Abstracts

Rationale: Seizure semiology (SE) is a simple, cost-effective tool that helps with localization of the symptomatogenic zone, which usually is in close proximity to the epileptogenic zone. (Tufenkjian and Lders et al., 2012). Analysis of semiology becomes even more important in the absence of a lesion on brain MRI. SE in neonates and infants may not show reliable information unlike in older patients due to the following reasons: 1. Subtle seizures, often masked by normal movements. 2. Immature dendritic development and brain myelination, and 3. Inability to verbalize subjective experience. Infants who suffered perinatal stroke and subsequent focal seizures were studied to determine the degree of correlation between seizure semiology and the location of cerebral insult in perinatal arterial ischemic stroke or focal hemorrhage. Methods: We retrospectively reviewed the charts of 47 subjects with diagnosis of stroke and epilepsy between January 2006 and December 2015 at Rainbow Babies & Children's Hospital. We excluded 16 pts with hemorrhagic stroke without any parenchymal involvement and multifocal, late onset ischemic stroke (4 pts) and those with hypoxic ischemic injury (6 pts). Seizure symptomatology was classified using the semiology classification (Lders et al., 1998) by an investigator (JP), who was blinded to all clinical data. This result was later correlated with electrographic and imaging data. Results: There were 21 patients (12 females). The mean duration of follow-up was 44.1 months (range 6 to 96 m). Age of seizure onset was within a week of stroke in all patients except in three, who developed seizures at 5, 15 and 20 m respectively. 8/21 patients presented with status epilepticus. 4/21 had evolution into focal seizures and epileptic spasms. Seizure semiology involving frontal, temporal, parietal or occipital lobes resulted mainly in focal or bilateral asymmetric clonic seizure occurring in isolation (6/21), or in association with other components (ocular, tonic, orolingual and, autonomic) (9/21), and hypermotor seizure in 2/21 pts. In 3/21 pts, hypomotor seizures arose from fronto-parietal, parietal and occipital lobe, respectively. Versive seizure was noted in 1 with frontal lobe involvement. All seizures were noted only in the para-sagittal channels (C3/C4 electrodes) in 9/13 with unilateral strokes and 4/8 pts with multifocal stroke. The most common cause of stroke was idiopathic (15/21). 16/21 pts had no motor deficit to mild (1-2/5) distal weakness, 4/21 had moderate-severe weakness (3-5/5) to spastic quadriplegia, and remaining patient had global hypotonia (Medical Research Council, 1943) (Table 1). Conclusions: Seizure semiology was useful in lateralizing but did not reliably correlate with the location of cerebral infarct. Clinical seizures involved para-sagittal EEG channels in majority of the cases, irrespective of the specific location of vascular distribution. Therefore, one should be cautious using seizure semiology in formulating hypothesis regarding the epileptogenic zone in neonates and infants. Funding: None