Significant Case Reports

Extrarolandic Origin of
Spike and Myoclonus
Activity in Epilepsia
Partialis Continua: A
Magnetoencephalographic
and Positron Emission
Tomography Study

Although there was no obvious phase relationship between interictal spikes and
myoclonic jerks, the myocloni were initiated from within the interictal spike area
in the extrarolandic cortex. The data demonstrate that motor symptoms may be a
remote effect of epileptic activity within
functional brain circuits.

J. Volkmann, MD
Department of Neurology
Heinrich-Heine-University
Moorenstr. 5
40225 Dusseldorf, Germany

Key words: Epilepsia partialis continua,
PET, EEG.

R. J. Seitz, MD

Volkmann J, Seitz RJ, Muller-Gartner HW,
Witte OW. Extrarolandic origin of spike
and myoclonus activity in epilepsia partialis
continua: a magnetoencephalographic and
positron emission tomography study.
J Neuroimag 1998;8: 103-106

Department of Neurology
Heinrich-Heine-University

H. W. Muller-Gartner, MD
Institute for Med1cine
Research Center JL.ilich
Julich, Germany

0. W. Witte, MD
Department of Neurology
Heinrich-Heine-University

ABSTRACT
The dipole sources of interictal spike activity were localized and the myoclonus
activity back-averaged by combined magnetoencephalography and surface electromyographic measurements in a child
who had epilepsia partialis continua without a structural brain lesion. Dipole
sources were matched with metabolic information obtained from interictal 5fluoro-o-glucose positron emission tomography (PET) and superimposed onto highresolution magnetic resonance images.
Dipole sources of interictal epileptic discharges clustered within the inferior parietal cortex, which also showed a regional
hypermetabolism on PET scans. The dipole
sources of reafferent activity following
myocloni in the postcentral gyrus were
associated with a local hypometabolism.

Received Mar 26, 1996, and in revised
form Sep 9, 1996, and May 6, 1997. Accepted for publication May 7, 1997.
Address correspondence to Dr
Volkmann.

Epilepsia partialis continua (EPC) is a
rare form of a partial epileptic status
characterized by continuous myoclonic
twitching of a body part. EPC can occur
with a variety of etiologically different
acute or chronic brain lesions including
Rasmussen's encephalitis. Most authors
describe a cortical origin of myoclonic
activity in EPC. 1 Time-locked myoclonic
and interictal spike activity in some patients with EPC suggests an initiation of
myoclonic activity from within the spikegenerating zone. In the majority of patients, however, polygraphic recordings
fail to demonstrate an obvious phase relation between epileptic and myoclonic
discharges. Using electromyography
(EMG)-triggered back-averaging of electroencephalographic (EEG) activity in
a patient with EPC without premyoclonic
EEG spikes, Celesia and colleagues 2 recently demonstrated that the source of
myoclonic activity fell within the foot
area of the primary motor cortex in close
vicinity to an arteriovenous malformation
presumably causing the disorder. The
present authors combined magnetoencephalography (MEG), [18F]2-fluoro-2deoxy-n-glucose positron emission tomography (FDG-PET), and high-resolution
magnetic resonance imaging (MRI) to
noninvasively localize the irritative
(spike-generating) zone and symptomato-

genic (myoclonus-generating) zone in a
child who had EPC without a structural
brain lesion.
Case Report
The 11-year-old girl was the seventh child
of healthy, nonconsanguineous parents,
born after uneventful pregnancy and labor. At the age of 2 years the girl had a
minor head injury as a result of a fall
from a chair and shortly later developed
stuttering. Motor and intellectual development was otherwise normal. Six
months before admission to the hospital
the girl experienced a first focal motor
seizure on the right side of her body with
secondary generalization. Within the following weeks, seizure frequency increased
up to 12/hr regardless of the high doses
of various anticonvulsive drugs administered. At the same time a rapidly progressive right-sided hemiparesis emerged.
On admission to the authors' hospital,
neurological examination revealed a
right-sided hemiparesis and a continuous,
approximately 1/sec myoclonic twitching
of the right fingers. Somatosensory
evoked potentials (SSEPs) showed increased conduction latencies (42.8 msec)
after right tibial nerve stimulation but no
giant potentials. Transcranial magnetic
stimulation could not elicit a motor
evoked response in the right anterior
tibial muscle, despite normal amplitude
and central conduction latencies to the
other extremities. MRis at the time of
the recording were normal but previous
MRis had shown a diffuse edema of the
left sensorimotor and parietal cortex.
Positive cerebrospinal fluid (CSF) oligoclonal bands with preserved normal cell
count suggested a chronic encephalitis.
Rasmussen's encephalitis was discussed
but the results of a brain biopsy were inconclusive.
Methods
Spontaneous interictal neuromagnetic activity was registered using a 74-channel
dual-probe MEG system (Biomagnetic
Technologies). The two MEG probes
were centered around C3/C4 covering
approximately two-thirds of each hemisphere (Fig 1). For dipole fitting, a
spherical head model was chosen; it was

Copyright© 1998 by the American Society of Neuroimaging

103

-31 ms

77 ms

MEG
(LH)

MEG
(RH)

EMG
0 ms

The back-averaged, jerk-locked MEG waveforms superimposed for all channels of the left (LH) and right hemispheric (RH) probe
are shown. Dipolar field patterns (shown in isocontour diagrams on the right) resulting in dipole fits with low residual variance could only
be observed around 30 msec prior and 80 msec after the onset of EMG activity over the left hemisphere. Solid lines in the isocontour
diagrams indicate ingoing, and dashed lines 1ndicate outcoming magnetic flux. For better orientation the position of each sensor of the
dual-probe 37-channel MEG system is displayed on the left with respect to the digitized head surface of the patient

Fig 1.

determined by digitizing the surface of
the subject's head and using an iterative
algorithm to calculate the best-fitting
sphere. Single equivalent dipole fits with
a goodness of fit > 0.95 were used to
compute mean dipole locations in a window of ±5 msec around peak latencies for
each of 27 interactively identified, interictal spikes. The rectified and integrated
EMG activity of the flexor digitorum
superficialis was used to trigger backaveraging of myoclonus activity (204
events). The MEG baseline was established in each epoch by DC-offset correction in the time period 500 to 250 msec
prior to EMG onset. From the backaveraged waveforms and for each hemisphere, independent single equivalent dipole (ECD) solutions were computed for
every sample point within a time window
of -150 to 500 msec around EMG onset.
Only those dipole fits that had a goodness of fit > 0.9 at latencies showing a
clear dipolar field distribution were accepted for further analysis.
Immediately prior to MEG, FDG-PET
was carried out using a Scanditronix PC
4096/7 WB PET camera as described in
104

detail elsewhere? No seizure activity
was noticed clinically or on the EEG during PET scanning. lnterictal spike rate
and myoclonus activity were comparable
during PET and MEG.
ECD and PET data were matched
with the individual MRI anatomy (Siemens Magnetom, T1 three-dimensional
[3D] flash sequence) using the surfacematching module of Analyze to align the
MRI and PET volumina and in-house
software to convert MEG dipole coordinates into MRI pixel coordinates based
on fiducial markers. 4 MRI sectioning,
3D rendering, and multimodal image integration were performed on a SUN
workstation using Analyze software.
Hypermetabolic and hypometabolic
areas on PET scans were defined as pixel
values lying outside of a 2-standard deviation range around the mean cortical metabolism of the contralateral hemisphere.
This statistical approach, which consists
of the only way of looking at focal
changes in brain metabolism in a single
case, is justified in two regards: (1) Heiss
and coworkers 5 showed that for normal
subjects, metabolic rates and kinetic con-

Journal of Neuroimaging Vol 8 No 2 April1998

stants in FDG-PET may vary substantially among different subjects and cortical regions of the same subject, but not
between the hemispheres of each individual. (2) In the present patient the ratio
of the metabolic rate in the left and right
hemisphere was 0.99 ± 0.02, proving a
significantly symmetrical hemispheric me·
tabolism like that in normal subjects.
Results
Evaluation of the raw MEG and EMG
traces revealed a random phase relationship between interictal spike and myoclonic activity. Less than 20% of myoclonic jerks were preceded by visible
spike activity at random time intervals.
On the other hand, spikes were often not
followed by myoclonic jerks.
For 27 interactively identified, interictal spikes, clear dipolar magnetic field
patterns were observed and allowed
source localization using a single equivalent dipole model. Dipole sources for
these spikes clustered within an area of
approximately 60 cm 3 within the left inferior parietal lobule.
The jerk-locked, averaged waveforms

(see Fig 1) showed one prominent de flecti on in the premyoclonic period with a
peak latency of 31 msec prior to E MG
onset and three deflections in the postmyoclonic period with peak latencies of
77. 172. and 300 msec after EMG onse t,
which were all maximal over the left
he misphere . A large deflectio n with
inward-dire cted flux in all channe ls over
th e left hemisphere and outward-directed
flux in all right hemispheric channels during myoclonic jerks was presumably
caused by a movement artifact and did
not result in a localiza ble field distribution. Stable dipolar fie ld p atte rns resulting in single eq ui va knt dipole fits with
low residual variance (goodness of fit
> 0.90) could only be observed for th e
premyoclonic field 50 to 20 msec prior to
EMG onset and for the first postmyoclonic field 56 to lOO msec after EMG
onset (see Fig I). Since the source locations varied little in the respective time
periods. the dipole coordinates at pea k
latencies were chosen for anatomical
coregistration with PET and MRI images.
The second and third postmyoclonic
fields at 172 and 300 msec after EMG
onset resulted in more diffuse fie ld patterns th at could no t be fitted using a
single equivalent dipole model. The
source of premyoctonic activity 3 1 msec
before EMG onset loca lized to the anterior border of the spike area within the
left inferior parietal lobule (goodness of
fit = 0.956). Seventy-seven milliseconds
after EMG o nse t of myoclo nic activity, a
dipolar generator in the left somatosensory cortex was obse rve d (goodness of fit
= 0.952) (Fig 2). This postmvoclonic field
was consistent in latency, field distribution , and source location with the movement evoked field I (MEF l) normally
obtaine d after voluntary finger movements.<' The MEF1 and its e lectrical
counterpart m ost probably reflect reafferent activity to area 3a from muscle
receptors 7 x
spindle and j~Jint
FDG-PET showed increased FOG uptake in the left infe rior posterior parie tal

cortex- only partially overlapping th e
spike area. In add ition, a focal hype rme tabolism was found within the premotor
cortex surrounded by decreased FOG
uptake in prefrontal and sensorimotor
cortex (sec Fig 2 ). As remote effects, a significant hype rmetabolism appeared in the
left dorsal thalamus and right cerebellum.
Discussion
Noninvasive MEG and PET scans in the
present patient suggest a cortical origin of
EPC in agreement with previous results
of intraoperati ve cortical recordin gs. 2 ·'J- J:>
D emonstrated were ( l) cortical gene rators o f in terictal epileptic fields.
(2 ) the initia tion of jerks in the interictal
spike area. and (3 ) focal cortical abnormalities in FOG-PET
Some featu res of the present patient
deserve special attention: Dipole an alysis
revealed that in terictal epile ptiform activity was ge nerated in the in ferior parie tal
cortex. D espite an absence of a phase
relationship between interictal spikes and
myoclonic jerks, premyoclonic activity
surprisingly originated from within the
intcrictal spike area. as re vealed by the
back-averaging technique. T ranscranial
magnetic stimulation of the prim ary motor cortex revea led normal corticomuscular conduction latencies (17.2 msec to the
left first dorsa l interosseous muscle and
16.4 msec to the right first dorsal interosseous muscle). The latency of 3 1
msec between the prem yoclonic fie ld and
myoclonus onset, th erefore, suggests a
polysynaptic tran sm ission to the spinal

cord, probably via activa tion of the premotor cortex , that was hype rme tabo \ic
but showed no dipolar activity on MEG.
Unfortunately, MEG did not identify the
site of fin al corticospinal o utflow. which
may be related to multiple fact o rs such as
the presence of myoclonus-induced movement artifacts, limitatio ns of the single
equi va le nt dipo le mode l, o r contributio n
of radi al curren t components to the generator.
While many reports e mphasize that an
abnormality within motor cortex itself is
responsible for EPC 1 Obeso and coautho rs ' :> argued tha t via dense corticocortica l connections other cortical areas
may exert a remote effect on the mo tor
cortex in patients with corti cal myoclonus. This would also explain the giant
SSEPs often observed in cortical reflex
myoclonus: Shibasaki and colleagues 14
suggested th at the enhanceme nt of the
Pl-N2 compo ne nt is generated by the
same mechanism that un derlies
m yoclonus-related cortical spikes. The
hyperexcitability in the somatosensory
cortex might provide an increased input
volley into the motor cortex, which the n
causes the jerk. In accordance with such
an indirect generation of m yoclonic activity, Cowan and associates 10 presented a
case of cortical myoclo nus, EPC, and motor epilepsy in which surface EEG, intraoperative corticography, and PETrevealed abnormal activity in the
postcentra l p arie tal cortex. Afte r resecti o n o f the parietal focus, the cortical myoclonus and EPC disappeared.

Fig 2. Summa ry of th e cortica l meta bol ic
and e lectrophys1o logical cha ng es re lat ed to
epilepsia part1alis continua in the investigated
child. In a semitransparent 3D MRI rendering ,
areas of re lat1ve cortica l hypermetabolism appea r 1n red, wh ile areas of corti ca l hypometabolism are in blue. Dipole sou rces of inte rict a l sp 1kes (b lu e squares), myoc lonus ini tiat ing activity (yellow dot), and myoclo nus
feedback act1v1ty (green dot) computed fro m
MEG are superimposed .

Significant Case Report: Volkmann et al: MEG and PET in Epilepsia Partialis Continua

105

Intuitively, one would expect a focal
hypermetabolism in the area of the epileptic focus. Such a hypermetabolism has
indeed been observed in some patients
with EPC, and can also be observed in
experimentally induced focal epilepsy. 15
However, just as common has been the
observation of a focal hypometabolism in
patients with EPC. 16•17 In the present patient areas with focal hypermetabolism
and areas with hypometabolism were
found. Although areas of regional hypermetabolism and interictal spike activity
were adjacent in the inferior parietal cortex, there was only little overlap between
them. A possible explanation would be a
decoupling of electrophysiological activity
and brain metabolism in this region,
which has been described in animal models of focal brain lesions [18]. An additional area of focal hypermetabolism was
situated in the premotor cortex and was
probably related to the generation of epileptic jerks. The hypermetabolic areas
were surrounded by areas of hypometabolism as also described for experimental
epilepsy. 15 Notably, the reafferent somatosensory activation fell into a zone
with reduced metabolism. This may be
due to the slow stimulus rate in addition
to a partial volume effect based on the
relatively low spatial resolution of PET
and the small area of activated cortex, 19
which were not sufficient to evoke focal
metabolic increases within an area of
pathologically reduced metabolism.
The present report shows that noninvasive techniques like MEG, PET, and
MRI may be useful in analyzing the functional circuits involved in focal epileptic
activity. In this respect the present case is
an illustration of the conceptual framework put forward by Ltiders 20 who differentiated between subregions of cortex
related to the generation of interictal epileptic discharges (irritative zone), seizure
onset (pacemaker zone), and ictal and
interictal symptomatology (ictal symptomatogenic and functional deficit zone) in
focal epilepsy. While the pacemaker
zone-as in the present patientusually resides within the irritative zone,
the ictal symptomatogenic and functional
deficit zone may be located remotely,
sometimes even in the contralateral
hemisphere, and affected by the spread of
pathological activity.
References
1. Thomas JE, Reagan TJ, Klass DW.
Epilepsia partialis continua. Arch
Neural 1977;34:266-275
2. Celesia GG, Parmeggiani L, Brigell
106

M. Dipole source localization in a
case of epilepsia partialis continua
without premyoclonic EEG spikes.
Electroencephalogr Clin Neurophysiol 1994;90:316-319
3. Seitz RJ, Schlaug G, Kleinschmidt A,
et al. Remote depressions of cerebral
metabolism in hemiparetic stroke:
topography and relation to motor and
somatosensory functions. Hum Brain
Mapping 1994;1:81-100
4. Walter H, Seitz RJ, Knorr U, et al.
Individual somatotopy of primary
sensorimotor cortex revealed by intermodal matching of MEG, PET and
MRI. Brain Topogr 1992;5:183-187
5. Heiss W-D, Pawlik G, Herholz K, et
al. Regional kinetic constants and cerebral metabolic rate of glucose in
normal human volunteers determined
by dynamic positron emission tomography of [18F]-2-fluoro-2-deoxy-o-glucose. J Cereb Blood Flow Metab
1984;4:212-223
6. Kristeva R, Cheyne D, Deecke L.
Neuromagnetic fields accompanying
unilateral and bilateral voluntary
movements: topography and analysis
of cortical sources. Electroencephalogr Clin Neurophysiol 1991;81:284298
7. Arezzo J, Vaughan HG, Koss BT.
Relationship of neuronal activity to
gross movement-related potentials in
monkey pre- and postcentral cortex.
Brain Res 1977;132:362-369
8. Arezzo J, Vaughan HG. Intracortical
sources and surface topography of the
motor potential and somatosensory
evoked potential in the monkey. Prog
Brain Res 1980;54:77-83
9. Kugelberg E, Widen L. Epilepsia partialis continua. Electroencephalogr
Clin Neurophysiol 1954;6:503-506
10. Cowan JMA, Rothwell JC, Wise RJS,
et al. Electrophysiological and positron emission studies in a patient with
cortical myoclonus, epilepsia partialis
continua and motor epilepsy. J Neural Neurosurg Psychiatry 1986;49:796807
11. Wieser HG, Graf HP, Bernouilli C, et
al. Quantitative analysis of intracerebral recordings in epilepsia partialis
continua. Electroencephalogr Clin
Neurophysiol1978;44:14-22
12. Siegfried J, Bernouilli C. Stereoencephalographic exploration and epilepsia partialis continua. Acta Neurochir (Wien) 1976;23:183-191
13. Obeso JA, Rothwell JC, Marsden
CD. The spectrum of cortical myoclonus. Brain 1985;108:193-224
14. Shibasaki H, Yamashita Y, Neshige
R, et al. Pathogenesis of giant somatosensory evoked potentials in
progressive myoclonus epilepsy. Brain
1985;109:225-240
15. Witte OW, Bruehl C, Schlaug G, et

Journal of Neuroimaging Vol 8 No 2 April 1998

al. Dynamic changes of focal hypometabolism in relation to epileptic
activity. J Neural Sci 1994;124:188197
16. Hefter H, Witte OW, Reiners K, et
al. High frequency bursting during
rapid finger movements in an unusual
case of epilepsia partialis continua.
Electromyogr Clin Neurophysiol
1994;34:95-103
17. Ochs RF, Gloor P, Tyler J, et al. Positron emission tomography measurements of glucose metabolism in epilepsia partialis. Neurology 1985;35:227
18. Ginsberg MD, Castella Y, Dietrich
WD, et al. Acute thrombotic infarction suppresses metabolic activation
of ipsilateral somatosensory cortex:
evidence for functional diaschisis. J
Cereb Blood Flow Metab 1989;9:329341
19. Ibanez V, Deiber MP, Sadato N, et
al. Effects of stimulus rate on regional cerebral blood flow after median nerve stimulation. Brain 1995;
118:1339-1351
20. Ltiders HO. Epilepsy surgery. New
York: Raven, 1991

Isolated Straight Sinus
Thrombosis: Clinical and
Neuroradiologic Correlates
James Valeriano, MD
Department of Neurology
MCP-Hahnemann School of Medicine
Allegheny General Hospital
Pittsburgh, PA

Kumar Bhagavatula, MD
Department of Neurology
MCP-Hahnemann School of Medicine
Allegheny General Hospital

Andrew Ku, MD
Department of Radiology
MCP-Hahnemann School of Medicine
Allegheny General Hospital

Peter J. Snyder, PhD
Department of Neurology
MCP-Hahnemann School of Medicine
Allegheny General Hospital

ABSTRACT
A rare case of straight sinus thrombosis in
a 36-year-old female with sudden onset of
status migrainosus is presented. This condition was demonstrated by a linear density in the midline on a noncontrast computed tomography scan, as a filling defect
at the location of straight sinus on mag-