318
Seventy-five infants participated in the study. Their gestational ages ranged from 37 to 41 weeks, except one who
was born at 36 weeks; birthweights ranged from 1670 to
4300 g. The exclusion of preterm infants was not intentional but may be due to preterm babies being prone to acute
symptomatic seizures. Also, the absolute number of
preterm babies is fewer than term babies. Seven patients
died before 1 year of age. The other patients were followed
up for a minimum of 3 years and their ages at the last followup ranged from 3 to 19 years.
PROCEDURE
Idiopathic
Benign familial neonatal
convulsions
Benign neonatal convulsions
Symptomatic
Early infantile epileptic
encephalopathy
Early myoclonic encephalopathy
Holoprosencephaly
Tuberous sclerosis
Othera
Cryptogenic
Total
Nr of
patients
16
16
2
5
4
25
7
75
2
2
4
25
7
63
0
3
0
0
0
9
0
0
0
0
0
3
EIEE
WS
EIEE
SGE
EIEE
SLRE
EIEE
Seizure free
EIEE
Early death
EME
WS
Holoprosencephaly
SGE
WS
SGE
LRE
SGE
Seizure free
SLRE
SLRE
Tuberous sclerosis
SLRE
SLRE
SLRE
WS
SLRE
WS
SLRE
WS
SLRE
WS
SLRE
WS
SLRE
SLRE
SLRE
Seizure free
Cryptogenic epilepsies
CLRE
CLRE
Nr of patients
SLRE
LGS
1
1
3
1
1
1
2
SLRE
LGS
1
1
1
2
1
1
2
SLRE
SGE/LGS
SGE
SLRE
Death
Seizure free
4
4
1
1
10
5
7
320
led to the concept of age-dependent epileptic encephalopathy, but other evolutionary changes can occur. Two of our
eight infants with EIEE eventually developed localizationrelated epilepsy with or without passing through the period
of West syndrome. Ohtahara et al. (1992) reported that eight
of 15 patients eventually had focal spikes with or without
passing through the period of hypsarrhythmia. We examined
a child with this syndrome in which suppressionburst patterns were replaced by continuous background EEG with
spindles displaying only focal spikes, which then evolved
into multifocal spikes and then to hypsarrhythmia (Watanabe
et al. 1987). This may contradict the notion that EIEE may be
West syndrome with an early onset (Lombroso 1990).
EME is also categorized as symptomatic generalized
epilepsy in the International Classification. It is characterized
by partial or fragmentary erratic myoclonus, massive myoclonia, partial motor seizures, and suppressionburst pattern on
EEG (Aicardi 1992). The infants under study did not demonstrate generalized seizures and showed only partial seizures
in addition to erratic myoclonia during the neonatal period,
and continued to show partial seizures thereafter. This syndrome may not necessarily be placed under the category of
generalized epilepsy, although some patients with this syndrome may develop West syndrome transiently.
Neonatal epilepsy with rhythmic alpha/theta activity is characterized by a peculiar EEG consisting of high-amplitude rhythmic alpha/theta activity, which is continuous during
wakefulness and becomes discontinuous during sleep, tonic
seizures associated with desynchronization, and recruiting
rhythm followed by oral automatism associated with slow
waves (Watanabe et al. 1976). This is typically seen in neonates
with semilober holoprosencephaly. In alobar holoprosencephaly, rhythmic alpha/theta activity is present only in anterior
regions. In such cases, seizure types are difficult to determine
due to the absence of the brain in posterior regions, although
they were provisionally classified into partial seizures by definition. Continuous high-voltage rhythmic alpha/theta/delta activities are also observed in other cortical dysplasia such as type I
lissencephaly, but are not usually observed in the neonatal
period (Dalla Bernardina et al. 1996). Moreover, they do not
show asynchronous discontinuity in sleep and have no anteriorposterior voltage gradient.
The International Classification classifies neonatal
seizures into undetermined epilepsies with both generalized
and focal seizures. But in this study, only three patients with
EIEE had both generalized and partial seizures in the neonatal period. Therefore, epilepsy with both generalized and
focal seizures is rather exceptional in neonatal epilepsies.
In our present study, seven infants were diagnosed as having cryptogenic epilepsy. They had no evidence of underlying disorders and exhibited normal or mildly delayed
psychomotor development, despite the presence of
intractable partial seizures. They seemed to have idiopathic
epilepsy because of the absence of any apparent cerebral
lesions or definite developmental retardation, but the refractoriness of their seizures suggests the presence of latent
minor cerebral lesions that could not be detected by existing
neuroimaging studies. The presence of such cryptogenic
localization-related epilepsy of neonatal onset has been
described by Natsume et al. (1996). Our subjects and those
studied by Natsume continued having partial seizures without evolutionary changes of epileptic syndromes.
322