Abstracts

Rationale: Surgical intervention is one of the last resources offered to patients with refractory epilepsy. Although resection of brain regions has the potential to cure epilepsy, it carries the usual risks of any intervention along with the potential permanent damage to a variety of brain functions. This risk can be estimated using Electrical Stimulation Mapping (ESM), which uses cognitive tasks and cortical stimulation to identify ‘eloquent’ and ‘resectable’ areas. In spite of well over 50 years of research, a standardized evidence-based set of tasks for predicting language decline does not yet exist. Furthermore, some of the tasks are modified to accommodate and better assess the patients according to their cultural background and cognitive level. Counting numbers is often used to screen and characterize language in patients when language abilities are limited. Here we report a patient who experienced selective impairment of both spoken and written production of letters, but not numbers, during posterior superior temporal gyrus (STG) stimulation. Methods: A 45-year-old female with intractable epilepsy presented for surgical care. Approximately two decades prior she underwent resection of a left anterior temporal cavernous angioma, and seizures had subsequently recurred. She had worked professionally and cognitive function was normal with the exception of impaired naming and fluency. Language fMRI suggested left language dominance with some limited bilateral representation, predominantly in Wernicke's area. ESM was completed across three days with stimulation increasing in 1mA increments until either functional change, 12mA was reached, or after-discharges occurred. Language was screened with counting, and at the maximum stimulation level a series of language tasks were completed. Results: Stimulation of the posterior superior-temporal gyrus (STG) had no effect on counting numbers (1-20), while reliably disrupting recitation of the alphabet (A-Z) on each of two testing sessions. Writing tasks were also spared for numbers, whereas writing the alphabet both spontaneously and to dictation was selectively disrupted during stimulation. Conclusions: To our knowledge, this is the first case clearly showing dissociation of letter but not number production in multiple modalities during posterior STG stimulation. Despite the clinical use of ESM over the last century, the set of cognitive tasks applied to evaluate language function lacks standardization. The results of this case are not easily accommodated in clinical models, in which function is comprised to an anatomical area. They also highlight a need to refine these models and the inadequacy of using counting for clinical testing in presurgical ESM. Counting does not accurately index language skill and estimation of surgical risk based on number production could lead to inaccurate results and potential irreparable language impairment. Funding: American Academy of Neurology