Abstracts

RATIONALE: Partial epilepsy is the most common form of epilepsy and can frequently become medically refractory. Interictal spike-triggered functional MRI (fMRI) can be a useful investigative tool to aid in the surgical planning of these patients. However, no studies have systematically investigated two common causes of refractory partial epilepsy, namely, hippocampal sclerosis and cortical dysplasia.
METHODS: Four patients with cortical dysplasia and three patients with hippocampal sclerosis with intractable seizures and frequent interictal discharges ([gt]1 discharge per minute) were recruited from our video-EEG telemetry unit. All patients had anatomical MRI and most also had PET and SPECT to facilitate seizure localization. Patients were studied two to eight weeks after discharge from hospital. Echo planar fMRI was performed at 3.0 Tesla with concurrent EEG. Blood oxygen level dependant (BOLD) weighted signals were measured. Whole brain fMR images (20-22 axial slices, 4+1 mm thick, TE/TR 40/3000 ms) were acquired immediately after interictal discharges (spike) or when at least fifteen seconds has passed without a discharge (rest). The images were then analysed using the Student[ssquote]s t-test to compare spike and baseline datasets to generate a statistical map of spike related BOLD activation.
RESULTS: The patient profiles are shown in the Table. Studies were attempted on an additional four patients with cortical dysplasia. However, the interictal discharges subsided in these patients between the time of discharge from hospital and the day of the study. Patients with cortical dysplasia and hippocampal sclerosis had 7-25 and 4-10 spike-triggered fMRI acquisitions, respectively. BOLD activation was seen in only two patients, both with cortical dysplasia. Notably, these patients had the most spike-triggered fMRI acquisitions in the study (20 and 25, see Table) The BOLD activation was centred on the dysplastic lesion in both cases.
CONCLUSIONS: Spike-triggered fMRI shows BOLD activation in patients with cortical dysplasia, but only if a large number of discharges are captured. In patients with hippocampal sclerosis, up to ten interictal discharges were not sufficient to produce BOLD activation. These results highlight the need for careful patient selection and timing of experiments when employing spike-triggered fMRI in these groups of patients.[table1]
[Supported by: National Health and Medical Research Council of Australia (Grant # 135400), Brain Imaging Research Foundation, Canadian Institutes of Health Research, and Alberta Heritage Foundation for Medical Research.]