Seizure (2004) 13, 317—321
Diffusion changes suggesting predominant
vasogenic oedema during partial status
Keun-Sik Honga,*, Yong-Jin Choa, Sang Kun Leeb, Sang-Wuk Jeonga,
Won Kyoo Kima, Eun Jin Oha
aDepartment of Neurology, Ilsan Paik Hospital, Inje University, 2240 Daewha-dong, Ilsan-gu,
Goyang-city, Gyonggi-do 411-706, South Korea
bDepartment of Neurology, Seoul National University College of Medicine, Gyonggi-do, South Korea
totoxic oedema during partial status epilepticus (PSE). However, vasogenic oedema
related to the breakdown of the blood-brain-barrier (BBB) and ictal hyperperfusion
could be the predominant DWI findings in the epileptogenic area during PSE. We report
a case of PSE with ictal aphasia, right hemiparesis, and repetitive focal motor seizure
of the right side. T2-weighted image (T2WI) and apparent diffusion coefficient (ADC)
maps obtained during PSE showed an increased signal in the left temporo-parietal
area, indicative of vasogenic oedema. EEG documented the ictal activities and single
photon emission tomography (SPECT) showed asymmetrically increased perfusion in
the corresponding area. Follow-up T2WI, DWI, and ADC maps obtained 3 months later
showed the disappearance of the previous abnormalities. However, T2WI showed cor-
tical atrophy and newly developed white matter changes in the corresponding area.
This case shows that DWI findings may be variable during PSE, dependent on the
predominance of cytotoxic and vasogenic oedema.
© 2003 BEA Trading Ltd. Published by Elsevier B.V. All rights reserved.
Diffusion-weighted imaging (DWI) has demonstrated a focal area of cy-
Diffusion-weighted image (DWI) can detect the
changes in water diffusion and it has been well
documented that the DWI can distinguish between
cytotoxic oedema and vasogenic oedema. It has
proven useful in the evaluation of ischemic stroke
and is also becoming increasingly useful in the
evaluation of a wide variety of other disease pro-
*Corresponding author. Tel.: +82-31-910-7277;
E-mail address: firstname.lastname@example.org (K.-S. Hong).
cesses including neoplasms, intracranial infections
and traumatic brain injury.
DWI studies in animal experiments and hu-
man with partial status epilepticus (PSE) have
demonstrated the decrease of apparent diffusion
coefficient (ADC) value in the epileptogenic foci,
indicative of cytotoxic oedema.1—6
However, blood-brain-barrier (BBB) breakdown
and subsequent vasogenic oedema have been
reported in humans and animals with PSE.7,8
Moreover, in several MRI reports of human with
PSE, the vasogenic oedema has been thought as
the most probable mechanism for the reversible
1059-1311/$ – see front matter © 2003 BEA Trading Ltd. Published by Elsevier B.V. All rights reserved.
318K.-S. Hong et al.
signal in the left MCA (empty arrow). (E—H) MRI obtained at 32h after the seizure onset. T2WI, DWI and ADC showed
abnormal signals in the left temporo-parietal areas, suggesting vasogenic oedema. However, a very small area of thin
cortical layer showed marked hyperintensity on DWI without increased signals on ADC maps (arrows). MRA showed
increased signal in the left MCA (empty arrow). (I—L) Follow-up MRI performed 3 months later showed atrophic
changes in the left temporo-parietal area with an increased signal in the underlying white matter (arrow head). MRA
no longer showed signal asymmetry in either MCA. MRA (A, E, I); T2WI (B, F, J); DWI (C, G, K); ADC maps (D, H, L).
R, right; L, left.
(A—D) Initial MRI performed on the day after the thrombolysis was unremarkable except for an increased
Thus, DWI findings during PSE could be vari-
able, dependent on the predominance of cytotoxic
oedema and vasogenic oedema. To the best of
our knowledge, diffusion changes suggesting pre-
dominant vasogenic oedema during PSE have not
been documented by DWI studies. We report a case
of PSE in which DWI studies showed predominant
vasogenic oedema in the epileptogenic area.
Seventy-six-year-old woman without a previous
history of seizure was admitted to the emer-
gency room because of right hemiplegia and global
aphasia, which developed 1h before admission.
Emergency CT scan did not reveal any remarkable
findings and intravenous thrombolysis with r-tPA
was performed under the impression of ischemic
stroke in the territory of left middle cerebral
artery (MCA). After thrombolysis, the right hemi-
plegia completely resolved, but sensory aphasia
persisted. MR angiography (MRA) performed on the
day after the thrombolysis showed the recanalized
left MCA (Fig. 1A). T2-weighted image (T2WI), DWI,
and ADC maps were unremarkable (Fig. 1B—D). The
patient was discharged on anticoagulation therapy,
but sensory aphasia of a mild degree persisted but
with decreasing severity.
After 1 month later, she was readmitted because
of aggravated sensory aphasia, confusion, right
hemiparesis, and brief, recurrent attacks of clonic
movement in right upper extremity without gen-
eralized tonic—clonic movement. EEG performed
at 28h after the seizure onset showed periodic
sharp waves with brief runs of low voltage fast
activities in the left temporal, parietal, and occip-
ital electrodes, compatible with PSE (Fig. 2A). MRI
performed at 32h after the seizure onset showed
increased signal in the left temporo-parietal area
on T2WI, DWI, and ADC maps (Fig. 1E—H). How-
ever, a very small area of thin cortical layer showed
marked hyperintensity on DWI without increased
signals on ADC maps (arrows in Fig. 1F—H). MRA
Diffusion changes suggesting predominant vasogenic oedema 319
with the intermittent runs of low voltage, fast activities in the left temporal, parietal, and occipital electrodes
(arrow). (B) Follow-up EEG 3 months later showed slowings in the corresponding electrodes.
(A) During PSE, EEG documented the ictal discharges, which consisted of frequent periodic sharp waves
showed an asymmetrically increased signal in the
left MCA (Fig. 1E).
Single photon emission tomography (SPECT) per-
formed 3 days later showed mild hyperperfusion in
the left parietal area (Fig. 3). Phenytoin infusion
(1000mg over 1h) was started at 1h after EEG and
then maintained with adjusted dose according to
serum level. Phenytoin and repeated doses of intra-
and EEG abnormalities and intravenous phenobar-
bital infusion with loading dose (900mg) was added
on hospital day 10. The periodic sharp waves dis-
appeared on EEG performed after the phenobarbi-
tal infusion and her neurologic deficit resolved over
several days, except for the sensory aphasia.
Follow-up MRI and EEG were obtained 3 months
later. T2WI showed regional atrophy in the left
temporo-parietal area with high signal changes in
the underlying white matter and a dilated lateral
ventricle (Fig. 1J). The previous abnormal signals
on the DWI and ADC maps disappeared (Fig. 1K
and L). EEG showed localized slowings in the cor-
responding area (Fig. 2B). The patient had persis-
tent sensory aphasia of mild degree and recurrent
episodes of confusion during follow-up.
Magnetic resonance imaging technique
All MRI studies were performed on a 1.5T MR unit
(Siemens, Erlangen, Germany) with an echopla-
nar imaging (EPI) capability. Fast spin-echo, T2WI
(TR/TE, 4000/96ms; field of view, 21cm × 21cm;
matrix, 256 × 192; and slice thickness, 5mm) was
obtained. DWI was obtained in the transverse
plane using a single-shot EPI (TR/TE, 5900/159ms;
field of view, 24cm × 24cm; matrix, 128 × 128;
slice thickness, 5mm; and two b values, 0 and
900s/mm2). The diffusion-gradients were applied
along the three axes (x, y, z) simultaneously.
In our patient, T2WI and ADC maps showed in-
creased signal in the left temporo-parietal region
in which EEG showed ictal activities. These findings
320K.-S. Hong et al.
(arrow). R, right; L, left.
SPECT performed 3 days after the onset of PSE showed asymmetric hyperperfusion in the left parietal area
suggest that vasogenic oedema might have been
the predominant pathology in our patient with PSE,
although we could not provide ADC values because
of technical problems. The high signal intensity
on DWI seems to be partially due to the T2-shine
through effect considering the high signal seen on
Our findings of predominant vasogenic oedema
in the epileptogenic area during PSE are in contrast
with previously reported DWI findings in human
and animal models. In animal models, increased
signal on DWI and decreased ADC values, reflecting
cytotoxic oedema, were observed in the epilepto-
genic area.1,2These findings were similar to those
of acute ischemia. In humans, few MRI studies
performed during PSE have documented DWI abnor-
malities suggesting cytotoxic oedema in the epilep-
togenic area.4—6The failure of Na+/K+-ATPase,
excessive release of excitatory amino acids, and in-
creased membrane permeability are the proposed
mechanisms of cytotoxic oedema.3,11,12
However, seizure can trigger acidosis and the
breakdown of the BBB, resulting in vasogenic
oedema. In animal models, seizures induced by var-
ious convulsive agents resulted in BBB breakdown
and subsequent oedema, which may have been
related to circulatory and metabolic changes.8In
humans with prolonged PSE, angiography showed
a capillary blush in the focal area, in which CT
showed the focal cerebral oedema. These findings
suggest BBB breakdown and vasogenic oedema in
the epileptogenic area.7Ictal hyperperfusion might
aggravate vasogenic oedema. In our patient, SPECT
performed during PSE documented the hyperperfu-
sion in the epileptogenic area (Fig. 3).
DWI changes suggesting vasogenic oedema have
been observed only in the case of PSE reported by
Wieshmann et al.4However, vasogenic oedema in
their case was limited in the remote subcortical
area compared with the more widespread area in
The timing of MRI acquisition after the seizure
onset is an important factor determining whether
the DWI changes could be the pattern of cytotoxic
or vasogenic oedema. The animal experiments
showed that the ADC value is maximally decreased
at 24h after the seizure onset. After then it re-
turn to the control level between 48 and 72h,
but is not increased compared to the control.1,2
In our patient, DWI was obtained at 32h after the
seizure onset and the timing of MRI acquisition is
not so delayed compared to those in the previous
Diffusion changes suggesting predominant vasogenic oedema321
animal experiments1,2and human reports.5,6In
other report of nonconvulsive status epilepticus,
MRI obtained on the third day of onset showed the
cytotoxic oedema.13Thus, our DWI finding of pre-
dominant vasogenic oedema could not be entirely
attributed to the delayed MRI scan.
In our case, diffusion changes suggesting cyto-
toxic oedema were also observed in the limited
area. The very small thin layer of cortex showed
marked hyperintensity on DWI without an increased
signal on ADC map, reflecting predominant cyto-
toxic oedema (arrows in Fig. 1F—H).
Thus, predominant diffusion abnormalities in the
epileptogenic area could be either vasogenic or cy-
totoxic oedema pattern, dependent on the predom-
inance of both types of oedema.
Vasogenic oedema is usually considered re-
versible in comparison with cytotoxic oedema.
However, in our case, some irreversible neuronal
damage was evident despite the prominent vaso-
genic oedema on DWI, as evidence by permanent
sensory aphasia of a mild degree and by localized
slowing in the left temporal area on follow-up EEG.
Furthermore, follow-up T2WI also showed enlarge-
ment of the lateral ventricle and sulci in the corre-
sponding area (Fig. 1J), which had not been found
on initial MRI, indicative of regional neuronal loss.
This irreversible damage seems to be induced by
the coexistent cytotoxic oedema which is largely
masked by the prominent vasogenic oedema, but
evident in small, thin cortical layer (arrows in
Fig. 1F—H). Another possibility is that the patient
might have had recurrent seizure attacks result-
ing in further neuronal damage, because she suf-
fered from recurrent episodes of confusion during
It is of interest that follow-up T2WI showed newly
developed high signal intensity in the underlying
white matters in left temporo-parietal area. As the
signal changes in the underlying white matters were
observed only in the limited area of the previous
epileptogenic focus, these changes seem to be re-
lated to the PSE rather than other pathology such
as ischemia. This indicates that the white matter
changes as well as cortical atrophy could be the
findings of the seizure-induced damage.
In conclusion, vasogenic oedema is commonly as-
sociated with PSE. Diffusion changes may be ob-
served as vasogenic oedema as well as cytotoxic
oedema during PSE and the recognition of these
variable diffusion changes may be important for
the diagnosis and understanding of the pathogenic
This work was supported by grant (00012002355)
from Inje University, 2002.
1. Righini A, Pierpaoli C, Alger JR, Di Chiro G. Brain
parenchyma apparent diffusion coefficient alterations as-
sociated with experimental complex partial status. Magn
Reson Imaging 1994;12:865—71.
2. Nakasu Y, Nakasu S, Morikawa S, et al. Diffusion-weighted
MR in experimental sustained seizures elicited with kainic
acid. AJNR Am J Neuroradiol 1995;16:1185—92.
3. Wang Y, Majors A, Najm I, et al. Postictal alteration
of sodium content and apparent diffusion coefficient
in epileptic rat brain induced by kainic acid. Epilepsia
4. Wieshmann UC, Symms MR, Shorvon SD. Diffusion changes
in status epilepticus. Lancet 1997;350:493—4.
5. Lansberg MG, O’Brien MW, Norbash AM, Moseley ME, Morrel
M, Albers GW. MRI abnormalities associated with partial
status epilepticus. Neurology 1999;52:1021—7.
6. DiehlB, NajmI,Ruggieri
diffusion-weighted imaging in a case of lesional epilepsy.
7. Sammaritano M, Andermann F, Melanson D, et al. Prolonged
focal cerebral edema associated with partial status epilep-
ticus. Epilepsia 1985;26:334—9.
8. Nitsch C, Klatzo I. Regional patterns of blood-brain barrier
breakdown during epileptiform seizures induced by various
convulsive agents. J Neurol Sci 1983;59:305—22.
9. Henry TR, Drury I, Brunberg JA, Pennell PB, McKeever PE,
Beydoun A. Focal cerebral magnetic resonance changes
10. Yaffe K, Ferriero D, Barkovich AJ, Rowley H. Reversible MRI
abnormalities following seizures. Neurology 1995;45:104—
11. Sloviter RS, Dempster DW. ‘‘Epileptic’’ brain is replicated
qualitatively in the rat hippocampus by central injection of
glutamate or aspartate but not by GABA or acetylcholine.
Brain Res Bull 1985;15:39—60.
12. McNamara JO. Cellular and molecular basis of epilepsy. J
13. ChuK,Kang DW,Kim
Diffusion-weighted magnetic resonance imaging in non-
convulsive status epilepticus. Arch Neurol 2001;58:989—
JY,Chang KH, LeeSK.