Abstracts

Rationale: Traditionally, visual object naming has been the primary task used to identify essential language cortex prior to surgical resection involving the language dominant hemisphere. However, we previously found that inclusion of auditory description naming in the cortical mapping protocol identifies additional language sites that tend to be located more anteriorly in the temporal region than sites identified with visual object naming. Given the relatively limited sample of 20 epilepsy patients in the original study, we aimed to assess the reliability of this finding by examining the topographical distribution of auditory and visual naming sites in a larger sample of patients.  Methods: Eighty-eight consecutive epilepsy patients who underwent cortical language mapping in the left temporal region were included in this retrospective analysis. All patients were native English speakers. A minimum of two trials each of visual and auditory naming was conducted at each site. For each task, electrical stimulation began immediately before presentation of pictures or auditory descriptions and terminated immediately upon the patient’s production of a correct response or at a maximum of 6 seconds. Sites were considered critical for task performance if the patient failed to name the item during stimulation, but provided the target word post stimulation. Positive sites, coded by centimeters from the temporal pole, were defined as “anterior” if 4cm and “posterior” if > 4 cm. T-tests and chi-square analyses were used to compare the distribution of auditory and visual (i.e., “dual” visual-auditory) naming sites.  Results: Fifty-six patients (35 extraoperative, 17 intraoperative, four both) had positive language findings (24 women, mean age at surgery: 35.0, mean age of epilepsy onset: 18.5).  Although both visual and auditory naming sites found across the left temporal region, on average, auditory naming sites were located more anteriorly (n = 67, mean: 4.8 cm, SD= 1.7) than visual naming sites (n = 85, mean: 59 cm, S= 1.8; P = 0.0002). Chi-square analysis indicated a greater proportion of auditory naming sites in anterior temporal cortex and a greater proportion of visual naming sites in posterior temporal cortex (c2 = 10.95, p = 0.001). Of note, patients without positive findings had fewer sites tested (mean: 20.7 SD = 11.9) than those with positive findings (mean: 28.4, SD = 18.6, P = 0.038); however, there were no significant differences in age or education between these two sets of patients. Conclusions: Consistent with earlier findings in a smaller sample, we found a topographical dissociation between auditory and visual naming sites in a larger group of epilepsy patients. Specifically, the auditory naming task identified language sites that would not have been identified using visual naming alone, and, on average, auditory naming sites were located more anteriorly relative to visual naming sites.  These results support the use of both auditory and visual naming tasks in cortical language mapping for more thorough identification of cortical areas critical for language. Funding: NIH BRAIN T35 grant: T35 AG044303