Abstracts

fMRI provides a non-invasive way to assess the re-organization of language networks commonly found in patients with localization related epilepsy. We sought to delineate the association between etiology and likelihood of atypical language representation. We studied 87 patients with a left hemisphere seizure focus (59 left temporal). Patients had vEEG, MRI, and fMRI (69 right handed, 18 ambidextrous or left handed; 46 males, 41 females; aged 22.3 years (range 8.5-45 years; 40[le]16years), mean seizure onset age 10.9 years (range 0.5-29 years)). fMRI, performed with whole brain EPI BOLD technique at 1.5 (50) and 3T (37), utilized a language task panel that included verbal fluency, category description, reading comprehension, and auditory comprehension and employed a box car design. Data were analyzed using semi-automated methods including SPM2 and displayed as t-maps. Individual fMRI activation maps were visually inspected and coded to determine activation in brain regions (e.g. left, right, bilateral; frontal lobe (IFG,MFG), or temporal lobe). Atypical language was defined as right dominant or bilateral activation patterns. Five percent of typically developing right handed volunteers have atypical language. Twenty-four of 87 patients (28%) had atypical language. Nine of 30 patients (30%) with normal MRI had atypical language; six of 26 (23%) with MTS; five of five (100%) with congenital stroke; none of two with vascular malformation (0%); two of fourteen 14 (14%) with tumor (1 resected age two, LST gyrus); one of eight (13%) with focal cortical/subtle dyspalsia; and one of two (50%) with a chronic inflammatory process (Rasmussen[apos]s Encephalitis). Tumors and focal dysplasia did not completely engulf any frontal or temporal language region. Stroke was the most significant factor associated with atypical language ([chi][sup2];p [lt]0.01). Atypical handedness was found more commonly in patients with atypical language ([chi][sup2];p [lt]0.006). There was a trend (t test) for earlier seizure onset with atypical language (8.0 vs 11.3 years) but onset before age six predicted atypical language ([chi][sup2];p [lt]0.002). Destructive processes that involve substantial portions of left hemisphere, such as stroke and chronic encephalitis, most readily result in language re-organization. Non-lesional neocortical epilepsy is associated with a high likelihood of atypical language, more so than other lesional epilepsy -- such as dysplasia, tumor, vascular malformation -- despite concurrent seizures. Early seizure onset and pathological lesions are important factors in language reorganization. (Supported by R01 NS44280 NINDS [amp] Clinical Epilepsy Section NINDS, NIH.)