Objectives: Uncontrolled seizures in children 1 to 36 months old have serious short-term health risks and may be associated with substantial developmental, behavioral, and psychological impairments. We evaluated the effectiveness, comparative effectiveness, and harms of pharmacologic, dietary, surgical, neuromodulation, and gene therapy treatments for infantile epilepsies.
Data sources: We searched Embase®, MEDLINE®, PubMed®, the Cochrane Library, and gray literature for studies published from January 1, 1999, to August 19, 2021.
Review methods: Using standard Evidence-based Practice Center methods, we refined the scope and applied a priori inclusion criteria to the >10,000 articles identified. We ordered full text of any pediatric epilepsy articles to determine if they reported any data on those age 1 month to <36 months. We extracted key information from each included study, rated risk of bias, and rated the strength of evidence. We summarized the studies and outcomes narratively.
Results: Forty-one studies (44 articles) met inclusion criteria. For pharmacotherapy, levetiracetam may cause seizure freedom in some patients (strength of evidence [SOE]: low), but data on other medications (topiramate, lamotrigine, phenytoin, vigabatrin, rufinamide, stiripentol) were insufficient to permit conclusions. Both ketogenic diet and the modified Atkins diet may reduce seizure frequency (SOE: low for both). In addition, the ketogenic diet may cause seizure freedom in some infants (SOE: low) and may be more likely than the modified Atkins diet to reduce seizure frequency (SOE: low). Both hemispherectomy/hemispherotomy and non-hemispheric surgical procedures may cause seizure freedom in some infants (SOE: low for both), but the precise proportion is too variable to estimate. For three medications (levetiracetam, topiramate, and lamotrigine), adverse effects may rarely be severe enough to warrant discontinuation (SOE: low). For topiramate, non-severe adverse effects include loss of appetite and upper respiratory tract infection (SOE: moderate). Harms of diets were sparsely reported. For surgical interventions, surgical mortality is rare for functional hemispherectomy/hemispherotomy and non-hemispheric procedures (SOE: low), but evidence was insufficient to permit quantitative estimates of mortality or morbidity risk. Hydrocephalus requiring shunt placement after multilobar, lobar, or focal resection is uncommon (SOE: low). No studies assessed neuromodulation or gene therapy.
Conclusions: Levetiracetam, ketogenic diet, modified Atkins diet, and surgery all appear to be effective for some infants. However, the strength of the evidence is low for all of these modalities due to lack of control groups, low patient enrollment, and inconsistent reporting. Future studies should compare different pharmacologic treatments and compare pharmacotherapy with dietary therapy. Critical outcomes underrepresented in the literature include quality of life, sleep outcomes, and long-term development.