The efficacy of vigabatrin (VGB) as the first, and adrenocorticotropin hormone (ACTH) or valproate (VPA) as the second, treatment of choice for newly diagnosed infantile spasms was evaluated in 42 infants treated at the University of Helsinki, Finland. Response was measured by total control of spasms for a minimal duration of 1 month and confirmation by video-EEG. Spasms were cryptogenic in 10 and symptomatic in 32. Vigabatrin (50-100 mg/kg/day) controlled spasms in 11 (26%); 5 were cryptogenic and 6 symptomatic. ACTH offered in combination with VGB in 22 and VPA in 4, not controlled by VGB alone, were effective in 11 (50%) and 1 (25%), respectively. Overall, 26 (62%) infants responded to treatment; 100% response with cryptogenic cases and 50% with symptomatic etiology. Side effects were more severe with ACTH than VGB or VPA. Relapse after a spasm-free period of >4 months occurred in only 1 infant treated with VGB, but in none who received ACTH combined with VGB. ACTH should be considered after an initial trial of VGB in increasing doses from 50 to 150 mg/kg for 10-14 days. 
COMMENT. Vigabatrin is suggested as the first treatment for all infants with infantile spasms. In non-responders, ACTH should be considered. Response should be confirmed by video-EEG of 3-4 hours during waking and sleep, and visual responses should be carefully monitored when practical. Etiology of spasms is the most important determinant of treatment outcome; cryptogenic cases benefit more frequently than those with acquired causes. Most infants with symptomatic etiology are mentally retarded at follow-up. See Ped Neur Briefs (June 1999; 13:47) for report of a previous study of ACTH cf vigabatrin in infantile spasms. In one study, VGB was recommended as first-line therapy for infantile spasms. In another, also from Finland, asymptomatic visual field constriction is reported in 2 children treated with VGB.
Vigabatrin-induced aminoaciduria. Fourteen children treated for epilepsy in Sheba Medical Center, Tel-Aviv University, Israel, had increased urinary excretion of amino acids, particularly B-alanine, g-aminobutyric acid, and B-aminoisobutyric acid, while receiving VGB . A metabolic screen, including amino acid and organic acid analyses, is recommended prior to starting VGB therapy for seizures.
Aminoaciduria and epilepsy. Hyperaminoaciduria has been reported previously in children with absence and other idiopathic epilepsies [3, 4]. The abnormal excretion was not associated with renal disease and could not be explained as a side effect of antiepileptic medication. In fact, the hyperaminaciduria was decreased during therapy with AEDs (trimethadione, phenobarbital, mephobarbital) or the ketogenic diet, suggesting a correlation between the aminoaciduria and epilepsy.
Ketosis and epilepsy. P31 magnetic resonance spectroscopic imaging studies were performed in 7 patients with intractable epilepsy, before and after the ketogenic diet. Significant increases in high-energy phosphates and metabolism occurred with the diet. 
Epidemiology of infantile spasms. Cumulative incidence of IS among Atlanta children is 2.9/10,000 live births. Among 10-year-olds with a history of IS, 83% are mentally retarded (MR), with no difference between cryptogenic and symptomatic cases. Of 10-year-olds with profound MR, 12% have a history of IS. Fifty percent of IS patients develop Lennox-Gastaut syndrome before age 11 years.