First and foremost is the fact that we have shown that the efficacy of corticosteroids in pediatric epilepsy is not limited to epileptic encephalopathies, such as infantile spasms and Rasmussen encephalitis. Rather, the effects seemed to be pronounced in focal epilepsy, including those due to focal dysplastic lesions. This is consistent with the results of recent studies suggesting the efficacy of corticosteroids in focal and generalized epilepsy of different etiologies. The impact of corticosteroids on seizures during pre-surgical subdural grid EEG monitoring in drug-resistant children was recently demonstrated. A reduced seizure frequency was found in dexamethasone-treated patient compared with untreated. The results of our experimental study support the BX517 hypothesis that inflammatory mechanisms and BBB damage could contribute to seizure generation and severity. We confirmed the pattern of WBC activation in the experimental model used. Interestingly, in rats pre-treated with dexamethasone, we observed a decrease in SE severity at time of decreased number of circulating T-cells and reduction of BBB damage. The efficacy of corticosteroids observed in patients supports the hypothesis that seizures of different etiologies may be aggravated by inflammatory mechanisms and BBB disruption. The hypothesis linking BBB damage to seizures is in agreement with histological studies that have shown BBB dysfunction in human epileptic tissue and with MRI studies showing changes corresponding to the location of EEG E2020 inhibitor activity in patients with partial status or focal epilepsy. Our clinical study design did not include the systematic evaluation of MRI before and after treatments, nevertheless the decrease in FLAIR hyperintensity concurrent with seizure reduction after steroid treatment suggests the hypothesis that restoring BBB integrity may be one of the mechanisms involved in the antiepileptic action of corticosteroids. A prospective ad hoc MR study is mandatory to confirm or disprove this notion. One obvious issue is that we used only pilocarpine as a model of seizures. This is not, in our opinion, a crucial limiting factor inasmuch that this model has been historically used to model human disease and also because several features of this model have common traits with human epilepsy.