Post-Stroke Epilepsy

Tom Skyhj Olsen, MD, PhD

optimism is building after the shocking experience of a

devastating stroke. In fact, some patients consider the complicating seizure and the subsequent epilepsy diagnosis the
most severe consequence of the stroke.

Address
Department of Neurology, Gentofte University Hospital,
DK 2900, Hellerup, Denmark.
E-mail: tso@dadlnet.dk
Current Atherosclerosis Reports 2001, 3:340344
Current Science Inc. ISSN 1523-3804
Copyright 2001 by Current Science Inc.

Seizures

Seizures occur in about 10% of stroke patients. Hence,

stroke is the most common cause of seizures and epilepsy
in the elderly population. Five percent are early-onset
seizures (peak onset within the first day after the stroke)
and another 5% are late-onset seizures (peak onset within
6 to 12 months after the stroke). Epilepsy (ie, recurrent seizures) develops in 3% to 4% of the stroke patients (in about
one third of the patients with early-onset seizures and
about one half of the patients with late-onset seizures).
There is a strong positive correlation between stroke
severity and the risk of post-stroke seizures; the risk is very
low in mild strokes. Seizures are more common in
hemorrhagic stroke and in stroke with cortical
involvement. Whether this is due to the hemorrhagic
component or the cortical involvement per se, or a
reflection of more severe strokes among patients with
hemorrhagic strokes and lesions involving cortical structures, is not clear. The influence of seizures on outcome is
still a matter of controversy. Although epileptic seizures are
considered easy to control, this is not supported by
evidence from randomized controlled trials.

Introduction
Cerebrovascular disease, including stroke, is by far the
most common cause of epilepsy in the elderly population,
accounting for 30% to 50% of the epilepsy cases in that age
group [1]. In stroke, seizures occur in the very acute state
and even at the very onset, as well as years after the stroke.
It occurs in patients with large hematomas and infarcts, as
well as in patients with only subtle structural abnormalities of the brain. Hence, seizures and epilepsy are frequen tly enc o u n t e r e d i n t h e s t r o k e un i t , w h e r e a s
management of stroke often has to be dealt with in the
epilepsy clinic.
The occurrence of an acute seizure is experienced as a
serious complication in the acute state of stroke, and many
have suggested prophylactic antiepileptic medication in
order to avoid acute seizures. The patient experiences the
occurrence of a late-onset seizure in a period when new

Seizures following stroke are usually divided into earlyonset seizures occurring within the first 2 weeks after the
stroke, and late-onset seizures occurring more than 2 weeks
after the stroke.
In sizeable, prospective hospital populations, earlyonset seizures occurred within the first 2 weeks in about
5% of the patients. The exact numbers are 5.2% of 1897
stroke patients in the international multicenter Seizure
After Stroke (SAS) study [2], 4.9% of 811 Australian
stroke patients [3], 4.2% of 1197 stroke patients in the
Copenhagen Stroke Study [4], and 6.1% of 1381 stroke
patients in the French Dijon study [5]. In the French
Besancon study [6], the incidence was only 1.8% in 3205
patients, but the authors stress that this low figure may be
due to selection bias.
In the Copenhagen Stroke Study [4], 66% of the earlyonset seizures occurred within the first 24 hours (2.8% of
the patients). In the Australian study [3], 98% of the earlyonset seizures occurred within 48 hours after the stroke. In
the population-based Oxfordshire Community Stroke
Project [7], 2% of 675 stroke patients had seizures within
24 hours after the stroke. In a Spanish study [8] of 1093
stroke patients, 2.5% had seizures within the first 48 hours,
and in a population-based study from Rochester, Minnesota [9], 4.7% of stroke patients had seizures within the
first 24 hours after the stroke.
Late-onset seizures occurring more than 2 weeks after
the stroke also occur in approximately 5% of stroke
patients. The exact numbers were 3.6% of 1897 stroke
patients in the SAS study [2], 7.3% of the 658 patients in
the Oxfordshire Community Stroke Project [7], and 3.2%
of 3205 stroke patients in the Besancon study [6]. In the
population-based Rochester study [9], 5% of 535 patients
developed seizures more than 1 week after the stroke. The
reported incidence is, however, strongly influenced by the
length of observation.
On average, seizures (both early-onset and late-onset)
occur in about 10% of stroke patients. In the Oxfordshire
Community Stroke Project [7], the 5-year actuarial risk of a
post-stroke seizure in survivors was 11.5%. In the Rochester
study [9], the cumulative probability of developing initial

340

Post-Stroke Epilepsy Olsen

late seizures was 7.4% by 5 years and 8.9% by 10 years.

Compared with the general population, the risk of developing seizures is substantial after stroke. The relative risk of
seizures in comparison with the general population was
estimated at 35 in the first year after stroke and 19 in the
second year after stroke in the Oxfordshire Community
Stroke Project [7]. In the Rochester study [9], the risk of
developing late-onset seizures (after 1 week) was increased
by 23 during the first year after the stroke, and the risk
remained increased over the next 3 years.

Recurrent SeizuresEpilepsy
In the Oxfordshire Community Stroke Project [7], recurrent seizures (ie, epilepsy) occurred in 3.8% of the
patients within an observation period of 5 years. In the
Rochester study [9], with a mean follow-up of 5.5 years,
3.4% (4.2% of the survivors) of the patients developed
epilepsy. In the SAS study [2], with a mean observation
period of 9 months, 2.5% of the patients developed
recurrent seizures and, therefore, epilepsy. In a number
of smaller studies, frequencies of 6% to 9% have been
reported [1013].
In an Australian study of 31 patients with early-onset
seizures occurring within 2 weeks after the stroke, 32%
had subsequently late seizures within an observation
period of 26 months [14]. In comparison, only 10% of 31
matched-stroke patients without early-onset seizures developed late seizures.
In a retrospective study of 90 patients with post-stroke
seizures with a mean follow-up of 30 months, 39% experienced seizure recurrence [15]. Half of those with earlyonset seizures and one third of those with late-onset seizures had recurrent seizures. Of the patients with recurrent
seizures, 43% had only a single recurrence whereas 57%
had multiple recurrences.
In the Besancon study [6], 135 patients with poststroke seizures were followed-up on average 47 months
post-stroke. Fifty percent of the patients experienced
recurrence of seizures: one third of the patients with
early-onset seizures and half of the patients with lateonset seizures; 18% had only one recurrence whereas
32% had more than two recurrences (multiple recurrence). In the study by Lancman et al. [16], with a mean
follow-up period of 12 months, 10% of 219 patients had
post-stroke seizures (early and late) and 3% had recurrent seizures. In the Rochester study [9], 21% of 23
patients with early seizures who survived more than 1
week developed epilepsy, whereas 59% of 23 patients
with initial late seizures developed epilepsy. In the
Oxford Community Stroke Project [7], 36% of 14
patients with early-onset seizures (within 24 hours after
the stroke) went on to have one or more post-stroke
seizures; 58% of 43 patients with late-onset seizures
(after 24 hours) had post-stroke seizures, but they
occurred infrequently.

341

In conclusion, 3% to 4% of all stroke patients develop

epilepsy as a consequence of the stroke, with about one
third of patients having early-onset seizures and about one
half of patients experiencing late-onset seizures.

Stroke Type
In the Copenhagen Stroke Study [4], early-onset seizures
(within 14 days) occurred in 8% of the patients with
hemorrhagic strokes and in 3% of the patients with
ischemic stroke. In the Australian study [3], 8% of the
patients with hemorrhagic stroke and 4% of the patients
with ischemic strokes had early-onset seizures within the
first 14 days after the stroke. In the Dijon study [5], earlyonset seizures within 48 hours were seen in 4.3% of the
patients with hemorrhagic stroke and in 2% of the
patients with ischemic stroke. In the Oxfordshire Community Stroke Project [7], onset-seizures within 24 hours
after the stroke were seen in 4% of the patients with
hemorrhagic stroke and in 1.8% of the patients with
ischemic stroke. In this study, the relative risk of seizures
was increased in patients with hemorrhagic strokes; the
hazard ratio for intracranial hemorrhage versus infarction being 10. In the SAS study [2], 10.6% of the
patients with hemorrhagic stroke had seizures, whereas
this was seen in 8.6% of the patients with ischemic
stroke. In the Besancon study [6], early-onset seizures
occurred in 9.3% of the patients with hemorrhagic
strokes and in 4.2% of the patients with ischemic stroke.
Thus, patients with hemorrhagic strokes are at a greater
risk for developing post-stroke seizures.
Among 112 consecutive patients with intracerebral
hemorrhage [17], 17% had seizures and all of these seizures occurred at or within a short time of hemorrhage
onset (mean follow-up of 60 days). Faught et al. [18] prospectively studied 123 patients with intracerebral
hematomas for an average time of 4.6 years and found
that 25% had seizures, and in half of the patients they
occurred with in the first 24 hours. In a study by Weisberg et al. [19], with a follow-up of 12 months, seizures
developed in 15% of 222 patients with intracerebral
hematomas. In 12% of patients, the seizures occurred
within 72 hours of the stroke. These findings are all in
agreement with the finding of a higher incidence of seizures among patients with hemorrhagic stroke. Although
this was also the finding in the Copenhagen Stroke Study
[4], it appeared from this study that hemorrhage per se
did not predict a higher risk of developing post-stroke
seizures. When controlling for relevant confounders and
potential predictors, hemorrhage per se did not predict
seizures. The higher risk of seizures in patients with hemorrhagic stroke was instead correlated to stroke severity,
which is definitely more severe in patients with hemorrhagic stroke. Accordingly, in the SAS study [2],
hemorrhagic infarction did not predict a higher risk of
post-stroke seizures.

342

Cardiovascular Disease and Stroke

Stroke Severity
Seizures are strongly related to size of the stroke lesion. In
the Copenhagen Stroke Study [4], mean infarct diameter of
patients with early-onset seizures was 62 mm, whereas it
was 39 mm in patients without early-onset seizures. In the
SAS study [2], infarct volume in patients with seizures
(early/late-onset) was 77 cm3, whereas it was 46 cm3 in
patients without seizures.
Stroke severity is likewise strongly related to occurrence
of seizures. In the Copenhagen Stroke Study [4], stroke
severity was graded into four areas, according to the Scandinavian Stroke Scale (SSS). In patients with mild stroke
(SSS score of 45 to 58, 41% of the 1197 patients) only 1%
of the patients went on to have early-onset seizures,
whereas in patients with very severe stroke (SSS score of 0
to 14, 19% of the patients) 10% had early-onset seizures.
In the SAS study [2], patients who had seizures had a
mean Canadian Neurological Score of 67 compared with a
score of 75 in the patients who did not develop seizures.
In the Copenhagen Stroke Study [4], stroke severity in a
multivariate analysis was the only predictor of early-onset
seizures, whereas lesion location and lesion type was not.
Also in the SAS study [2], stroke severity was a predictor
of post-stroke seizures, but in this study this was the case
with cortical localizations as well.
The strong relation between stroke severity and size of
lesion appears also from the fact that the risk of post-stroke
seizures is very low in lacunar infarction. In the Dijon
study [5], the incidence was six times lower in lacunar infarction compared with all other infarcts. In the Australian
[3] and in the Spanish [8] studies, early-onset seizures did
not occur in patients with lacunar infarcts, and the incidence was only 3% in the Oxfordshire Community Stroke
Project [7].

Lesion Location
Most seizures are produced by lesions affecting the cortex
[10,16,17]. In the Australian study [3], seizures occurred
exclusively in patients with lesions involving cortical structures. In the Spanish study [8], however, a multivariate
analysis that did not include stroke severity in the analysis
showed that cortical involvement leads to a sixfold relative
increase of the risk of seizures. In the Besancon study [6],
87% of the patients with seizures had lesions with cortical
localization. In the SAS study [2], a multivariate analysis
showed cortical localization to be a predictor of poststroke seizures in patients with infarcts (hazard ratio of
2.1), as well as in patients with hemorrhagic strokes (hazard ratio of 3.2). In the Copenhagen Stroke Study [4], cortical localization was a strong predictor of seizures in a
univariate analysis, but in a multivariate analysis, where
among other factors stroke severity was included in the
analysis, cortical localization as well as stroke subtype were
outnumbered by stroke severity. Strokes with cortical
involvement are definitely much more severe than strokes

caused by deeply located lesions. Hence, the relation

between cortical localization and development of seizures
could be linked to stroke severity or the amount of tissue
disturbance/destruction. The role of the cortex in the development of post-stroke seizures, therefore, is still not clear.

Seizure Type
Partial seizures are dominating in stroke patients. In the
Copenhagen Stroke Study [4], 68% of the early-onset
seizures were partial seizures or partial seizures with secondary generalization, whereas 22% were primary generalized
seizures. In the Australian study [3], 48% of the early-onset
seizures were partial and 11% were secondarily generalized.
In the SAS study [2], 52% of the seizures were partial or
partial with secondary generalization. In the Dijon study [5],
61% of the seizures were simple partial seizures, 28% were
secondarily generalized, and 11% were primarily generalized
seizures. In the Besancon study [6], 36% were partial, 28%
were partial with secondary generalization, and 37% were
presumed to be generalized. However, in the Spanish study
[8], primarily generalized seizures were most often reported,
as 62% of the early-onset seizures were primarily generalized
seizures, apparently without a focal onset.

Time of Onset
Early-onset seizures (within the first 2 weeks) occur most
often in the first 24 hours after the stroke. In the Australian study [2], 98% of early-onset seizures occurred
within 48 hours after the stroke, and more than half of
them occurred "shortly" after the stroke. In the Oxfordshire Community Stroke Project [7], 78% of early-onset
seizures occurred within the first 24 hours, and in the
Copenhagen Stroke Study [4], 66% occurred within the
first 24 hours and 86% within 72 hours. In series of
patients with hematomas [17,19], seizures are nearly
always reported to occur in close relation to the stroke
onset. In a study by Weisberg et al. [19] of 33 patients,
70% had seizures within the first 24 hours and 81% had
seizures within 72 hours. In another series of 19 patients,
all but one of the patients had their seizure prior to the
emergency room presentation [17].
Late-onset seizures occur most often within the first year,
with a peak 6 to 12 months after the stroke [2]. The risk of
having a seizure was 5.7% in the first year after stroke in the
Oxford Community Stroke Project [7], and 11.7% within the
first 5 years, with an incremental risk averaging 1.5% after the
first year. In the Besancon study [6], seizures of late-onset
type (after 2 weeks) occurred in 64% of the patients within
the first year and in 76% within 2 years.
Thus, time of onset of seizures after stroke has two
peaks: one during the first day and a second in the 6 to 12
months after the stroke. In the Rochester study [9], the
relative risk of having seizure was still increased 3 years
after the stroke.

Post-Stroke Epilepsy Olsen

Seizures Before Stroke

"Vascular precursive epilepsy" has been used as a term to
describe seizures predating stroke [20]. In a case-control
series, Shinton et al. [21] found this phenomenon in 4.5%
of stroke patients. The relevance of this phenomenon has,
however, been questioned by others [22]. In the Oxfordshire Community Stroke Project [7], seizures predated
stroke in 2% of the patients, but the phenomenon could
have been elicited by silent infarcts as well as other known
risk factors for stroke. For example, 29% of the stroke
patients included in the community-based Copenhagen
Stroke Study had silent infarcts [23].

Prognostic Influence
Whether or not the seizure per se is harmful to the brain
tissue in the stroke patient is still a matter of debate. Bogousslavsky et al. [24] described persistent worsening of neurologic
deficits following a seizure. In the Copenhagen Stroke Study
[4], patients with seizures had more severe strokes (Scandinavian Stroke Scale, 21 vs 37) and they more often died within
the hospital (50% vs 20%), but those who survived had the
same neurologic deficit at discharge (Scandinavian Stroke
Scale, 47 vs 49). However, in a multivariate analysis where
initial stroke severity was taken into account, patients with
early-onset seizures had the same relative risk of dying,
whereas the surviving patients with early-onset seizures had a
better neurologic outcome, equivalent to an increased score
of 6 points at the Scandinavian Stroke Scale. The authors
explained this surprising finding as the result of larger
ischemic penumbras in patients with early-onset seizures. A
large penumbra is related to a possible good outcome, but
also to an enhanced risk of seizures in metabolically
disturbed tissue because neurons in the penumbra are still
able to discharge. Initial stroke severity is determined by the
amount of dead tissue in the core of the infarct, as well as the
amount of nonfunctional but potentially viable tissue in the
penumbra. If neurons of the penumbra survive, a better outcome (gain) will be positively correlated with the relative size
of the penumbra. A Spanish study [25] came to the opposite
conclusions: seizures at the onset of a first-ever stroke are an
independent prognostic factor for in-hospital mortality. The
study did not, however, include initial stroke severity
expressed by a validated score in their analysis.

Electroencephalogram
The clinical use of an electroencephalogram (EEG) in postinf a r c t i o n s e i z u r e s a n d e p i l e p sy i s l i m i t e d . I n t h e
Australian study [2], six of 26 patients had paroxysmal features. The rest were either normal or showed focal slowing. In
the Besancon study [6], the earlier the EEG was performed
after the seizure, the more epileptiform abnormalities were
observed; the EEGs were either normal or showed focal or
diffuse slowing in 87% of the recordings. Periodic lateralized
epileptic discharges (PLEDS) were seen in 3% of the record-

343

ings. In the SAS study [26], epileptiform changes were found

in 10% of seizure patients compared with 2% of nonseizure
patients. Sensitivity, as well as specificity, of epileptiform
changes of EEG was poor, as was the predictive value of EEG
for seizures. The authors conclude that the clinical use of
EEG is questionable and should not be used as a basis for
treating patients with antiepileptic medication.

Pathophysiology
The cause of epileptic seizures, early-onset as well as lateonset, is poorly understood. The ischemic penumbra surrounding the infarct is probably the substrate of earlyonset seizures. Epileptogenic processes, such as enhanced
release of excitotoxic glutamate, ionic imbalances, breakdown of membrane phospholipids, and release of free
fatty acids, all characterize the penumbra. In hemorrhagic
stroke, a combination of focal ischemia, blood products,
and the sudden development of a space-occupying lesion
may elicit seizure activity. The cause of late-onset seizures is
thought to be the epileptogenic effect of gliotic scarring.

Treatment
The effect of treating post-stroke epileptic seizures has not
been tested in randomized clinical trials. As only one third
of patients with early-onset seizures and one half of the
patients with late-onset seizures develop epilepsy, there is
no evidence justifying antiepileptic medication after the
first seizure. This is further stressed by the fact that there is
not yet any firm evidence to support the view that a single
seizure affects outcome unfavorably. Finally, the influence
of antiepileptic drugs on the recovery process is not clear.
If recurrent seizures develop, they are often few and
they are usually controlled by monotherapy [14,15], using
drugs like carbamazepine, oxcarbazepine, phenytoin,
valproic acid, and lamotrigine [27]. In case of the very rare
cases of status epilepticus [15], most protocols include
intravenous benzodiazepines in conjunction with intravenous fosphenetoin loading [27].

Conclusions
Ten percent of patients with stroke experience a seizure event,
and in about half of them the seizure recurs and epilepsy
develops. Because stroke is so common, this is a leading cause
of epilepsy in the elderly. Based on this, it is surprising that so
little is known about the consequences of a seizure in the acute
state of stroke and about the effect of treatment. It is still not
known whether a seizure is harmful or not, so the question
about prophylactic antiepileptic treatment in the acute state of
the stroke is not yet settled. The effect of antiepileptic
treatment in respect to seizure control and in respect to poststroke function has not been tested in a randomized controlled trial. Although much is known about the epidemiology
of post-stroke seizures, research in this area is still needed.

344

Cardiovascular Disease and Stroke

References and Recommended Reading

Papers of particular interest, published recently, have been
highlighted as:

Of importance
Of major importance
Krmer G: Epilepsy in the elderly. Stuttgart: Georg Thieme
Verlag; 1999.
2. Bladin CF, Alexandrov AV, Bellavance A, et al. for the Seizure
After Stroke Study: Seizures after stroke. A prospective
multicenter study. Arch Neurol 2000, 57:16171622.
Sizeable study, with well-described patients using
multivariate statistics.
3. Kilpatrick CJ, Davis SM, Tress BM, et al.: Epileptic seizures in
acute stroke. Arch Neurol 1990, 47:157160.
4. Reith J, Jrgensen HS, Nakayama H, et al.: Seizures in acute
stroke: Prediction and prognostic significance. The
Copenhagen Stroke Study. Stroke 1997, 15851589.
5. Giroud M, Gras P, Fayolle H, et al.: Early seizures after acute
stroke: a study of 1,640 cases. Epilepsia 1994, 35:959964.
6. Berges S, Moulin T, Berger E, et al.: Seizures and epilepsy
following strokes: recurrence factors. Eur Neurol
2000, 43:38.
7. Burn J, Dennis M, Bamford J, et al.: Epileptic seizures after first
stroke: the Oxforshire community stroke project. BMJ
1997, 315:15821587.
8. Arboix A, Garcia-Eroles L, Massons JB, et al.: Predictive factors
of early seizures after acute cerebrovascular disease. Stroke
1997, 28:15901594.
9. So EL, Annegers JF, Hauser WA, et al.: Population-based study
of seizure disorders after cerebral infarction. Neurology
1996, 46:350355.
10. Olsen TS, Hgenhaven H, Thage O: Epilepsy after stroke.
Neurology 1987, 37:12091211.
11. Louis S, McDowell F: Epileptic seizures in nonembolic
cerebral infarction. Arch Neurol 1967, 17:414418.
12. Fentz V: Epileptic seizures in patients with cerebro-vascular
accidents. Nordisk Medicin 1971, 86:10231025.
13. Richardson EP, Dodge PR: Epilepsy in cerebral vascular
disease. Epilepsia 1954, 3:4974.
14. Kilpatrick C, Davis SM, Hopper JL, Rossiter SC: Early seizures
after acute stroke. Risk of late seizures. Arch Neurol
1992, 49:509511.

15.
16.

17.

1.

18.
19.

20.

21.

22.
23.

24.

25.

26.

27.

Gupta SR, Naheedy MH, Elias D, Rubino FA: Postinfarction

seizures. A clinical study. Stroke 1988, 19:14771481.
Lancman ME, Golimstok A, Norscini J, Granillo R: Risk factors
for developing seizures after a stroke. Epilepsia
1993, 34:141143.
Berger AR, Lipton RB, Lesser ML, et al.: Early seizures following
intracerebral hemorrhage: Implications for therapy. Neurology
1988, 38:13631365.
Faught E, Peters D, Bartolucci A, et al.: Seizures after primary
intracerebral hemorrhage. Neurology 1989, 39:10891093.
Weisberg LA, Shamsnia M, Elliott D: Seizures caused by nontraumatic parenchymal brain hemorrhages. Neurology
1991, 41:11971199.
Pohlman-Eden B, Cochius JI, Hoch DB, Hennerici MG. Stroke
and epilepsy: critical review of the literature. Part II: Risk factors, pathophysiology and overlap syndromes. Cerebrovasc Dis
1997, 7:29.
Shinton RA, Gill JS, Zezulka AV, Beevers DG: The frequency of
epilepsy preceding stroke. Case-control study in 230 patients.
Lancet 1987, 1:1113.
Starkey IR, Warlow CP: Epilepsy preceding stroke [letter].
Lancet 1987, 1:742743.
Jrgensen HS, Nakayama H, Raaschou HO, Olsen TS: Silent
infarction in first-ever stroke patients. Stroke
1994, 25:22932294.
Bogousslavsky J, Martin R, Regli F, et al.: Persistent worsening
of stroke sequelae after delayed seizures. Arch Neurol
1992, 49:385388.
Arboix A, Comes C, Massons J, et al.: Relevance of early
seizures for in-hospital mortality in acute cerebrovascular
disease. Neurology 1996, 47:14291435.
Bladin CF, Norris JW: Stroke and seizure/epilepsy. In Primer on
Cerebrovascular Diseases. Edited by Welch KMA, Caplan LR, Reis
DJ, Siesj BK, Weir B. San Diego: Academic Press;
1997:355358.
Smith B: Management of seizures in acute stroke. In Primer on
Cerebrovascular Diseases. Edited by Welch KMA, Caplan LR, Reis
DJ, Siesj BK, Weir B. San Diego: Academic Press;
1997:706708.