Journal of Psychosomatic Research 61 (2006) 343 – 347

Clinical predictors of intractable childhood epilepsy

Aithala Gururaja,4, Laszlo Sztrihab, Joseph Hertecantb, Valsamma Eapenc
a
Department of Paediatrics, Faculty Perubatan, Universiti Teknologi MARA, Shah Alam, Selangor, Malaysia
b
Department of Paediatrics, Faculty of Medicine and Health Services, United Arab Emirates University and Tawam Hospital, Al Ain, United Arab Emirates
c
Department of Child Psychiatry, Faculty of Medicine and Health Services, United Arab Emirates University, Al Ain, United Arab Emirates
Received 7 July 2006

Abstract
Objective: This study aimed to determine the clinical, electro-
encephalographic, and radiological factors associated with medi-
cally intractable seizures in children in the Al Ain Medical District
in the United Arab Emirates. Methods: This work used a
prospective case-control study of children referred to pediatric
neurology and neurodevelopmental clinics at Tawam and Al Ain
University Hospitals. Results: There were 55 children with
intractable epilepsy; their data were compared with 50 children
who responded well to antiepileptic drugs and who were seizure-
free for at least 2 years. Onset b1 year of age, a high seizure
frequency at onset, positive history of neonatal seizures, devel-
opmental delay and status epilepticus, neurological deficits, and
abnormal brain imaging results were found to be significantly more
common in the study group. Symptomatic localization-related
epilepsy was more common in children in this group than in the
control group. Conclusion: Our study suggests that children who
present with idiopathic localization-related and generalized epi-
lepsy syndromes with few seizures at onset and with no neuro-
logical deficits tend to have a relatively good prognosis.
D 2006 Published by Elsevier Inc.

Keywords: Childhood epilepsy; Seizures; Clinical factors; United Arab Emirates

Introduction with three or more AEDs, used alone, serially, or in

Epilepsy is one of the most common neurological
conditions in children. In the United States, approximately
125,000 new cases of epilepsy are seen each year; 30% of
this group will be younger than 18 years [1,2]. The largest
number of newly diagnosed patients with epilepsy is found
in the age group between birth and 2 years. A large majority
of these children with epilepsy respond very well to
monotherapy with antiepileptic drugs (AEDs); a number
of others have epilepsies that are, however, difficult to
control even with two or more AEDs. It is estimated that
more than 30% of patients become refractory to medical
therapy [3–6]. Intractable epilepsy (IE) is defined as
continued seizures in children despite adequate therapy
combinations [4,7]. There are few studies that have
attempted to determine the possible risk factors of intract-
ability in childhood epilepsy [8–14]. There is no published
data regarding IE in the Arab population.
This study was designed to determine the characteristic
features of IE in the Emirati, Arab, and Asian expatriate
children seen over a 5-year period in the pediatric neurology
referral clinics at Al Ain and Tawam hospitals in the Al Ain
Medical District.
Methods
This was a prospective controlled study of children
referred to pediatric neurology and neurodevelopmental
clinics at Tawam and Al Ain University Hospitals.

4 Corresponding author. Faculty Perubatan, Universiti Teknologi Inclusion criteria

MARA, Aras 20, Menara 1, Komplex Sains & Teknologi, 40450 Shah
Alam, Selangor, Malaysia. Tel.: +60 19 255 9689. From among the children diagnosed with epilepsy
E-mail address: g.aithala@gmail.com (A. Gururaj). between 2 months and 13 years of age who were referred

0022-3999/06/$ – see front matter D 2006 Published by Elsevier Inc.

doi:10.1016/j.jpsychores.2006.07.018
344 A. Gururaj et al. / Journal of Psychosomatic Research 61 (2006) 343–347

to pediatric neurology clinics from December 1997 to Table 1

General characteristics of Group 1 (with intractable seizures; n=55) and
December 2002, only the ones whose first unprovoked
Group 2 (responders; n=50)
seizure occurred after the first month of life were included.
Group 1, n (%) Group 2, n (%)
The criteria for IE were met if a patient had at least one
seizure per month over an observation period of 2 years or Sex
Male 30 (54.5) 29 (58)
more, despite treatment with three or more AEDs admin-
Female 25 (45.5) 21 (42)
istered singly or in combination. Patients with poor Age (years)
compliance or with low serum AED levels were excluded. 2–5 18 (33) 2 (4)
Detailed history as noted from the parents and seizure 5–10 20 (36) 25 (50)
semiology was evaluated, and deficits on neurological N10 17 (31) 23 (46)
Time of first seizure
examination were recorded by a pediatric neurologist and
b12 months 44 (80) 1 (2)
a pediatrician experienced in epileptology (LS and GA). The 1–5 years 7 (13) 29 (58)
EEG records of the patients were reviewed (by LS and GA), 5–10 years 4 (7) 20 (40)
and the brain imaging studies were visually analyzed by a Type of onset
trained pediatric neurologist (LS). The patients were treated Generalized tonic/clonic/tonic clonic 40 (73) 24 (48)
Myoclonic 22 (40) 4 (8)
according to the internationally accepted regimens.
Partial 5 (9) 17 (34)
A control group consisted of children with well-controlled Absences 2 (4) 4 (8)
epilepsy, defined as having achieved at least 2 seizure-free Atonic 1 (2) 1 (2)
years before the last follow-up. Clinical and neurological Initial frequency/day
evaluations were done as part of the clinical work, and verbal Less than 5 15 (27) 5 (10)
5–10 7 (13) 1 (2)
consent was obtained prior to interviewing parents.
10–20 or more 15 (27) 0
Total 37 (67) 6 (12)
Definitions Weekly 13 (24) 8 (16)
Monthly 5 (9) 36 (72)
Seizures were classified using the International League Drug therapy (drugs/day)
1–2 0 50 (100)
Against Epilepsy (ILAE) classification of epileptic seizures,
3 30 (55)
and epilepsy and epilepsy syndromes were classified using 4 13 (23)
the ILAE classification (1989) [15,16]. Remote sympto- 5 or more 12 (22)
matic cases meant epilepsies with known or identifiable Outcome
cause or pathology in the brain; cryptogenic cases described No seizures 0 50 (100)
Daily seizures 16 (29)
those that are probably the result of some undetermined
Weekly seizures 10 (18)
brain disorder, whereas idiopathic cases represented those Monthly seizures 29 (53)
where the seizures are due to possible genetic propensity to
generate seizures [3,13].

Data analysis
Out of 550 children with epilepsy studied during the
period, there were 76 cases (14%) that could be termed as
IE. Out of this, only 55 cases were eligible for the study
since follow-up data were insufficient, drug compliance was
poor, or clinical or EEG data were incomplete in the other
cases. EEG was not possible in some of these children in
spite of sedation. Their data were compared with 50 children
who responded well to AED and who were seizure-free for
at least 2 years.
Results
General characteristics
Majority of the patients in both groups were UAE
nationals (73% and 60% in the study and control groups,
respectively). The age at presentation varied from 2 to
12 years in both groups. Significantly, the onset of the
seizures was in the first 12 months in 80% of the children
with IE as compared with only 2% in the control group. In
the study group, 73% of the children had a generalized
tonic/clonic/tonic clonic onset of seizures and 40% had a
myoclonic presentation, whereas 48% of the children in the
control group had a generalized tonic clonic onset and only
8% had myoclonic seizures at onset. The initial seizure
frequency tended to be quite high in the study group, with
more than 67% of the children experiencing daily seizures
and with the frequency varying from 5 to more than
20 episodes daily. In the control group, only 12% of the
children had daily seizures, which responded promptly to
medications. The children in the study group needed three
or more AEDs, with 22% of the children requiring five or
more medications to improve seizure frequency. In spite of
this, 47% of the children continued to have daily or weekly
seizures (Table 1).
A comparison of both groups illustrates the difference
between the children in the groups in the following variables
(Table 2). A history of hypoxic ischemic encephalopathy
and/or neonatal seizures and status epilepticus at a later age
 A. Gururaj et al. / Journal of Psychosomatic Research 61 (2006) 343–347 345

Table 2 Table 4
A comparison of case and control groups Classification of epilepsy
Group 1, Group 2, Group 1 Group 2
Factors n (%) n (%) P value Type of epilepsy/epilepsy syndrome (n=55) (n=50)
Age of onset b5 years 18 (32) 2 (4) .001 Localization-related epilepsy/epilepsy syndromes
Time of first seizure b12 months 44 (80) 1 (2) b.001 Idiopathic 1 20
Initial seizure frequency N5/day 22 (40) 1 (2) b.001 Cryptogenic 3 10
Type of onset Symptomatic 23 5
Generalized tonic/ clonic/tonic clonic 40 (73) 24 (48) b.001 Generalized epilepsy/epilepsy syndrome
Partial 5 (9) 17 (34) Idiopathic 10 11
Myoclonic/infantile spasms 22 (40) 4 (8) Cryptogenic/symptomatic 16 2
History of status epilepticus 27 (49) 4 (8) b.001 West syndrome — cryptogenic 1 0
History of febrile seizures 17 (31) 7 (14) .039 West syndrome — symptomatic 3. 0
Developmental delay 52 (94.5) 7 (14) b.001 Lennox Gastaut syndrome 3 0
Associated ADHD 15 (27) 5 (10) .027 Myoclonic astatic seizures 6 2
Neonatal seizures 10 (18) 0 .001 Early infantile myoclonic encephalopathy 3 0
History of HIE 14 (25) 3 (6) .008 Undetermined 2 0
Family history of seizures 9 (16) 10 (20) .800 Group 1 includes children with intractable seizures and the responders
Consanguinity 37 (67) 39 (78) .845 make up Group 2.
Abnormal EEG at onset 50 (91) 45 (90) .008
Abnormal brain imaging 49/55 (89) 6/41 (15) b.001
Neurological deficits 43 (78) 4 (8) .001 children in the control group. Of the possible causes among
Group 1 (n=55) includes children with intractable seizures and the the children with IE, perinatal events (hypoxic ischemic
responders make up Group 2 (n=50). encephalopathy, perinatal infections, and neonatal hyper-
bilirubinemia) were present in 27% of the children in the
was significantly common in the study group. Similarly, a
study group and in 4% of the children in the control group.
significant developmental delay was found in 94% of the
The other possible causes included structural malformations
children with IE as against 14% of the children in the control
(31%), tuberous sclerosis (7.3%), hippocampal sclerosis
group. Abnormal brain imaging was seen in 89% of the
(3.6%), and vascular malformations (5.5%; Table 3).
children in the study group but only in 14.6% of the children
in the control group. The children with IE were more likely
Classification of epilepsy/epilepsy syndrome
to have significant neurological deficits (78%, as against 8%
among the controls). However, the following variables were
As shown in Table 4, 70% of the children in the control
not found to be significantly different in the groups: a history
group had localization-related epilepsy/epilepsy syndrome.
of febrile seizure, presence of attention deficit hyperactivity
Among the children with IE, 48% had localization-related
disorder, and a family history of seizures or mental
epilepsy, a majority of which belonged to the symptoma-
retardation. Interestingly, parental consanguinity was pos-
tic group.
itive in the majority of the children in both groups.

Etiology of epilepsy
Discussion
The etiology of epilepsy was unclear or idiopathic in IE is a serious condition in children, leading to
21.9% of the children in the study group and in 92% of the significant impairment in quality of life (QOL), as well as
behavioral and psychiatric problems including depression
Table 3 and poor scholastic performance. Most patients become
Etiology of epilepsy in patients and controls severely emotionally affected, and coping with this chronic
Group 1, Group 2, handicap may pose a burden on the family as well as on
Cause n (%) n (%)
social, educational, and health services [7]. In addition, there
Idiopathic 12 (21.9) 45 (90) are numerous medical problems, some of which are
Perinatal 15 (27.2) 2 (4)
potentially life-threatening, arising as a result of uncon-
(HIE, neonatal infections/meningitis,
hyperbilirubinemia) trolled seizures including aspiration, cardiac arrhythmias,
Structural malformations of the brain 17 (30.9) 0 electrolyte imbalance, brain edema, renal failure, unex-
Tuberous sclerosis 4 (7.3) 0 plained sudden death, and refractory status epilepticus [8].
Hippocampal sclerosis 2 (3.6) 0 Besides, these children are more prone to drug toxicity and
Associated arachnoid cysts 2 (3.6) 2 (4)
drug interactions resulting from polypharmacy and are at
Vascular 3 (5.5) 0
Post-meningitis/encephalitis 2 (3.6) 1 (2) higher risk of developing behavioral and academic diffi-
beyond neonatal period culties [17].
Group 1 (n=55) includes children with intractable seizures and the The exact cause or the contributing factors for intract-
responders make up Group 2 (n=50). ability in childhood seizures are far from clear but are
346 A. Gururaj et al. / Journal of Psychosomatic Research 61 (2006) 343–347
believed to be complex [7]. Few studies done in the United
States, Canada, Western Europe, and India have identified
the following risk factors: early age of onset; gross EEG
abnormalities; simple partial, tonic, and myoclonic seiz-
ures; a history of status epilepticus; developmental delay
at onset; and abnormal brain imaging results from
computerized tomographic scans and magnetic resonance
imaging [4,8–13].
In the Al Ain Medical District, during the study period,
on average, 100 to 120 new cases of epilepsy in children
were seen annually in the two major tertiary-level
university hospitals. The average incidence was around
1 in 400 to 450 children. A number of cases of refractory
epilepsy are encountered in these hospitals. Earlier studies
from this region had indicated that the fairly high incidence
of refractory epilepsy in this population is because of
an increased prevalence of congenital malformations
of the brain and perinatal hypoxic ischemic brain injury
(HIE) [18,19].
In this study, the majority of the children with IE had
onset of seizure during infancy. This finding has been
observed in other studies [8,12,13]; this possibly reflects the
high incidence of HIE and congenital malformations of the
brain in this group of children. High initial seizure
frequency of N5 per day was another significant feature in
the study group. Coupled with an early onset of seizure, the
high seizure frequency at onset possibly results to greater
damage to the brain, resulting to intractability of seizures.
Two other factors, namely, a history of neonatal seizures and
status epilepticus, were found to be significantly positive in
the study group. This is in keeping with the findings of Berg
et al. [9], who found that the presence of these two variables
made the outcome in IE especially poor. Kwong et al.
suggested that status epilepticus should be considered as a
marker rather than as a cause of intractability in these
children since the outcome and morbidity after status
epilepticus have been shown to be dependent on the
underlying pathology rather than on the event itself
[12,20]. Furthermore, the presence of neurological deficits
is considered an important predictor of intractability
[8,12,13]. In this regard, our finding that 78% of the
children with IE had one or more neurological deficits as
against 8% in the control group is noteworthy. This possibly
reflects the severity of brain damage. Our finding of an
association between intractability and developmental delay
at onset may also support this notion of diffuse and severe
brain damage in these children. However, a more recent
study observed that a normal neurological examination
cannot adequately predict the response to treatment in
cryptogenic epilepsy in children [14].
The classification of epilepsy and epilepsy syndromes
offers an important predictive value [9]. As in most studies,
we found that 50% of the children with IE belonged to
localization-related symptomatic epilepsy and epilepsy
syndromes, but the risk was also high among children with
generalized cryptogenic/symptomatic syndromes. On the
other hand, the majority of the children in the control group
had localization-related idiopathic/cryptogenic syndromes.
A history of febrile seizure was not found to be
associated with intractability in this study. Similarly, there
was no association with a family history of seizures;
parental consanguinity was found in the majority of the
children in both groups.
It is also known that the etiology of IE plays an important
role in determining the prognosis. We were able to find an
association between specific brain pathology and seizures in
43 of 55 cases. Perinatal causes, including HIE, perinatal
infections, and hyperbilirubinemia, were found in 27% of
the children, and structural brain malformations were found
in 31%. The other causes included vascular pathology,
tuberous sclerosis, hippocampal sclerosis, and post-menin-
gitic sequelae beyond the neonatal period. As against this,
90% of the children in the control group had no identifiable
pathology. The high prevalence of structural brain malfor-
mations in this population has been observed by us earlier
[18]. The incidence of perinatal HIE is declining in the
Emirates due to improving maternal and child health
services; as compared with figures from India and China,
the percentage of children with HIE as a result of IE is much
less in this study [12,13].
Our study suggests that children who present with
idiopathic localization-related epilepsy syndromes with
few seizures at onset and with no neurological deficits tend
to have a relatively good prognosis.
It is important to identify those children who are at an
increased risk of developing IE early in their course in this
region so that a more appropriate and aggressive therapeutic
option may be designed early during the course of their
illness; newer AEDs may have a role in the treatment of
these children. On the other hand, in those children where
the prognosis is good, one should be able to taper off the
AED therapy early to avoid the potential side effects of the
drugs on the developing brain and to limit the costs involved
in long-term therapy [21,22]. Early surgical intervention
may be considered in some of these children to minimize
brain damage due to repeated seizures and to improve their
QOL [4,11].
References
[1] Hauser WA. Epidemiology of epilepsy in children. Neurosurg Clin N
Am 1995;6:419 – 29.
[2] Holmes GL, Engel J. Predicting medical intractability of epilepsy in
children: how certain can we be? Neurology 2001;56:1430 – 1.
[3] Aicardi J. Epilepsy in children, vol 12. Philadelphia7 Lippincott-
Raven, 2004. pp. 351 – 2.
[4] Ko T-S, Holmes GL. EEG and clinical predictors of medi-
cally intractable childhood epilepsy. Clin Neurophysiol 1999;110:
1245 – 51.
[5] Brodie MJ, Leach JP. Success or failure with antiepileptic drug
therapy. Neurology 2003;60:162 – 4.
[6] Kwan P, Brodie MJ. Early identification of refractory epilepsy. N Engl
J Med 2000;342:314 – 9.
 A. Gururaj et al. / Journal of Psychosomatic Research 61 (2006) 343–347
[7] Alving J. What is intractable epilepsy? In: Johannessen SI, editor.
Intractable epilepsy. Petersfield7 Wrightson Biomedical Publishing,
1995. pp. 1 – 12.
[8] Berg AT, Levy SR, Novotny EJ, Shinnar S. Predictors of intractable
epilepsy: a case control study. Epilepsia 1996;37:24 – 30.
[9] Berg AT, Shinnar S, Levy SR, Testa FM, Smith-Rapaport S,
Beckerman B. Early development of intractable epilepsy in children:
a prospective study. Neurology 2001;56:1445 – 52.
[10] Dlugos DJ, Sammel MD, Strom BL, Farrar JT. Response to first drug
trial predicts outcome in childhood temporal lobe epilepsy. Neurology
2001;57:2259 – 64.
[11] Casetta I, Granieri E, Monetti VC, Gilli G, Tola MR, Paolino E,
Govoni V, Lezzi E. Early predictors of intractability in childhood
epilepsy: a community based case-control study in Copparo, Italy.
Acta Neurol Scand 1999;99:329 – 33.
[12] Kwong KL, Sung WY, Wong SN, So KT. Early predictors of medical
intractability in childhood epilepsy. Pediatr Neurol 2003;29:46 – 52.
[13] Chawla S, Aneja S, Kashyap R, Mallika V. Etiology and clinical
predictors of intractable epilepsy. Pediatr Neurol 2002;27:186 – 91.
[14] Tang-Wai R, Oskoui M, Webster R, Shevell M. Outcomes in pediatric
epilepsy: seeing through the fog. Pediatr Neurol 2005;33:244 – 50.
347
[15] Commission on Classification and Terminology of the International
League Against Epilepsy (ILAE). Proposal for revised clinical and
EEG classification of epileptic seizures. Epilepsia 1981;22:489 – 501.
[16] Commission on Classification and Terminology of the International
League Against Epilepsy (ILAE). Proposal for revised classification
of epilepsy and epileptic syndromes. Epilepsia 1989;30:389 – 99.
[17] Holmes GL. Epilepsy in the developing brain: lessons from the
laboratory and clinic. Epilepsia 1997;38:12 – 30.
[18] Sztriha L, Gururaj A, Bener A, Nork M. Temporal lobe epilepsy in
children: etiology in a cohort with new-onset seizures. Epilepsia
2002;43:75 – 80.
[19] Gururaj AK, Sztriha L, Bener A, Dawodu A, Eapen V. Epilepsy in
children with cerebral palsy. Seizure 2003;12:110 – 4.
[20] Maytal J, Shinnar S, Moshe SL, Alwarez LA. Low morbidity
and mortality of status epilepticus in children. Pediatrics 1989;83:
323 – 31.
[21] Wingerchuk DM, Sirven JI. Remission, intractability and antiepileptic
drug withdrawal. Lancet Neurol 2005;4:390 – 1.
[22] Berg AT, Langfitt J, Shinnar S, Vickrey BG, Sperling MR, Walczak T,
Bazil C, Pacia SV, Spencer SS. How long does it take for partial
epilepsy to become intractable? Neurology 2003;60:186 – 90.