Abstracts

Rationale: Outcomes in epilepsy surgery are considered less favorable with multifocal etiologies. Perinatal insults including prematurity and associated comorbidities such as hypoxic ischemic encephalopathy (HIE) often imply multifocal disease. We studied the effect of pre/perinatal injury on epilepsy surgery outcome in children. Methods: A retrospective chart review was performed of all patients who underwent surgical treatment for epilepsy at Rainbow Babies and Children’s Hospital from 2001-2007. Results: For the 106 children who underwent surgical treatment for intractable epilepsy, the mean age was 191 months (range 32-443 months). 50 had a resection, and 46 had VNS insertion.15 had both. 12 of 106 were preterm with median gestational age of 33 weeks (range 24-36 weeks). Comorbidities included HIE (3)and stroke (1). 11 of 12 underwent invasive monitoring. 8 had a resection and 3 had bilateral foci and had VNS placement. 1 had VNS insertion after resection and 1 had VNS as initial treatment. 2 were lost to follow-up (LTF). Of the 8 who had resections, 2 were seizure free at 1 year follow-up (Engel Class I), 4 had >80% reduction and 2 had no change. With VNS alone 1 had rare seizures (Class II), and 1 had no change. 5 full term (FT) children had strokes. 2 underwent resection, 2 had VNS, and 1 had both. The 2 with resection alone were Class I. 1 with resection plus VNS and 1 with VNS alone had 50% seizure reduction and another with VNS had no change. 18 FT patients had HIE with nonspecific encephalomalacia, and volume loss. In the resection alone group, 2 of 4 were seizure free, and 2 had a >80% reduction. In the VNS group, 2 of 8 were seizure free, and 6 had no improvement. In the group with resection plus VNS, 1 of 4 was seizure free, and the remaining 3 had no change. 21 FT patients had focal cortical dysplasia. 5 of 11 in the resected group were seizure free at 1 year, 3 had an 80% reduction, and 3 had no change. 4 in the VNS only group had no change. 1 was LTF. 11 of 106 patients had neonatal seizures (NS). 1 was preterm. Most frequent etiologies included HIE (4). 3 with NS with an unknown etiology were not included in the above categories. 7 of 11 underwent resection alone, 1 had resection plus VNS and 3 had VNS alone. Outcomes were Class I:4, Class II:1, >80% reduction in 1, and no change in 5. 47 of 106 patients had no evidence of pre or perinatal insult. Most had no known etiology (30). In the resection only group, 14 of 22 were seizure free, and 5 had <80% seizure reduction. In the VNS only group, 4 of 19 were seizure free. Among those children who underwent resection, or resection plus VNS, those with a known perinatal insult had worse seizure outcomes (Class I/II:12, Class III/IV:24), compared to those with a normal pre or perinatal course (Class I/II:15, Class III/IV:7; Chi-square analysis, p=0.009). Conclusions: Early brain insult may adversely affect epilepsy surgery outcomes. Our limited numbers in each subgroup of pre/perinatal etiologies resulting in refractory epilepsy preclude subanalyses. Future studies in the clinical and lab arenas will further optimize the management of these children.