Abstracts

Rationale: Few studies have examined use of multiple techniques to map eloquent cortex in young children prior to epilepsy surgery (Stapleton 97; Kesavadas 07). We report on congruency of results of three methods of speech/language localization, functional MRI, cortical stimulation and high frequency ECoG mapping (Miller 07) and their effectiveness with a preliterate child with intractable epilepsy and focal cortical dysplasia. Methods: The patient, age 6, a right-handed female preschooler, had clusters of complex partial seizures up to 40/day, on Topomax and Keppra with Diastat prn. Video EEG telemetry indicated left central region seizure onset. MRI identified left frontal dysplasia. Neuropsychological evaluation yielded Full Scale IQ=75 and adaptive level of 3 yr, 9 mo. She was cooperative but distractible and easily confused. Functional MRI tasks included finger to thumb tapping alternating with the opposite hand, and silent verb generation from pictures alternating with cross fixation. Non-novel stimuli were chosen for reliability of patient verbal response. She rehearsed tasks extensively in office, scanner and home, where parents presented visual stimuli on laptop and she practiced holding still. The fMRI was performed at 1.5 T. Data analysis of fMRI with Siemens commercial software generated maps of t-statistics for activation. After the fMRI, picture naming for mapping was introduced which she rehearsed repeatedly. She was implanted with a left central intracranial 64-electrode grid with strip and depth electrodes. Using the grid, speech mapping was performed with cortical stimulation and high-frequency ECoG examining the >70 Hz ECoG signal triggered to onset of successful picture naming. Results: The child cooperated with all procedures. The fMRI verb generation task yielded bilateral activation (left superior temporal, right anterior and right parietal). Cortical stimulation mapping identified left superior temporal language and posterior frontal speech motor areas. High frequency ECoG identified left temporal areas with no frontal activation beyond motor areas. The patient underwent resection of the frontal ictal area without post operative alteration in language and has been 8-mo. seizure-free. Neuropsychological re-evaluation yielded FSIQ=75 and better memory scores. Conclusions: Functional mapping for language can be performed with careful planning and extensive rehearsal in young, preliterate children. The bilateral activation seen with verb generation on fMRI in this child with left lesion and early seizure onset is consistent with previous reports of atypical language patterns (Gaillard 07). Three methods of speech localization were congruent in identifying a left superior temporal language site; use of several methods enhanced chances of an accurate map. High-frequency electrocoticography mapping, although novel, may be particularly useful with a pediatric poulsation. Gaillard et al, Neurology 2007;69:1761. Kesavadas et al; Ped Rad 2007;37:964. Miller et al, J Neurosci 2007;27:2424. Stapleton et al, Ped Neurosurg 1997;26:68.