Abstracts

Rationale: Seizures are the most common neurological complication in neonatal intensive care, and are associated with high risk of death, epilepsy, cerebral palsy, and cognitive impairments. Phenobarbital (PB) remains the first-line anti-epileptic drug (AED) for neonatal seizures in spite of its known association with neurotoxicity and long-term cognitive impairment. Levetiracetam (LEV) is a newer AED FDA approved for adults and children, but not neonates. LEV has demonstrated a favorable safety profile in pediatrics and adults, and is commonly used off-label to treat neonatal seizures. Several small retrospective and pilot studies have investigated the use of LEV in neonatal seizures, with mixed results regarding efficacy. There is a paucity of randomized controlled trials of AEDs in neonates. Our objective was to compare the efficacy of LEV to PB in neonatal seizures based upon published data.Methods: We searched Pubmed to perform a systematic review and meta-analysis in accordance with the PRISMA statement. We included studies of neonates; use of LEV (primary or secondary); electrographically proven seizure diagnosis and response. Exclusion criteria were: single case reports or seizures due to correctable electrolyte abnormalities. LEV was considered successful with complete or ≥80% seizure resolution. We found no studies of LEV with comparison or control groups, so utilized data from two randomized controlled trials of PB as our comparison group. We calculated odds ratios (OR) comparing efficacy of LEV to PB, using PB as the reference value.Results: Our search yielded 155 results. Five studies of LEV met all inclusion/exclusion criteria and were included in the final analysis. The pooled sample size for LEV was 102 (48 received primary LEV and 54 received secondary LEV). The pooled sample size for PB was 52. Some patients were treated with LEV as monotherapy, though the majority received additional AEDs. The overall OR for seizure cessation with primary LEV was 2.02 (95% confidence interval (0.89, 4.57), p=0.09, Figure 1). The overall OR for seizure cessation with secondary LEV was 1.14 (95% confidence interval (0.57, 2.28), p=0.72, Figure 2). This data demonstrates no significant difference in the odds of achieving seizure cessation with LEV compared to PB. A total of 3 mortalities were reported in the 102 LEV patients and 8 mortalities were reported in the PB group.Conclusions: We found no significant association between LEV and improved seizure reductions compared to PB. However, the lack of significant results may be due in part to the small sample size, as the association between primary LEV and seizure cessation was near significance (p=0.09). Regardless, LEV may offer equivalent seizure control efficacy as PB without known neurotoxicity. Our meta-analysis, though limited, is the first to examine LEV efficacy compared to PB in neonates. Results support further investigation of LEV and the use of off-label LEV when deemed clinically appropriate in the setting of neonatal seizures. No funding was utilized in the synthesis of this publication.