Abstracts

This preliminary study aimed to assess the ability of fMRI to lateralize language and memory in epilepsy patients undergoing presurgical evaluation and to compare the outcomes based on memory lateralization with either intracarotid amobarbital procedure (IAP) or fMRI.
37 healthy and 7 epilepsy subjects underwent fMRI using two fMRI-language [verb generation (VG), semantic decision/tone decision (SDTD)] and two fMRI-memory tasks [picture encoding (PE) and word retrieval (WR)] presented in block design paradigm along with control tasks. Images were acquired on a 3T Bruker MRI scanner. Regions of interest (ROI) for each task were defined based on regional activations in healthy subjects identified from the global composite map for each task (through cross-correlation analysis). The ROI were then used to calculate language and memory laterality indices (LI = (L-R)/(L+R)) in healthy and epilepsy subjects. We accepted fMRI of memory LI[le]1 SD to be symmetric (b), LI between 1-2 SD to be mildly (m) lateralizing, and LI[ge]2 SD to be strongly (s) lateralizing. LI for IAP of language and memory were calculated based on hemispheric scores using the formula above. Language and memory lateralization with IAP and fMRI and the predictions of surgical outcome based on memory lateralization with either IAP or fMRI were compared.
In healthy subjects the LI for VG=0.19 [plusmn]0.08; for SDTD=0.26 [plusmn]0.12; for PE=-0.05 [plusmn]0.06; WR=-0.01 [plusmn]0.09. In epilepsy patients, there was agreement between language lateralization with IAP and fMRI (for VG R=0.853; P=0.015; for SDTD R=0.91; P=0.004). All patients underwent surgical resection. In the first 4 patients IAP of memory was strongly lateralizing (predicting seizure-free outcome) while fMRI was symmetric in one patient and mildly lateralizing (PE) or symmetric (WR) in another patient (both had not-seizure-free outcome). In two other patients, fMRI was strongly lateralizing and predicted seizure-free outcome. Outcome data are not yet available for the last three patients (see Table).
Preliminary data suggest that fMRI of language and memory at 3T may augment presurgical staging in patients with refractory epilepsy. Although fMRI of memory was better in predicting seizure-free outcome than IAP in our study, larger groups of patients need to be evaluated before fMRI replaces IAP. [table1]
[Supported by: The Neuroscience Institute, Cincinnati, OH (JPS), NIH RO1-HD38578-03 (SKH)]