Abstracts

Malformations of cortical development (MCDs) are frequently associated with epilepsy. Epileptogenicity may be restricted to the lesions, or also involves immediate adjacent areas and distant brain regions. Spike related EEG-fMRI studies can disclose distant responses and help understanding the dynamics of epileptogenicity in patients with MCDs. We studied 32 patients with different types of MCD. EEG and fMRI data were simultaneously and continuously acquired in a 1.5T scanner using MR compatible EEG electrodes and amplifier during 2-hour sessions. Each type of spike, based on spatial distribution and morphology, was analyzed separately and constituted one EEG-fMRI study. We looked at positive (activation) and negative (deactivation) changes in BOLD signal inside the lesion, in the perilesional areas and at distance. Eleven of the 32 patients had more than one type of spike, giving a total of 53 EEG-fMRI studies. Four studies had less than three spikes, and were not further analyzed; therefore a total of 49 studies (28 patients) were included: 4 related to focal cortical dysplasia, 11 to periventricular nodular heterotopia, 17 to polymicrogyria, 2 to hemimegalencephaly, 11 to double cortex or subcortical band heterotopia, and 4 to a ganglioglioma or DNET. BOLD responses were seen in 45/49 analyzed studies (92%): 7 studies showed only activations, 5 only deactivations and 33 had both.
Activations were observed in 40/45 studies (89%). In 9/40 (22.5%) the response was clearly not related to the lesion. In 31/40 studies (77.5%) there was lesional or perilesional responses (8 had only perilesional involvement). Activation in distant areas was seen in 38/40 studies (95%).
Deactivations were seen in 38/45 studies (84.5%). In 20/38 (52.5%) the response was clearly not related to the lesion. In 18/38 studies (47.5%) there was lesional or perilesional responses (5 had only perilesional involvement). Deactivation in distant areas was seen in 36/38 studies (94.5%). BOLD changes to interictal spikes in MCDs frequently involved the lesion and the perilesional areas. Lesional and perilesional responses occurred more often in the form of activation, and may be correlated to increased neuronal activity in those regions, reflecting hyperexcitability at the time of the spike generation. BOLD changes were also found in distant brain regions, more frequently in the form of deactivation, maybe reflecting distant inhibition, and confirming earlier results regarding the lesser specificity of deactivations. Our data support the existence of intrinsic epileptogenicity in MCDs, with indication of a widespread effect of the foci on other brain areas. (Supported by Canadian Institutes of Health Research, Preston Robb MNI fellowship and Savoy Foundation for Epilepsy.)