Abstracts

RATIONALE: Functional magnetic resonance imaging (fMRI) has been proposed as a potential noninvasive alternative to the intracarotic amobarbital procedure (IAP); however, detecting anterior hippocampal activity during fMRI memory paradigms has been difficult. Using a memory paradigm that compares novel to previously learned words, we have found consistent left greater than right hippocampal activation in normal individuals. In this study, we compared lateralization of fMRI memory activation and IAP memory performance in a group of presurgical epilepsy patients.
METHODS: Participants were five patients (mean age 38; 3m/2f) with left temporal lobe epilepsy undergoing evaluation for surgery. IAP and fMRI were performed on different days. IAP: Participants underwent left sided IAP at a variable dosage. Sixteen memory stimuli (8 verbal, 8 visual) were presented during sedation and multiple choice recognition memory was measured following return to baseline EEG. FMRI Procedure: Fifteen minutes prior to the functional scanning, participants were shown four words repeatedly for eight minutes. Words were presented visually, one at a time, every three seconds. During the fMRI scan, participants saw novel words intermixed pseudorandomly with the overlearned (old) words and null events. Scanning parameters were as follows: TE=40ms, TR=3000ms, FOV= 240mm, matrix= 64x64, number of slices = 32, slice thickness =3.8, in-plane resolution 3.75x3.75mm, repetitions =160. The scan was repeated twice using a different set of novel words for the second scan (counterbalanced across subjects). Data from each participant were motion corrected, normalized to standard atlas space (ICBM), and spatially smoothed. An event-related analysis was used to determine the activation to the novel words relative to the old words and null events. An anatomical mask was used to constrain the analysis to the mesial temporal region. The number of activated voxels in each search region was obtained and a laterality score calculated using the formula (R-L)/(R+L).
RESULTS: The laterality scores correlated with IAP memory performance (Spearman r = .80, p=.052). Three of the five patients showed right lateralized fMRI activation and had IAP performance of 12 or better. One patient showed the opposite pattern (left more than right activity) and had a IAP memory score of nine. The remaining patient had bilateral activation with an IAP memory score of seven.
CONCLUSIONS: Mesial temporal lobe activation to novel verbal events was highly correlated with the IAP in this preliminary sample. That is, patients who performed better on the IAP, had greater right lateralized fMRI activation. This study suggests that fMRI shows promise as a less invasive alternative to the IAP in left TLE patients.
Support: Barrow Neurological Foundation