The pathophysiology of neonatal seizures, and evidence for seizure-induced brain damage are reviewed by researchers from Montreal Children's Hospital, and Universite de Montreal, Quebec, Canada. Electrographically documented seizures with or without clinical manifestations are the most accurate concept of neonatal seizures. Incidence is 1.5-3.5 per 1000 live births, varying with risk factors such as low birth weight, prematurity, perinatal complications, and NICU availability. Etiology is the major determinant of outcome, but the seizure itself may be a factor. In animal models, neonatal seizures impair cognition, alter behavior, increase anxiety, and are associated with epileptogenesis. Clinically, the reported prevalence of epilepsy and abnormal neurodevelopment after neonatal seizures varies, ranging from 6.5% to 56% for epilepsy and from 19% to 67% for neurological abnormalities. Electrographic neonatal seizures, with or without clinical manifestations, correlate with increased morbidity and mortality. Risk factors for epilepsy include diffuse abnormalities on cranial imaging (more than focal), prolonged use of anticonvulsants, poor response of neonatal seizures to phenobarbital, abnormal EEG background, and acquired CNS infections. The efficacy of both phenobarbital and phenytoin in neonatal seizures is 50%, and no randomized controlled trial is reported to show improvement in neurodevelopmental outcome or prevention of epilepsy. Controlled trials of newer anticonvulsants and neuroprotective agents are needed. 
COMMENT. Although the experimental evidence from animal studies suggests that neonatal seizures can damage the developing brain, clinical reports are less convincing. Nonetheless, given the potential for seizure-induced brain damage, early and more effective treatment of neonatal seizures, including electrographic seizures, is generally advocated.
A study at Boston Children's Hospital on etiology and outcome of neonatal seizures found that global cerebral hypoxic-ischemia is the most frequent cause of neonatal seizures and a strong predictor of poor long-term outcome (Ped Neur Briefs April 2006;20:29-30) . An abnormal neurologic examination in the neonatal period was an unreliable predictor of outcome.
Perilesional brain oedema and seizure activity: cause or effect? In response to a publication concerning seizures with calcified neurocysticercosis , Das A et al. question the proposed causative correlation of epileptogenesis and perilesional edema as an inflammatory response to calcified granulomas . Das et al agree that calcified lesions are seizure-causing foci but attribute the perilesional edema to the effect of the seizure per se. Reversible peri-ictal MRI abnormalities are known to develop immediately after a seizure and may resolve within a few days or weeks. Their precise pathogenesis is unknown.
Hippocampal swelling demonstrated by MRI within 48 hours of a prolonged febrile seizure is transitory and thought to be caused by vasogenic edema. The swelling resolves within 5 days and the shrinkage of the hippocampus that follows at 4-8 months is thought to represent a preexisting developmental hippocampal abnormality that predisposes to the prolonged FS . Alternatively, the hippocampal shrinkage is consistent with brain damage caused by the seizure with subsequent development of mesial temporal sclerosis and epilepsy. (See Ped Neur Briefs Oct 2006;20:77; and Nov 2003;17:83).