' potentials in any meaningful way. Alternative references such as Oz or mastoids were unsuitable since the amplitude of the subcortical potentials (P2~, P31, and N34 ) decreased substantially over the more posterior and lateral scalp regions.
SEPs elicited by stimulation of peripheral nerves in the legs have been recorded from the spinal cord using electrodes on the skin surface (Cracco 1973; Delbeke et al. 1978; Jones and Small 1978) and by using intrathecal and epidural electrodes (Magladery et al. 1951; Ertekin 1976; Lueders et al. 1981 ; Jones et al. 1982). We were able to record consistent and highly reproducible responses with surface electrodes at positions spanning the entire extent of the spinal cord from the L~ spinous process to the second cervical vertebra. The onset of the negative deflection at the recording electrode represents the arrival of the afferent volley transmitted in the fastest conducting fibers (Lorente de N6 1947; Patton 1982). From Fig. 2 it is apparent that for the fastest recorded fibers, the afferent volley travels up the spinal cord at constant velocity. This was true in all subjects tested. Linear regression analysis demonstrates that the data (see Table I) closely approximate a straight line. Our data do not support the slowing of conduction velocity as the volley travels rostrally reported by Cracco (1973) and Lueders et al. (1981). Based on surface length measurements, the mean cord conduction velocity for our group was
326
66 m / s e c (S.D. 0.78 m/sec). Using intrathecal recording electrodes placed adjacent to the dorsal columns, Ertekin (1976) measured slower conduction velocities ranging from 35 to 50 m/sec. Studies in both animals (Loeb 1976) and humans (Desmedt and Cheron 1980) indicate that conduction velocity in the dorsal columns is slower than in the large myelinated fibers of peripheral nerves. Since the maximum sensory conduction velocity for the posterior tibial nerve is 56.5 m / s e c (Behse and Buchtal 1971), one would expect the cord conduction velocity to be less than this. Desmedt and Cheron (1980) showed that if surface measurements of interspinous distances are used for calculations rather than actual cord length, an erroneously high estimate of spinal cord conduction velocity is obtained. Using direct measurements of cervical cord length, they found that the conduction velocity of the afferent cord volley following median nerve stimulation was approximately 19% slower than the mean conduction velocity from wrist to spinal cord. Our surface measurements probably give an excessive estimate of actual spinal cord length. The average cord is 45 cm long in men (Truex and Carpenter 1964). From the C6/C v segment to the C 2 segment, the typical distance is 5.5 cm (Desmedt and Cheron 1980). The actual spinal cord length caudal to SC5 may therefore be estimated to be about 39.5 cm. The surface distance between L~ and SC5 in our subjects was about 44.5 cm which is 5 cm (12%) longer than the probable actual cord length. If our calculated velocity is proportionately reduced from 66 m / s e c to 58 m/sec, direct measurements are more closely approximated (Ertekin 1976).
M. SEYAL ET AL.
The onset of P28 occurs 2.9 msec (S.D. 0.7 msec) after the onset of the negativity at SC5 (Fig. 3A). Assuming that conduction velocity in the dorsal columns rostral to SC5 is about 58 m/sec, it is unlikely that P28 arises from the dorsal column nuclei since the afferent volley would be expected to arrive in the nucleus gracilis 1.2 msec after the SC5 potential (mean measured distance from C 6 / C 7 to the dorsal column nuclei = 70 mm; Desmedt and Cheron 1980). Even a synaptic delay across the n. gracilis of 0.5-0.8 msec would not account for this disparity. It is more likely that P2s results from current flow associated with activity in more rostral structures, possibly the medial lemniscus, and may therefore may be analogous to the FF4 or P~4 wave seen following median nerve stimulation (Desmedt and Cheron 1980). We were able to record simultaneously from electrodes on the scalp and electrodes stereotactically implanted in nucleus ventralis posterolateralis (VPL) and in nucleus centromedianum (CM) of the thalamus in a single patient being treated for intractable pain. Electrode position was determined by stereotactic coordinates and verified by transaxial and coronal computed tomography. From the VPL electrode (referred to a non-cephalic lead) we recorded a high voltage diphasic positive-negative potential with multiple overrid-
, 30~V I
VPL
SC6
, , ,
Subcortical potentials The earliest scalp potentials recorded using a non-cephalic reference are two positive waves, P28 and ~ , followed by a scalp negativity of higher a m p l i t u d e , N34. We were unable to identify P2s in 2 of 11 subjects, but P3~ and N34 w e r e invariably present. These early potentials have a widespread distribution over both hemispheres but are not seen over the cervical spine (SC5 or SC2) or in scalp to scalp recordings. They are, therefore, most probably 'far-field' potentials generated at subcortical levels.
Fig. 8. SEP following right PTN stimulation recorded simultaneously from the contralateral scalp and thalamus. The peak of the VPL positivity occurs simultaneously with the peak of the scalp-recorded P31. V P L nucleus ventralis posterolateralis; CM, nucleus centromedianum. Note that calibration for the VPL trace is different from the others.
POSTERIOR TIBIAL SEP ing spike-like deflections (Fig. 8). A similarly configured wave of much lower amplitude was recorded from the CM electrode. The onset of the VPL positivity in this patient occurred at 30.2 msec with the peak at 34.2 msec. The peak of the VPL positivity exactly coincided with the peak of the scalp-recorded Psi. The peak of the subsequent VPL negativity occurred 1.8 msec after the peak of the N34 recorded at the scalp. P3~ may, therefore, reflect initial activation of the thalamic ventrobasal nuclear complex. However, the scalp-recorded N34 and the principal thalamic negativity do not correspond closely enough to indicate that N34 is entirely the result of the thalamic negativity (Emerson et al. in preparation). Allison and Hume ( 1981) suggested that the initial positivity recorded from VPL reflects the lemniscal volley arriving at the thalamus with the aftergoing negativity representing summated postsynaptic potentials. Desmedt and Cheron (1981b) described a 'far-field' N~s following median nerve stimulation that they speculated might arise in thalamus or thalamocortical radiations. It is likely that our N34 is the PTN analog of this. Using a midfrontal reference, Jones and Small (1978) observed a 'negative' potential of fairly constant amplitude over the spine between T 3 and C 2 with no apparent shift in peak latency. This potential was not recorded using a non-cephalic reference and presumably represents P31 seen at Fz but recorded via the inverting input of the amplifier. Using an SC5-Cz montage, Lueders et al. (1981) described a series of waves widely distributed over the scalp which they attributed to 'far-field' subcortically generated events. Our subcortical components P3~ and N34 probably correspond to their P27 and N30 waves (compare their Fig. 1 with our Fig. 3). Lueders et al. (1981), however, described an N24 wave that appears similar to our P28 in both configuration and relationship to subsequent components. They suggested that this potential arises from the afferent volley in the cervical cord. This conclusion was based largely on a non-linear spinal cord conduction velocity. In our subjects, cord velocity remained constant. In addition, we demonstrated an earlier spinal potential localized to SC5 corresponding to the afferent sensory vol-
327 ley at that level (Fig. 3A). Furthermore, our Pz8 was seen equally well in scalp to elbow or shoulder derivations (Fig. 3A) and in scalp to SC5 recordings. It was not present on SC5 to shoulder derivations. Therefore, we view the P28 as a true positivity recorded from the scalp electrode rather than a negativity recorded at the SC5 electrode. Kakigi et al. (1982) recorded a P:5 wave using a non-cephalic (hand) reference which appears to be identical to our P28 and the N24 of Lueders et al. (1981).
The cortical potentials Following N34 , a large localized positivity, ~8, was consistently recorded from Cz and adjacent sites ipsilateral to the stimulated leg in all subjects. In most individuals, a localized negativity, N38, occurred synchronously over the homologous contralateral scalp. The amplitude of the contralateral N38 relative to the ipsilateral P~8 was variable from person to person and indeed varied in the same individual depending on which leg was stimulated. Cruse et al. (1982) demonstrated a similar potential distribution for the P2 wave (peak latency 38 msec) and a late N 2 (peak latency 55 msec). The P38/N38 probably corresponds to the N20 component of the median SEP. In some recordings using non-cephalic references (Fig. 5A, arrow), there was a clear inflection marking the onset of contralateral N38 negativity and distinguishing it from the preceding N34. This appears to be analogous to the inflection in the contralateral negativity which marks the onset of the N20 and differentiates it from the preceeding N~s following median nerve stimulation (Desmedt and Cheron 1981b). Initial activation of the somatosensory cortex in mammals including primates results in a stereotyped initial surface positivity followed by a surface-negative response (Perl and Whitlock 1955). Localized cellular events occurring in deeper cortical layers characteristically result in a surface potential of opposite polarity. This polarity inversion results from the vertical orientation of pyramidal cells and the consequent large extracellular current flows occurring perpendicular to the surface which behave as virtual dipoles. Depolarization of cell bodies in layer IV (site of termination of specific thalamocortical relay afferent fibers) creates an active sink in the cortical
328 depths and a corresponding passive current source at the cortical surface. This is seen by a superficial or scalp electrode as a localized positivity (Goff et al. 1978). Since the primary sensory area for the leg and foot is located on the mesial aspect of the postcentral gyrus within the interhemispheric fissure, the initially positive primary cortical response representing depolarization of pyramidal cell bodies would be seen principally by electrodes located ipsilateral to the stimulated limb (Fig. 9). Slight differences in the location of the leg area (Penfield and Rasmussen 1950) would result in considerable variation in field of the scalp-recorded potential. If the leg area is located at the superficial edge of the interhemispheric fissure, the dipole orientation would be vertical or oblique at the scalp (Fig. 9A). On the other hand, if the leg area lies deeper in the interhemispheric fissure, the dipole orientation would be more horizontal with the initial positivity again projecting ipsilateral to the side of stimulation, but with the negative end of the dipole projecting contralaterally (Fig. 9B). In our view, ~8 and N3s simply represent two ends of a single dipole with variable orientation in the vertical plane. In our group of subjects, we observed a wide range of variation in P3JN38 distribution that would be expected from differences in dipole orientation (see also Fig. 6). Similar results have been reported by Cruse et al. (1982). Unlike early components of the SEP, potentials occurring after P38 varied considerably from trial to trial even within the same subject (see for example response at Cz in Fig. 3A). This variation appeared to be related to the subject's level of alertness but this was not studied systematically. In some individuals the peaks of P38 and N38 did not occur simultaneously: the peak of P38 appeared to be slightly delayed relative to N38. We hypothesize that this apparent asynchrony is caused by an additional component which begins about 3.5 msec after the onset of P3s/N38 and is seen as an initial positivity at both C~ and C~. We propose that this later potential is generated by a different neuronal population, perhaps the result of intracortical spread to secondary cortical areas. Soon after the onset of the primary cortical potentials (p~8/N3s), other cortical areas may be
M. SEYAL ET AL.
÷
A
4"
B
Y Fig. 9. Hypothesizedvariations in the field of distribution of the P38/N3spotential based upon the known anatomic variability of the location of the leg area. See text for further details.
activated resulting in a complex summation of potentials at the scalp. The scalp distribution and nature of these later components are presently under investigation. We conclude that stable, reproducible potentials can be recorded over the spine and scalp following stimulation of the PTN. Clinical application of these SEPs requires knowledge of the normal topography and variability of these potentials, and an understanding of their probable neural generators. The following montage might be considered for routine clinical purposes using a 4-
POSTERIOR T1BIALSEP channel evoked potential system: L1 spine to an electrode 10 cm above Channel 2: Fpz referred to SC5 Channel 3: C 3 or C a (ipsilateral to stimulated leg) referred to Fpz Channel 4: Cz referred to Fpz. Channel 1 :
This montage allows identification of the peripheral volley as it enters the spinal cord (channel 1), demonstrates the subcortical potentials (channel 2) and isolates the localized cortical responses (channels 3 and 4). Both channels 3 and 4 are necessary given the normal range of variability in the potential field of the P38/N38 and later waves. Some degree of sedation is required in most individuals to reduce muscle activity sufficiently to permit interpretable recording of the subcortical potentials.
Summary Somatosensory evoked potentials (SEPs) were elicited by stimulation of the posterior tibial nerve (PTN) in 12 normal adults. Recording using both ceptialic and non-cephalic references were obtained from multiple electrodes placed over the spine and scalp. Following PTN stimulation, the fastest recorded potentials of the afferent sensory volley proceeds up the spinal cord at constant velocity. After arrival of the volley at cervical cord levels, 3 widely distributed waves, P28, Pal and N34, are recorded from scalp electrodes. These 'far-field' potentials are followed by a localized positivity (P38) which has a peak voltage either at the vertex or just laterally toward the side of stimulation. A contralateral negativity (N38) was present in most individuals. We propose that P28 arises from medial lemniscus; that P31 is generated by ventrobasal thalamus; and that N34 is probably the result of further activity in thalamus a n d / o r thalamocortical radiations. The P38/N38 complex represents the primary cortical response to PTN stimulation. Its most consistent characteristic is a positivity at the vertex or immediately adjacent scalp areas ipsilateral to the stimulated leg. The
329 topography of the P38/N38potential varies slightly from individual to individual in a manner consistent with a functional dipole situated in the leg and foot area on the mesial aspect of the postcentral gyrus, whose exact location and orientation changes in accordance with known variations in the location of the leg area. R6sum6
Potentiels dvoquks somatosensoriels a la stimulation du nerf tibial postdrieur." composantes spinales et composantes du scalp Des potentiels 6voqu6s somatosensoriels (PES) ont 6t6 produits par stimulation du nerf tibial post6rieur (NTP) chez 12 adultes normaux. Des enregistrements utilisant des r6f6rences h la fois c6phaliques et non c6phaliques ont 6t6 obtenus partir d'61ectrodes multiples plac6es sur la moelle 6pini6re et le scalp. Apr6s stimulation du NTP, le plus rapide des potentiels de la vol6e aff6rente sensorielle remonte la moelle +pini6re ~ vitesse constante. Apr6s l'arriv6e de la vol6e au niveau de la moelle cervicale, 3 ondes ~t large distribution, P28, P31 et N34 sont recueillies par les 61ectrodes de scalp. Ces potentiels de champs lointains sont suivis par une positivit6 P38 localis~e qui a un voltage maximal soit au vertex, soit juste lat6ralement du c6t6 de la stimulation. Une n6gativit6 contralat+rale N38 6tait pr6sente chez la plupart des individus. Nous proposons que P28 a son origine dans le lemnisque m6dian, que P31 provient du thalamus ventrobasal et que N34 r6sulte probablement d'une activit6 plus tardive du thalamus e t / o u des radiations thalamo-corticales. Le complexe P38/N38 repr6sente la r6ponse corticale primaire h la stimulation du NTP. Sa principale caract6ristique est une positivitb au vertex ou sur le scalp imm6diatement ipsilat6ral h la jambe stimul6e. La topographie du potentiel P38/N38 varie l+g~rement d'une individu ~ l'autre; ces r6sultats sont compatibles avec le mod61e d'un dip61e fonctionnel situ6 sur l'aire de la jambe et du pied de la face m+diane du gyrus post-central, et dont la localisation et l'orientation changerait comme on sait que varie la localisation de l'aire de la jambe.
330
References Allison, T. and Hume, A.L. A comparative analysis of shortlatency somatosensory evoked potentials in man, monkey, cat and rat. Exp. Neurol., 1981, 72:592-611. Arrezo, J.C., Vaughn, H.G. and Legatt, A.D. Topography and intracranial sources of somatosensory evoked potentials in the monkey. II. Cortical components. Electroenceph. clin. Neurophysiol., 1981, 51: 1-18. Behse, F. and Buchthal, F. Normal sensory conduction in the nerves of the leg in man. J. Neurol. Neurosurg. Psychiat., 1971, 34: 404-414. Cracco, R.Q. Spinal evoked response: peripheral nerve stimulation in man. Electroenceph. clin. Neurophysiol., 1973, 35: 379 386. Cruse. R., Klem, G., Lesser, R. and Lueders, H. Paradoxical lateralization of cortical potentials evoked by stimulation of posterior tibial nerve. Arch. Neurol. (Chic.), 1982, 39: 222-225. Delbeke, J., McComas, A.J. and Kopec, S.J. Analysis of evoked lumbosacral potentials in man. J. Neurol. Neurosurg. Psychiat., 1978, 41: 293-302. Desmedt, J.E. and Brunko, E. Functional organization of farfield and cortical components of somatosensory evoked potentials in normal adults. Prog. clin. Neurophysiol., 1980, 17: 27-50. Desmedt, J.E. and Cheron+ G. Central somatosensory conduction in man: neural generators and interpeak latencies of far-field components recorded from neck and right or left scalp and earlobes. Electroenceph. clin. Neurophysiol., 1980, 50: 382-390. Desmedt, J.E. and Cheron, G. Prevertebral (oesophageal) recording of subcortical somatosensory evoked potentials in man: the spinal PI3 component and the dual nature of spinal generators. Electroenceph. clin. Neurophysiol., 1981 a, 52: 257-275. Desmedt, J.E. and Cheron, G. Non-cephalic reference recording of early somatosensory potentials to finger stimulation in adult or aging normal man: differentiation of widespread Now and contralateral N20 from the pre-rolandic P22 and N3o components. Electroenceph. clin. Neurophysiol., 1981 b, 52: 553-570. Ertekin, C. Studies of the h u m a n evoked spinogram. II. The conduction velocity along the dorsal funiculus. Acta neurol. scand., 1976, 53: 21-38. Goff, W.R., Allison, T. and Vaughn, H.G. Functional neuroanatomy of event-related potentials. In: E. Callaway, P. Tueting and S.H. Koslow (Eds.), Event-related Potentials in Man. Academic Press, New York, 1978: 1-79. Jones, SJ. and Small, D.G. Spinal and subcortical evoked potentials following stimulation of the posterior tibial nerve in man. Electroenceph. clin. Neurophysiol., 1978, 44: 299-306.
M. SEYAL ET A U Jones, S.J., Edgar, M.A. and Ransford, A.O. Sensory nerve conduction in the h u m a n spinal cord: epidural recordings made during scoliosis surgery. J. Neurol. Neurosurg. Psychiat., 1982, 45: 446-451. Kakigi, R., Shibasaki, H., Hashizume, A. and Kuroiwa, Y. Short latency somatosensory evoked spinal and scalp-recorded potentials following posterior tibial nerve stimulation in man. Electroenceph. clin. Neurophysiol., 1982, 53: 602-611. Loeb, G.E. Decreased conduction in the proximal projections of myelinated dorsal root ganglion cells in the cat. Brain Res., 1976, 103: 381-385. Lorento de N6, R. A study of nerve physiology. XVI. Analysis of the distribution of the action currents of nerve in volume conductors. Stud. Rockefeller Inst. reed. Res., 1947, 132: 384-482. Lueders, H., Andrish, J., Gurd, A., Weiker, G. and Klem, G. Origin of far-field subcortical potentials evoked by stimulation of the posterior tibial nerve. Electroenceph. clin. Neurophysiol., 1981, 52: 336-344. Magladery, J.W., Porter, W.E., Park, A.M. and Teasdall, R.D. Electrophysiological studies of nerve and reflex activity in normal man. IV. The two-neurone reflex and identification of certain action potentials from spinal roots and cord. Bull. Johns Hopk. Hosp., 1951, 88: 499-519. Noel, P. and Desmedt, J.E. Cerebral and far-field somatosensory evoked potentials in neurological disorders involving the cervical spinal cord, brainstem, thalamus and cortex. Prog. clin. Neurophysiol., 1980, 7: 205-230. Patton, H.D. Special properties of nerve trunks and tracts. In: T.R. Ruch and H.D. Patton (Eds.), Physiology and Biophysics, Vol. IV. Saunders, Philadelphia, Pa., 1982: 104-124. Penfield, W. and Rasmussen, T. The Cerebral Cortex of Man. A Clinical Study of Localization of Function. Macmillan, New York, 1950. Perl, E.R. and Whitlock, D.G. Potentials evoked in cerebral somatosensory region. J. Neurophysiol., 1955, 18: 486-501. Rossini, M.P., Caracco, R.Q., Cracco, J.B. and House, W.J. Short latency somatosensory evoked potentials to peroneal nerve stimulation: scalp topography and the effect of different frequency filters. Electroenceph. olin. Neurophysiol., 1981, 52: 540-552. Truex, R.C. and Carpenter, M.B. Strong and Elwyn's H u m a n Neuroanatomy. Williams and Wilkins, Baltimore, Md., 1964. Tsumoto, T., Hirose, N., Nonaka, S. and Takahashi+ M. Analysis of somatosensory evoked potentials to lateral popliteal nerve stimulation in man. Electroenceph. clin. Neurophysiol., 1972, 33: 379-388. Vas, G.A.,Cracco, J.B. and Cracco, R.Q. Scalp recorded short latency cortical and subcortical somatosensory evoked potentials to peroneal nerve stimulation. Electroenceph. clin. Neurophysiol., 1981, 52: 1-8.