Abstracts

Most data regarding mortality and epilepsy come from studies of adults or of mixed age groups. There are only a few reports concerning predictors of death in individuals with childhood onset epilepsy studied and characterized according to modern standards. We report the mortality experience in a community-based cohort.
A prospectively ascertained cohort of children (N=613) with onset of epilepsy [lt]16 years and first diagnosed during 1993-1997 was assembled. To date, the group has been followed a median of 7 years. Epilepsy syndromes were classified according to the 1989 classification of epilepsy syndromes both at initial diagnosis and two years later. Epileptic encephalopathy (EE) was defined per Dulac (Epilepsia, 2001:42:S23-6).
Ten deaths have been observed occurring at 7 months to 7 years after initial diagnosis median ~4 years . The age at death ranged from 1.8 to 16.5 years (median=4.5yrs). With one exception, all deaths were related to the occurrence or presumed occurrence of seizures (SUDEP, N= 2), the causes or consequences of seizures (N=4) or a neurodegenerative condition as the underlying cause of the epilepsy (N=3). Factors associated with death were age at onset [8/113 (7.1%) [lt]2 at initial onset of epilepsy versus 2/500 (0.4%) who were older (p[lt]0.001)]; EE [8/69 (11.6%) with EE versus 2/544 (0.4%) with other forms of epilepsy (p[lt]0.0001)]; remote symptomatic etiology [8/121 (6.6%) versus 2/492 (0.4%) (p[lt]0.0001)]; and seizure outcome [6/57 (10.5%) who were clearly intractable immediately after diagnosis versus 3/545 (5.5%) who were not]. One child died shortly after diagnosis and had insufficient follow-up to characterize seizure outcome). In a multivariable proportional hazard model analysis, remote symptomatic etiology (Rate Ratio (RR) = 6.6, 95% CI, 1.3, 34.4 p=0.02 ) and EE diagnosed within two year of initial diagnosis (RR=17.0 , 95% CI= 3.3, 88.6, p=0.0008) were the primary predictors of mortality and accounted for 9 of the 10 deaths and 8 of the 9 epilepsy-related deaths.
Epilepsy-related death occurs in a small but important proportion of children with epilepsy. Because of the relatively few cases affected even in large series, results of any single study are often less than definitive. Our findings, however, are highly comparable to those of a similarly conducted study in Holland (Callenbach, Pediatrics 2001;107:1259-63) and a population-based study in Nova Scotia (Camfield, Lancet, 2002). Together, these studies strongly comfirm that almost all of the mortality that occurs within in the first several years of onset of epilepsy in children is associated with a symptomatic etiology or an epileptic encephalopathy. The risk of SUDEP is fortunately very low in individuals with childhood onset epilepsy.
[Supported by: National Institututes of Health - NINDS-RO1-NS31146]