Abstracts

Controversy remains over whether neonatal seizures are an independent cause of brain injury. This has resulted in continued uncertainty about how aggressively neonatal seizures should be treated, particularly electrical seizures without clinical correlate. One mechanism by which seizures may injure the brain is by causing a mismatch between cerebral oxygen supply and demand. This study investigated the cerebral hemodynamic consequences of neonatal seizures, in order to obtain insights into their potential contribution to brain injury., Newborn infants (age [le]30 days) diagnosed with seizures or suspected seizures of all gestational ages and seizure etiologies underwent up to 3 days of simultaneous recording with video-EEG and two-channel near infrared spectroscopy (NIRS) (NIRO 200, Hamamatsu Photonics), in addition to continuous measurement of ECG and systemic oxygen saturation. NIRS optodes were placed bifrontally. NIRS measurements of oxyhemoglobin (HbO2), deoxyhemoglobin (Hb) and tissue oxygenation index (TOI) were performed at 2 Hz. A calculated hemoglobin difference signal (HbD = HbO2-Hb) was used to estimate changes in regional cerebral blood flow., Of 12 neonates enrolled in the study, seizures were captured on both EEG and NIRS in 4 subjects. In 2 subjects with electrical (subclinical) seizures, focal epileptiform activity was associated with lateralized increases in HbD, likely reflecting increased regional CBF. Sustained ictal rhythms were not required to observe a change in HbD. Brief (3-5 sec) bursts of high amplitude (400[micro]V) epileptiform discharges were sufficient to cause sustained (20-30 sec) lateralized increases in HbD. The magnitude of HbD increase appeared to correlate with the amplitude or power of the epileptiform activity, rather than the degree of organization into a sustained ictal rhythm. In one subject with tonic seizures, the NIRS signals were uninterpretable due to accompanying systemic oxygen desaturation and movement artifact. In one subject with myoclonic jerks without EEG correlate, there were no associated changes in NIRS signals., These preliminary observations challenge the traditional view that sustained rhythmic epileptiform discharges have greater physiologic consequences than brief bursts of interictal epileptiform activity. No lateralized ictal decreases in HbD were observed in any of these subjects, implying that regional CBF was sufficient to meet the cerebral metabolic demands of these seizures. The observation that myoclonic jerks without EEG correlate were not associated with any change in HbD supports the hypothesis that these movements are mediated by subcortical rather than cortical structures., (Supported by Duncan L. Gordon Fellowship, The Hospital for Sick Children Foundation, Toronto; William Randolph Hearst Fund, Harvard Medical School; National Epifellows Foundation.)