Abstracts

The transition from interictal to ictal discharges is poorly understood. One possibility involves the brain changing into a facilitating state promoting seizures. Non-linear mathematical analysis of EEG frequency components has confirmed the presence of a pre-ictal state in focal epilepsy, however, spatial resolution was limited. We report on fMRI analysis of the pre-ictal state in three patients with focal epilepsy.
Patient #1 was a 22 year old male with daily complex partial seizures originating from the left face sensory area, based on video EEG-monitoring (VEM), PET, and SPECT. MR imaging showed no lesion. Surgical resection of suspected focus resulted in seizure cure, confirming clinical localization. Patient #2 was a 14 year old girl with frequent nocturnal frontal lobe seizures based on VEM, with MRI showing a possible dysplastic lesion in the right superior frontal gyrus. Ictal SPECT showed hyperperfusion in the same area. Patient #3 was an 11 year old girl with frequent nocturnal complex partial seizures originating from the right supplementary sensorimotor area based on VEM. MRI showed no obvious abnormality, but ictal SPECT showed hyperperfusion in the left (contralateral) motor/premotor area.
Blood oxygen level dependant (BOLD) fMRI was performed at 3 T, while the patients slept. Imaging was stopped at the clinical onset of a seizure. Block design analysis was carried out comparing BOLD signals for the one minute prior to seizure onset ([lsquo]task[rsquo]) to a span of one minute beginning five minutes prior to seizure onset ([lsquo]rest[rsquo]). Individual subject maps of BOLD signal changes were based on a t-test comparison between the two pre-ictal time periods (task versus rest).
One typical complex partial seizure was captured in each patient. In patient #1, a striking pre-ictal BOLD signal increase was seen over the left (ipsilateral to seizure focus) frontocentral region, maximal at the seizure focus. No significant BOLD signal decreases were observed. Patient #2 showed a robust pre-ictal BOLD signal increase over the left (contralateral to seizure focus and lesion) posterior frontal region as well as a focal BOLD decrease at the seizure focus on the right. In patient #3, a pre-ictal BOLD signal increase was seen in the premotor area contralateral lateral to the presumed seizure focus on the right. Notably, this BOLD increase co-localized precisely with the site of hyperperfusion seen on an ictal SPECT study. No significant BOLD decreases were seen in this patient.
We demonstrate that the pre-ictal state is associated with robust BOLD signal changes, although they were not consistent between patients. These findings support the concept of a pre-ictal physiological state, although the mechanisms through which the transition to seizures occur remain unknown.
[Supported by: National Health and Medical Research Council of Australia, Canadian Institutes of Health Research, Alberta Heritage Foundation for Medical Research.]