Abstracts

Rationale: Phenobarbital (PHB) is widely viewed as first-line therapy for treatment of neonatal seizures despite limited efficacy data and a substantial side-effect burden. Among contemporary anti-seizure drugs, levetiracetam (LEV) has gained popularity among some neurologists, with the anecdotal impression that it is effective, and safer than phenobarbital. The goal of this study was to document treatment practices and outcomes at UCLA, where both LEV and PHB are frequently used in the treatment of neonatal seizures.Methods: Full-term newborns with hypoxic-ischemic encephalopathy (HIE) that underwent a standardized therapeutic hypothermia protocol, which includes continuous video-EEG monitoring, were retrospectively identified. For each patient, we tabulated video-EEG confirmed seizure burden, parameters of anti-seizure medication exposure, and pertinent clinical and demographic factors. Time to seizure freedom among LEV- vs PHB-treated newborns was contrasted using Cox proportional hazards regression (CPHR). Comparisons of medians were accomplished with the Wilcoxon rank-sum test (WRS).Results: 78 consecutive newborns with HIE who followed the aforementioned protocol between 2008 and 2014. Of the 44 patients who exhibited video-EEG confirmed seizures, 34 became seizure-free. There were 10 deaths. At the discretion of treating neurologists and neonatologists, neonates received PHB only (n = 13), LEV only (n=18), LEV followed by PHB within 24 hours of first LEV administration (n=2), or PHB followed by LEV within 24 hours (n=10). One additional patient received LEV 34 hours after first PHB infusion. Among patients treated with LEV or PHB only, time to EEG-proven seizure freedom was significantly shorter among LEV recipients (p < 0.01, univariate CPHR); more than 40% of LEV recipients exhibited no seizures after initial infusion. See Figure 1. However, this effect is at least partially confounded by HIE severity/disease burden, as the incidence of death during hospitalization was substantially higher among PHB-treated newborns (p = 0.043, WRS). After adjustment for measures and potential proxies of epilepsy/injury severity (seizure frequency on first day of life, highest seizure frequency on any day, APGAR scores, birthweight, demographic factors), time to seizure-freedom was similar among PHB- and LEV-treated newborns (p >