The rapid sequential phenobarbital treatment of neonatal seizures was examined in 120 newborns and the efficacy of high dose monotherapy was compared with the addition of a second anticonvulsant for persistent seizure activity. Patients were examined in three participating neonatal intensive care units: Comprehensive Epilepsy Center, Pharmacokinetics Laboratory, Miami; Greensboro Area Health Education Center; and Department of Neonatal Medicine, Moses H. Cone Memorial Hospital, Greensboro, North Carolina. A single loading dose of phenobarbital 15-20 mg/kg was administered initially and nonresponders received sequential bolus doses of 5-10 mg/kg until seizures ceased or a serum concentration of 40 mg/mL was obtained. Infants with refractory seizures received additional phenobarbital to a maximum serum concentration of 100 mg/mL. The majority of neonates with recurrent seizure activity (77%) responded to phenobarbital monotherapy administered in a rapid sequential dosing schedule that achieved a serum concentration of 40 mg/mL. In 40%, seizures were controlled with a single 15-20 mg/kg initial loading dose and a serum concentration in the range of 10-30 mg/mL. Of 28 subjects refractory to phenobarbital, 13 (46%) were controlled by a second anticonvulsant (phenytoin or lorazepam) and four were controlled by three or more agents. Eleven were resistant to medication and ten died. There was no significant difference in drug responsiveness among patients with different seizure patterns and seizure etiology was not a significant determinant of phenobarbital responsiveness. Subjects less than 32 weeks gestational age responded better than those 32 weeks or greater in gestation. 
COMMENT.. The therapeutic effect of phenobarbital monotherapy in the treatment of neonatal seizures is dose dependent but the effect plateaus at 40 mg/mL and further increases only induce sedation and compromise neurologic assessment. A second anticonvulsant should be given promptly if seizures persist when phenobarbital serum concentrations are 40 mg/mL or above. The availability of rapid serum determinations of drug levels is important because delays in additional drug therapy were thought to predispose to further seizures and risks of serious neurologic sequelae. The authors emphasize that adequate patient monitoring and slow infusion rates of phenobarbital should always be used to avoid possible cardiovascular toxicity.
In our own experience of 63 newborns with seizures admitted January 1985 to December 1987 in the high risk nursery at SIU School of Medicine, monotherapy with phenobarbital was used in 70% and polytherapy in 30%. The mean serum phenobarbital levels after loading doses of 10 and 20 mg/kg were 20 and 40 mg/mL, respectively. Factors predictive of a poor prognosis included 1) polytherapy, 2) Apgar score less than 5 at five minutes, 3) abnormal EEG, and 4) abnormal brain ultrasound. Normal EEG and ultrasound were predictive of a normal follow-up examination. The addition of phenytoin or other polytherapy did not appreciably increase the degree of seizure control or improve prognosis (unpublished observations).