- Journal Watch
- Levetiracetam Versus Phenobarbital for Neonatal Seizures
Levetiracetam Versus Phenobarbital for Neonatal Seizures
Commentary by:
Leslie H. Hayes1 and Archana A. Patel1,2
1Department of Neurology, Boston Children’s Hospital, Harvard Medical School, Boston, MA 02115
2Division of Epilepsy and Clinical Neurophysiology, Boston Children’s Hospital, Boston, MA 02115
Seizures affect 1-4/1000 newborns and are associated with higher rates of mortality and neurologic morbidity (1,2). Unfortunately, neonatal seizures are often refractory despite aggressive treatment with anti-seizure medications (ASMs)(2). Even with a widening array of ASMs available for use in the management of acute neonatal seizures, phenobarbital remains the most commonly used first-line medication(3). However, with phenobarbital having limited efficacy in this population (approximately 25-72%) and mounting evidence demonstrating the neurotoxic effects on the brain through inducing apoptosis, there is high interest in finding alternative ASMs for the management of neonatal seizures(4).
Levetiracetam has emerged as a leading option given its efficacy in status epilepticus in adults and children and its favorable safety profile. Multiple small and/or retrospective studies suggest efficacy around 30-84%, seemingly equal if not better than phenobarbital(5-8). For example, Rao et al retrospectively studied levetiracetam in neonates with hypoxic-ischemic encephalopathy (HIE) and showed a shorter duration to seizure freedom when compared with phenobarbital(5). With this, levetiracetam use for neonatal seizures is increasing in clinical practice, ahead of the first randomized control trial(9).
NEOLEV2 is the first randomized control trial comparing phenobarbital and levetiracetam for treatment of neonatal seizures(10). Conducted in the United States, this multicenter, blinded trial enrolled neonates under two weeks of age with a corrected gestational age of 36-44 weeks and weight of at least 2.2kg, after being identified as being at risk for seizures or suspected to have clinical seizures. Hence, about 50% of those enrolled were suspected to have hypoxic ischemic encephalopathy, the most common risk factor for seizures in this population. Neonates with metabolic or kidney dysfunction or prior ASM administration were excluded. Continuous video electroencephalogram (cvEEG) monitoring was used to confirm seizures and assess response to ASMs. 83 neonates completed the study and were included in the efficacy analysis: 53 subjects received levetiracetam (40mg/kg initially followed by an additional 20 mg/kg if seizures persisted) and 30 subjects received phenobarbital (20 mg/kg initially followed by an additional 20mg/kg if seizures persisted). After receiving two doses of the assigned randomized first-line therapy , the neonates would then receive the other ASM if required for additional management.
The primary outcome was seizure freedom at 24 hours, which is the same primary outcome used in the hallmark study that determined efficacy of phenobarbital in neonatal seizures though not using cvEEG(3). 80% (24 of 30) of patients assigned to phenobarbital arm achieved seizure freedom at 24 hours compared to 28% (15 of 53) in the levetiracetam arm, which was statistically significant (p<0.001). Given that the optimal dose of levetiracetam for neonatal seizures is not known, the authors also assessed what additional benefit the second dose of levetiracetam (60 mg/kg total) provided and found an additional 7.5% improvement in efficacy. However, this is still significantly lower than the 70% response rate of phenobarbital after just one dose. Interestingly, 25% of patients continued to have seizures despite a total of 40 mg/kg of phenobarbital and 60 mg/kg of levetiracetam (28% of the levetiracetam arm, 17% of the phenobarbital arm), demonstrating the refractory nature of some neonatal seizures and overall poor efficacy of available ASMs in general as seen in prior studies(2). Children with HIE often present with an explosive onset of refractory seizures and the rates of HIE were similar in the two treatment arms and sub-analysis showed similar or slightly improved efficacy in each group (90% efficacy in phenobarbital group, 35% efficacy in levetiracetam group). Of note, the severity of HIE in these patients was not provided.
Although safety was a secondary outcome, it was of high interest in this study. Previous studies report adverse events such hypotension and respiratory depression with phenobarbital as compared to levetiracetam(7). This study similarly showed higher rates of hypotension and need for vasopressor support, respiratory abnormalities including need for ventilatory support, and poor feeding in the phenobarbital arm. The study size was too small to adequately determine statistical significance.
This first randomized control trial of levetiracetam versus phenobarbital in neonatal seizures is a significant step forward in understanding the best approach to managing neonatal seizures. While this study suggests an increased efficacy for phenobarbital over levetiracetam, it also raises important issues regarding the safety profile for both agents. Not only does phenobarbital have more short-term side effects, but also the long-term effects of phenobarbital are not well understood despite is long history of use. In fact, the side effect profile may play a larger role in clinical decision making for medically tenuous neonates and in the context of available resources. Given the better safety profile and some studies suggesting efficacy, there may still be a role in trying levetiracetam first at optimal doses, particularly in certain populations (different severities of HIE, cardiac disease, perinatal stroke). Overall, with increased accessibility of levetiracetam globally, it is more important than ever to further study the potential broad applicability of the agent, and especially in the very vulnerable population of neonates.