Abstracts

Rationale: Early life epilepsies (ELE) are among the most refractory forms of epilepsy and carry a high risk of early mortality. Most research emphasis and clinical guidelines focus on specific electronical syndromes such as West/Infantile Spasms (WS/IS) and Dravet. We sought to understand the initial outcomes in the full spectrum of these epilepsies. Methods: Children with newly diagnosed ELE (onset < 3y) were prospectively recruited at 17 US pediatric epilepsy centers (2012-2015) and followed for up to 1-year after diagnosis. Clinical data were extracted from medical records, coded, and entered in a REDCAP© database. We defined pharmacoresistance as failure of 2 medications and without adequate response to a subsequent therapies. In infants ( < 1y), we also ascertained evolution to WS/IS from non-WS/IS and evolution to new seizure types for children initially presenting with WS/IS. Standard univariate descriptive and bivariate methods were used for analysis. Results: 729/775 (94%) enrolled patients had adequate follow-up to characterize seizure outcome. 386 (53%) were boys. At seizure onset, 484 (66%) were < 1y, 142 (20%) 1y and 103 (14%) 2y-old. Epilepsy diagnosis was WS/IS in 222 (31%); 45 (6%) other specific epilepsy syndromes (5 each Dravet, Doose, Myoclonic Epilepsy of Infancy and 1-4 each of other syndromes); 462 (63%) had nonsyndromic epilepsies (NSE). Pharmacoresistance (PR) occurred in 189 (26%) patients and was more common with younger age at onset (30% if< 1y; 23% if 1y; 13% if 2y; p = 0.0002). In Infants (Table ), PR was less common in WS/IS (24%) than in NSE (33%) and other epilepsies (45%, p=0.03). In infants, developmental delay was associated with ~2-fold increase risk of PR, independent of syndrome type (p < 0.0001). Of 248 infants with NSE, 34 (14%) evolved to WS/IS. Evolution to WS/IS occurred in 9% with normal development, 19% with equivocal delay and 23% with definite delay (p=0.004). Of infants who initially presented with WS/IS, 46 (22%) developed other seizure types; 11% if development was normal, 14% if equivocal, and 29% if definitely delayed (p=0.006). Having an identified cause was a weaker predictor of seizure outcomes. 21 (3%) patients died; 20 infants and 1 in the 1y group (p=0.006). In infants, type of epilepsy did not influence risk of death (p=0.66). Deaths occurred in 0.5%, 1.5%, and 8.1% of infants with normal, equivocal, and delayed development (p < 0.0001); delay and identified cause independently predicted mortality. 9% with PR seizures versus 2% of others died (p=0.0009). Conclusions: In the age range studied ( < 3y), younger onset age is associated with worse outcomes. Although WS/IS is considered one of the most serious forms of ELE, as currently treated, the outcomes of nonsyndromic epilepsies in infants are equally poor if not worse. In the very young, developmental delay at initial diagnosis may provide a more robust indicator of poor outcomes than the type of epilepsy Funding: Pediatric Epilepsy Research Foundation