Temporal lobe epilepsy is the most common type of epilepsy in adults and is also associated with cognitive comorbidities, such as memory and language impairments, especially in temporal lobe epilepsy (TLE). In this population, a higher incidence of people with atypical language lateralization (ALL) is expected; 20% against 5% of the healthy people. Some factors are pointed as predictors: manual dominance, left-hand familial history, as well as early insult and left hemisphere lesion. Language in these patients has been studied for decades. Lately functional magnetic resonance (fMRI) technique has been replacing the WADA test and has provided new methods and paradigms to study this function in vivo, including in healthy volunteers. Therefore, the objective of this study was to delineate the language aspects in TLE with left (LHA) and right (RHA) hippocampal atrophy and patients without hippocampal atrophy (nonAH), comparing them to controls. The results were set out in two manuscripts. The first, showed the translation of a well-established language fMRI paradigm based on semantic decision and comparison of two versions: easy and complex task in 24 healthy subjects. Both versions were effective, but the complex version produced more robust results for the evaluation of the language in our patients. Therefore, complex version was chosen to conduct the study with TLE patients. The second result was presented in the manuscript which compared patients with RHA (n=31), LHA (n=32), nonHA (n=30) and controls (n=101). In this paper analyzes were performed to investigate the impact of HA on language to obtain a profile of the language in different subtypes of TLE. Analyzes were conducted to acquire the incidence of language lateralization, predictive factors and correlation with fMRI activation pattern in regions associated with language; comparison between groups regarding language activation pattern and functional connectivity, as well as a comparison of performance on the Boston naming test and its correlation with the task activation pattern. The results pointed to a similar frequency of ALL in all groups and only manual dominance as a predictor of hemispheric dominance for language. However, each group was different according to the specific regions of interest. Language activation pattern and ROI-to-ROI analysis showed that LHA presented reduced activation compared with controls and nonHA groups. Seed-based-voxel analysis showed that all groups of patients had reduced functional connectivity in relation to controls, as well as differences among patients. The higher scores on naming performance was correlated with greater activation in language areas only in RHA and nonHA groups. We concluded that patients with LHA presented worse performance in naming, activation and functional connectivity, followed by RHA, and nonHA patients were more similar to controls. This thesis shows that despite of several language fMRI studies, further investigations using different methodologies and paradigms are necessary for better undestanding of language alterations in different TLE subtypes (AU) |
