Abstracts

Rationale: Seizures in the neonatal intensive care unit are very challenging. 'Electrographico-clinical dissociation' unique to this age group due to minimal myelination and subtle clinical presentation, if any, makes diagnosis and treatment of seizures difficult.  This is particularly so  with extremely premature babies, for whom 'normal' EEG is not defined. It is well-known from translational and human longitudinal studies, that neonatal seizures are associated with significant neurological morbidity in later life.  Thus, detecting and treating those subtle events appears to be crucial, as part of neonatal care in babies at high risk for seizures. As a quality improvement project, we set out to screen with EEG one group particularly at risk for seizures: premature babies with known intraventricular hemorrhage grade III and IV. The objective of the study was to assess whether those newborns were having subclinical seizures. Methods: IVH III and IV patients were identified and put on continuous video EEG monitoring for 48 hours. Those with clinically suspected seizures were excluded; they were studied with EEG and treated clinically per standard of care.  For extremely premature babies in isolettes, minimal  EEG montage was used, in order not to handle their heads and necks and minimize hookup time. All of usual frontal and midline electrodes were employed, omitting occipital and posterior parietal contacts.  The recordings were read and interpreted by staff pediatric epileptologists. Data were entered into Red Cap clinical database for future analysis.  Forty to fifty cases were anticipated in one year, but the actual number enrolled was only 11, making quantitative statistical analysis not feasible.  Results: Eleven patients were monitored in days 7-11 of life.  Gestation age was 23- 33 weeks.  Four had grade III IVH, the rest had grade IV.None of the patients had seizures as defined  'a rhythmic evolving discharge over 10 seconds in duration,' which is a classic definition of neonatal seizures.  Eight out of eleven patients, however, had myoclonic-like jerks, whole-body, maximal in arms and head/neck, which had a clear EEG correlate of generalized sharp wave followed by decrement.  These jerks were mostly seen on video and also were recognized in limb and chin EMG recordings.  Some jerks were so subtle that only EEG and EMG findings correlated; other events were clearly seen on camera through the isolette. None of the babies were treated.High-frequency ripples at 100-200 Hz were seen in all patients with myoclonus.  Conclusions: Neonatal seizures in at-risk premature babies are defined very narrowly, and possibly due to this, multiple subtle myoclonic seizures are considered normal and not treated.Multiple myoclonic jerks with a clear EEG correlate were found in IVH III and IV babies, however without a control group of premature babies without IVH or term babies with IVH.The significance of these findings is unclear: it could be an epiphenomenon of prematurity  itself, of IVH itself, or it could be a neonatal ictal variant which warrants attention, monitoring, and prophylaxis with antiepileptic drugs.More studies are needed to evaluate the possibility that by sticking to a very strict definition of neonatal seizures, we are missing a significant number of them and failing to treat and prevent neurological sequelae. High-risk neonates ought to be screened with EEG, and subtle myoclonic events may warrant treatment. Funding: No funding