Indian Pediatrics 2001; 38: 390-396
Epileptic Encephalopathies of Early Childhood
The term ‘Epileptic Encephalopathy’ (EE) refers to a heterogenous group of conditions in which even in absence of progressive metabolic and/or structural brain abnormali-ties, the extremely abnormal brain electrical activity may not only be the cause of seizures, but also interfere with cognitive functions, leading to an arrest or regression in intelligence/behavior(1). The disorders sharing these characteristics, and thus included in this group are, early myoclonic encephalopathy (EME), early infantile epileptic encephalo-pathy (EIEE), West Syndrome (WS), Lennox Gastaut Syndrome (LGS), severe myoclonic epilepsy in infancy (SMEI) and certain myo-clonic or myoclonic astatic epilepsies(1,2). The clinical and electroencephalographical features of these conditions may overlap. Also, some of these disorders may with age, evolve into another. The evolution of EIEE into WS and further into LGS is well discribed(3). Thus, EE of children pose difficult problems of diagnosis and prognosis due to the polymorphism of their electroclinical and evolutive pattern.
The present work was undertaken to study the etiology and the clinical spectrum of patients with epileptic encephalopathies, to classify them under various epileptic syn-dromes and, to assess their neurodevelop-mental status and response to antiepileptic drug therapy.
The present study was conducted over a period of 1 year from July 1997 to June 1998. The subjects of the study were selected from the Epilepsy Clinic of the Pediatrics depart-ment. Children who had recurrent, difficult to control seizures associated with arrest or regression of development in absence of a progressive brain pathology were considered to be suffering from EE. All such children with onset of seizures under 3 years of age were included in the study.
These patients were evaluated by a detailed history with special reference to antenatal, natal and postnatal events, the age at onset of seizures and the type and evolution of seizures. A detailed neurodevelopmental assessment was carried out in all children at presentation and during the follow up visits at 3 monthly intervals. An EEG and CT/MRI scan was done at presentation and repeated later if indicated. Investigations for inborn errors of metabolism were done where indicated by performing blood gas analysis, serum ammonia, serum and CSF lactate, urinary amino acidogram, muscle and skin biopsy, etc.
On the basis of the clinical profile, EEG and neuroimaging, the patients were categorized under various epileptic syndromes as per the International League Against Epilepsy (ILAE) classification(4). Patients diagnosed as cryptogenic WS were given a course of prednisolone/ACTH. Patients who did not respond, or had seizure recurrence on cessation of therapy, received sodium valproate or benzodiazepines. Patients in other diagnostic categories were initially given either sodium valproate or benzodiazepines. Vigabatrin was used in selected cases of refractory WS. Response to drug therapy was assessed in patients with a minimum follow up of 6 months and graded by noting the reduction in seizure frequency.
Over the 1 year study period, 37 children, 28 males and 9 females, satisfying the inclusion criteria, were enrolled in the study. Of these, seventeen patients (45.9%) had onset of seizures within 3 months of age (Group A), and the remaining twenty (54.1%), after 3 months (Group B). Infantile spasms were the presenting feature in twenty cases. The remaining seventeen patients had myoclonic seizures alone, or in combination with other seizure types (Table I). In thirteeen (35.1%) cases no definite underlying etiology could be established. Birth asphyxia and central nervous system infections accounted for a majority of cases with known etiology. In some patients, more than one implicating factors were present (Table I).
As per the inclusion criteria, none of the patients had a normal development. However, the extent of retardation varied amongst them. Twenty six patients had a global delay in development dating back to birth. Of these, sixteen belonged to group A and ten to group B (94.1% and 50% of all group A and group B patients, respectively). Three patients in group B had an isolated language delay with a normal motor development prior to seizure onset. In the remaining 7 Group B patients, development remained normal till onset of seizures after which it showed an arrest or varying degree of regression. One patient with tuberous sclerosis in group B had autism.
Three patients in Group A (17.6%) and nine in group B (45%) did not show any other neurological abnormality apart from neurodevelopmental delay. The salient physical findings in the remaining are included in Table I. Five patients had no visual fixation. Visual evoked responses were studied in all of them. While no wave form could be picked up in one case, three showed a delayed response. VER was normal in one case. BERA performed on one patient with hearing deficit showed bilateral moderate hearing loss. This patient had suffered from neonatal meningitis.
MRI/CT scan was performed in all cases and was found to be abnormal in 28 (75.7%). The neuroimaging and predominant EEG findings are summarized in Table I.
The classification of patients under various epileptic syndromes is presented in Tables II and III. Eight patients in group A could not be classified under any defined epileptic syndrome presenting during this age. Of these, five had a definite history of perinatal asphyxia/postnatal meningitis which could have led to the present condition. The diagnosis of EIEE was made retrospectively in one patient who had presented at the age of two years with LGS with history of primarily tonic seizures since one month of age. Twenty patients were diagnosed as WS. Of these, six belonged to group A. All of them had an identifiable underlying cause while five out of fourteen cases with WS in group B were cryptogenic.
Response to antiepileptic drugs was assessed in 31 cases who had been followed up for atleast 6 months. Only one case with cryptogenic WS had a complete remission of seizures. Nine cases (29%) showed no appreciable reduction in seizure frequency. Other cases showed a variable reduction in seizure frequency (Tables II and III). In general, response in group A patients was poorer as compared to group B. Within group A, all patients exhibited a similar response irrespective of the syndromic diagnosis.
Several attempts have been made to classify epileptic syndromes and these definitions continue to be refined. Overlapping clinicoelectrical features make classification of patients within a specfic syndrome often difficult. The electoencephalogram of these cases may show evolutive changes with time and often serial tracings are needed to reach a definite diagnosis. The CT/MRI scan may also be normal initially and show abnormality only later on(5,6). Thus, these children, who suffer from profound physical, social and mental handicap, often shift from one physician to another, lacking a proper diagnosis and receiving multiple drugs. Proper classification of these patients by understanding their etiology and clinical features is a keenly felt issue. There is paucity of Indian data on this subject(7,8). In this study, we have docu-mented the clinical features and outcome in children with various epileptic encephalo-pathies, who presented to our epilepsy clinic.
Since age at onset of seizures is an important factor in differentiating various epileptic syndromes, it was used to classify the patients into two groups. In group A, two patients were diagnosed as EME on the basis of typical clinical features and documentation of burst suppression pattern on EEG. In one of the cases there was a history of one sibling dying at the age of ten months with similar illness. No definite etiology could be established for either case despite an extensive metabolic workup. CT performed within three months of age in both cases was normal. Other workers have also often been unable to find the etiology for EME. Undetected inborn errors of metabolism may be responsible for some cases since this disorder can be familial(6,9,10).
Eight patients in this group could not be classified into any syndrome. While five of them had an underlying history of asphyxia or meningitis, no cause could be established for the remaining three patients. All of these patients had myoclonic seizures alone, or in combination with other types of seizures. The EEG of these patients, though showing generalized epileptiform discharges did not show burst suppression pattern character-istically seen in EIEE or EME; the two encephalopathies classically presenting during early infancy. However, this pattern may not be seen during early stages, becoming manifest only later during the course of illness(6). Again, this pattern may not persist indefinitely and may get replaced by hypsarrhythmias for a variable duration of time(11). For most of these cases, serial EEGs were not available, which could have aided in reaching a definite diagnosis.
Twenty cases were diagnosed as WS. Of these, five were cryptogenic (25%), all belong-ing to group B and the remaining symptomatic. Six cases with symptomatics WS had presented with seizures within three months of age. All of them had characteristic flexor spasm and hypsarrythmia on EEG.
Six cases were diagnosed as LGS. One of them had shown an evolution from EIEE to LGS. The other cases had presented with multiple types of seizures after one year of age and had typical EEG features.
In 26 patients (70.2%), etiology was related to perinatal events or a central nervous system infection during early infancy. Structural malformations accounted for only 4 cases (10.1%). This is at variance with data from developed nations where a majority of cases are due to prenatal causes (12,14). Similar observations have been made in another Indian study(7).
While no patient was developmentally normal as per the inclusion criteria, group A patients were more serverely retarded as compared to group B. The extent of retardation in the former was comparable regardless of syndromic classification. Severe develop-mental delay in patients with early onset EE has been documented in other studies too(3,5,6). In group B, developmental retarda-tion was more in patients with symptomatic WS as compared to cryptogenic WS, a finding corroborated by other studies too(3,12,15). In all patients with cryptogenic WS, the development was normal till seizure onset. Similarly, while development in three patients with cryptogenic LGS was normal till seizure onset, it was globally delayed in two patients with preceding birth asphyxia. Thus developmental outcome in EE seems to be related to both etiology and the age at onset of seizures, both of which may, in turn, be related.
Response to drug therapy was again different in the 2 groups. Patients in Group A showed a poor response to therapy with only 14.2% having more than fifty per cent reduction in seizure frequency and 42.9% having no change in seizure frequency. Other studies on EME and EIEE have also reported a poor response of seizures to antiepileptic therapy(3,5,6). In group B, overall response to therapy was better than group A. Patients with cryptogenic WS had a better response as compared to those with symptomatic WS. In the former sub-group, while one out of 5 patients had total seizure control, others had >50% reduction in seizure frequency. All patients in the symptomatic group showed less than 50% reduction in seizure frequency. Amongst the cases with LGS, only one case showed more than 50% reduction in seizure frequency, no patient being seizure free. In Ohtsuka’s series also 71% of LGS cases had refractory epilepsy(3).
Thus, it is apparent that patients with EE, in general, have a poor developmental outcome and response to therapy. Children with early onset of seizures (within 3 months of age) have the poorest prognosis in both respects. Other workers have also found age at seizure onset to be of prognostic significance in EE (3,16). This may be due to two important reasons. First, prognosis of epilepsy is influenced by severity of brain damage. A severe brain damage is likely to be associated with earlier onset of seizures. Second, brain damage cuased by seizures also varies with age. Earlier the onset of seizures, more is the resultant damage(3).
To conclude, in the present study, we found a high incidence of perinatal asphyxia and postnatal infections in children with EE. Classification of the cases under specific epileptic syndromes was difficult with eight cases (21.6%) defying classification despite an extensive workup and follow up. Most cases exhibited a grim developmental outcome and a poor seizure control. The age at onset of seizures was found to be an important determinant of seizure control and future developmental outcome.
Contributors: AS planned the study, collected and interpreted the data and drafted the paper. SA helped in planning the study, data collection and drafting the paper. VT helped in data collection.