Epilepsy and brain abscess

Christine Kilpatrick MD FRACP

Department of Neurology, The Melbourne Neuroscience Centre, The Royal Melbourne Hospital, Parkville, Australia

Although epilepsy is a well recognized complication o f brain abscess, the true incidence o f seizures and the factors
which predispose to seizure occurrence and recurrence are not well established. This study retrospectively assessed
the incidence of seizures and seizure recurrence in 35 consecutive adult patients diagnosed surgically with brain
abscess between 1984 and 1994. Following diagnosis, all patients were treated with phenytoin. There was no
relationship between the site of the abscess, organism cultured, surgical treatment, presumed aetiology, age or sex o f
the patient and seizure occurrence. It is concluded that epilepsy is a c o m m o n complication o f cerebral abscess,
frequently occurring at presentation. Early seizures predispose to late seizures and in these patients long-term
anticonvulsant treatment should be considered. If a patient remains seizure free at discharge, the chance of
developing seizures is relatively small.
Journal of Clinical Neuroscience ] 997, 4 (1) : 26-28 © Pearson Professional Ltd 1997

Keywords: brain abscess,seizures, early, late

Introduction
Epilepsy has traditionally b e e n t h o u g h t to be a c o m m o n Follow-up of patients ranged from 0.5 months to 108
complication of brain abscess such that most patients months, with a mean follow-up of 31 months. Patients
with this condition are treated with prophylactic anti- were followed to J u n e 1994 (20 patients), or death
convulsant therapy. T h e r e p o r t e d incidence of seizures (10 patients), and in 5 patients who were lost to follow-up
however varies, ranging from as low as 3% to as high as a more limited follow-up was achieved, with a mean o f 29
95%, with most r e p o r t i n g approximately 35%. 1-9 Apart months. Ten patients died during the follow-up period.
from abscess site, the factors which predispose to Six patients died as a result of the acute illness. All had
seizure o c c u r r e n c e and the incidence of seizure recur- associated medical problems, chronic renal failure (1),
rence in patients presenting with brain abscess are not acute leukaemia (1), polyarteritis nodosa on long-term
well established. T h e aims o f this study are to assess the immunosuppressants (1), associated spinal abscess (1),
incidence of seizures, both early (within 2 weeks o f and two elderly patients (77 and 81 years old) developed
presentation) and late (more than 2 weeks after pre- p n e u m o n i a following initial improvement. Two patients
sentation), and seizure r e c u r r e n c e in adult patients died secondary to unrelated illnesses and a further 2
with brain abscess, and identify those patients with a patients died related to HIV infection, but unrelated to
higher risk of seizures and h e n c e the n e e d for long- the acute illness.
term anticonvulsant treatment. O f the 35 patients, 30 had intracerebral, 3 extradural,
1 subdural and 1 cerebellar abscess. T h e male:female
ratio was 8:1 and mean age 51 years (range 13-81 years).
Patients and methods T h r e e patients had HIV infection.
Thirty-five consecutive adults diagnosed surgically with
brain abscess between January 1984 and January 1994 at Statistical analyses
The Royal Melbourne Hospital, were assessed retrospec-
tively. Information was sought from hospital medical The difference between two i n d e p e n d e n t proportions
records, treating physicians and direct contact with was tested using Fisher's exact test. T h e proportion of
patients or relatives. All patients had a head computed patients with early seizures who had late seizures was
tomography (CT) scan and some patients had a magnetic compared with the proportion of patients without early
resonance image (MRI) scan. All patients were treated seizures who developed late seizures.
with phenytoin whether or not they had a seizure, follow-
ing presentation and prior to surgical treatment. Only
Results
patients in whom the diagnosis was confirmed pathologi-
cally were included. Seventy percent of patients were O f the 34 supratentorial abscesses, 12 (35%) had seizures.
recruited from 2989 to December 1993. Seizures did not complicate the one case of cerebellar

26 J. Clin. Neuroscience Volume 4 Number 1 January 1997

Epilepsy and brain abscess Clinical studies

abscess. O f the 12 patients with seizures, in 6 the seizures Table 3 Organism cultured and seizure occurrence
were partial and in 6 clinically generalized. O f the 12
Bacterial Fungal No growth Parasitic
patients with seizures, 10 were early and 2 late in onset. O f
the 10 patients with early seizures, in 9 patients seizures Seizures 8 1 3 0
No seizures 14 6 2 1
occurred at presentation, and in one case 10 days postop-
eratively. O f the 2 patients with late seizures without early Total 22 7 5 1
seizures, in 1 the seizure occurred 18 months postdiagno-
sis and in the other case 3 years after presentation, when
the patient presented with cerebral metastases. In this
case the seizures were probably unrelated to the abscess.
The clinical features of the patients with early seizures Table 4 Surgical t r e a t m e n t and seizure occurrence
were similar to those without early seizures. Excision Drainage
Seizures r e c u r r e d in 7 of the 10 patients with early
Seizures postsurgery 3 7
seizures; most were infrequent a n d readily controlled.
Seizures were controlled in 6 of the 7 patients by either Total 13 22
increasing the dose o f p h e n y t o i n or introducing a second
antiepileptic drug. The 6 patients had only one or two
r e c u r r e n t seizures during the follow-up period. O n e
patient was not con'trolled despite the introduction of
a second antiepileptic drug. Late seizures were signifi-
Nine o f the 10 patients with early seizures with or with-
cantly m o r e c o m m o n in patients with early seizures
out late seizures r e m a i n e d on long-term anticonvulsant
(7 of 10) than in those without early seizures (2 o f 25)
therapy. Ten patients who h a d b e e n seizure-free ceased
(P< 0.001) (Table 1).
anticonvulsant t r e a t m e n t 12-24 m o n t h s after surgery.
In this study no relationship was f o u n d between the
These patients were followed for a m e a n period of 19
site of the snpratentorial abscess and seizure o c c u r r e n c e
m o n t h s after ceasing therapy. Only I of these 10 patients
(Table 2). Seizures, however, were not associated with
h a d a late seizure, 6 m o n t h s after ceasing phenytoin.
infratentorial abscess. T h e r e was no correlation between
the organism cultured and the o c c u r r e n c e of seizures.
Streptococcus milleri was the most c o m m o n organism Discussion
cultured (Table 3). T h e type of surgical t r e a t m e n t was
Epilepsy is a relatively c o m m o n complication of cerebral
not a risk factor for d e v e l o p m e n t of subsequent seizures
abscess, occurring in 33% of patients in this c u r r e n t
(Table 4). T h e r e was no relationship between p r e s u m e d
study, c o m p a r a b l e to previous r e p o r t s ) -6, s, 9 Some studies,
aetiology and seizure o c c u r r e n c e (Table 5).
however, have r e p o r t e d a higher incidence of seizures.
T h e r e was no relationship with age or sex of a patient
This study aimed to identify those patients with cere-
and seizure recurrence (Table 6). A predisposition to
bral abscess who have a higher risk of seizure occurrence
abscess f o r m a t i o n was evident in 27 of the 35 patients. O f
and h e n c e should be maintained on long-term an ticon-
note, bacterial infection was m o r e c o m m o n in this group
vulsant treatment. Results of this study suggest that those
(19 of 27), whereas a fungal aetiology was m o r e c o m m o n
patients with early seizures have a high risk of late seizures
in those without a recognized predisposing factor (5 of 8).
and hence should remain on long-term anticonvulsant
therapy. Previous studies have also r e p o r t e d early seizures
as a risk factor for late seizures in patients with cerebral
abscess.5, a0 Others, however, have not shown this relation-
ship.S, 11 In contrast to a recent report, however, s the cur-
Table I Early seizures and t h e risk of late seizures rent study did not identify age, sex, site or size of abscess
Early seizures No early seizures as risk factors for seizure occurrence. This may in part be
due to the smaller n u m b e r of patients studied.
Late seizures 7 2
As m a n y patients r e m a i n e d on anticonvulsant treat-
No late seizures 3 23
m e n t in the long-term, it could be argued that this
Total 10 25
accounts f o r the relatively low incidence of late seizures.

Table 2 Abscesssite and seizure occurrence

Temporal Parietal Frontal Occipital Multiple Cerebellar

Seizures 1 3 6 1 1 0
No seizures 3 7 10 1 1 1
Total 4 10 16 2 2 1

J. Clin. N e u r o s c i e n c e Volume4 Number 1 January 1997 27

Clinical studies Epilepsy and brain abscess

Table 5 Presumed aetiology and seizure occurrence

Spread of contiguous infection Trauma Postcraniotomy Haematogenous Unknown
Seizures 0 1 2 5 4
No seizures 4 1 2 6 10
Total 4 2 4 11 14

Table 6 Age and sex of patient and seizure occurrence remain seizure-free at discharge, however, the findings sug-
gest the chance of developing seizures is relatively small and
Age (Years) Sex
in these patients, cessation of therapy after 24 months could
13-40 41-59 60-81 Male Female be considered.
Seizures 4 5 3 10 2
No seizures 9 8 6 21 2 Acknowledgements
Total 13 13 9 31 4 The author wishes to acknowledge Professor Andrew Kaye,
Director of Neurosurgery, The Royal Melbourne Hospital,
for allowing these patients to be studied.

Received 24 March 1995; Accepted 11 August 1995

Correspondence and offprint requests: Dr C.J. Kilpatrick,

The Melbourne Neuroscience Centre, Royal Melbourne Hospital,
O f note, however, 9 o f 10 patients with early seizures c o n - Parkville, Victoria 3050, Australia, Tel: 61 3 9342 8448,
tinued on anticonvulsant t r e a t m e n t and yet a high per- Fax: 61 3 9342 8427
centage of this g r o u p laad late" seizures, whereas a
n u m b e r of patients without early sei2ures later ceased References
t r e a t m e n t and only one patient developed late seizures. 1. Dohrmann P, EMck W. Observations on brain abscess:
Although the n u m b e r s are small, in this study the risk of review of 28 cases. MedJ Aust 1982; 2: 81-83.
2. Cowie R,.Williams B. Late seizures and morbidity after
late seizures or seizure r e c u r r e n c e was not related to stop- subdural empyema.J Neurosurg 1983; 58: 569-573.
ping anticonvulsant treatment. 3. Nielsen H, Harmsen A, Gyldensted C. Cerebral abscess. A
it is possible that the incidence of late seizures would long-term follow-up. Acta Neurol Scand 1983; 67: 330-337.
have b e e n greater h a d the patients b e e n followed for a 4. Bradley PJ, Manning KP, Shaw MD. Brain abscess secondary
longer period of time. Northcroft and Wyke 1° p r o p o s e d to otitis media.J Laryngol Otol 1984; 98: 1185-1191.
5. Renier D, Flandin C, Hirsch E, HirschJF. Brain abscess in
that the incidence o f seizures increased with time and neonates. A study of 30 cases. J Neurosurg 1988; 69: 877-882.
that a m e c h a n i s m of fibroglial reaction resulted in scar 6. Ibrahim AW, al-Rajeh SM, Chowdhary UM, Ammar A. Brain
f o r m a t i o n responsible for the seizures. This mechanism, abscess in Saudi Arabia. Neurosurg Rev 1990; 13: 103-107.
however, does not account for the high incidence of 7. Aebi C, Kaufmann F, Schaad UB. Brain abscess in childhood:
long-term experience. EurJ Pediatr 1991; 150: 282-286.
seizures at presentation, which was seen in the present
8. Koszewski W. Epilepsy following brain abscess. The
study and in other reports. evaluation of possible risk factors with emphasis on new
In the c u r r e n t study there was a striking male pre- concept of epileptic focus formation. Acta Neurochir
dominance. All previous reports of cerebral abscess 1991; 113: 110-117.
where the sex of the patients has b e e n indicated, have 9. Yang S-Y, Zhao C-S. Review of 140 patients with brain
r e p o r t e d a male p r e d o m i n a n c e ) , 6-9, 13 None, however, abscess. Surg Neurol 1993; 39: 290-296.
10. Northcroft GB, Wyke BD. Seizures following surgical
has r e p o r t e d a ratio as high as that in the current study. treatment of intracranial abscesses. J Neurosurg 1957; 14:
T h e reason for this is unclear and the coexistence of HIV 249-263.
infection does n o t explain this finding. 11. Legg NJ, Gupta PC, Scott DE Epilepsy following cerebral
The results of this study indicate that epilepsy is a com- abscess. A clinical and EEG study of 70 patients. Brain
1973; 96: 259-268.
m o n complication of cerebral abscess, frequently occurring
12. Nicolosi A, Hauser WA, Musicco M, Kurland LT.
at presentation. Early seizures predispose to the develop- Incidence and prognosis of brain abscess in a defined
m e n t of late seizures and in these patients long-term anti- population: Olmsted County, Minnesota 1935-1981.
convulsant treatment should be considered. In patients who Neuroepidemiology 1991; 10: 122-131.

28 J. Clin. Neurosdence V o l u m e 4 N u m b e r 1 January 1997