Abstracts

Rationale: Previous research supports the involvement of distinct neural pathways for the processing of concrete and abstract nouns. Patients with dominant hemisphere temporal lobe epilepsy (TLE) have been shown to display different inter-hemispheric and intra-hemispheric organization of these pathways. Previous research has also shown that concrete nouns are processed faster in comparison to abstract nouns in a variety of cognitive tasks; commonly referred to as the concreteness effect. Event-related functional magnetic resonance imaging (fMRI) was employed to determine whether ongoing seizures causes a reorganization of these neural pathways in patients with non-dominant (right) TLE compared to non-epileptic controls. Methods: Six patients with right TLE and eight age- and gender-matched non-epileptic controls were recruited and studied. All participants completed secondary school, were native English speakers, and were born right-handed. Event-related fMRI was performed at 3T during a lexical decision task in which participants were asked to decide between words and non-words. The word stimuli consisted of abstract and concrete nouns that were matched on printed frequency, length, and orthographic neighborhood size. Response latencies recorded during the task were used to examine the existence of the concreteness effect. Analysis was conducted to compare the regions of activation seen between the two groups of participants for abstract and concrete nouns. Results: Patient demographic data are shown in Table 1. Patients with right TLE had significantly slower reaction times in comparison to non-epileptic controls for both abstract and concrete word stimuli (p < 0.05). Concrete nouns were processed significantly faster than the abstract nouns for both the right TLE patients and the controls (p < 0.001). Preliminary group analyses comparing fMRI activation patterns where p = .01 (uncorrected), showed a significant reduction in specific cortical regions in patients with right TLE for abstract nouns. Activation within the right middle superior temporal gyrus, left superior temporal gyrus as well as the left middle and inferior frontal gyri was reduced in patients with RTLE in comparison to non-epileptic controls. Conclusions: These results suggest that patients with non-dominant (right) TLE show intra-hemispheric reorganization for the processing of abstract semantic information. However, the patients’ longer reaction times suggest that this reorganization does not completely compensate for normal language processing. Thus, the presence of seizures in the non-dominant hemisphere appears to have a detrimental effect on some aspects of language function. (Sources of funding: Canadian Institutes of Health Research and Alberta Heritage Foundation for Medical Research)