Abstracts

Rationale: Recent reports have described the potential application of magnetoencephalography for identification of verbal memory processing in the mesial temporal (MT) region. Using a verbal listening task requiring recognition of recurring target words across multiple trials, we previously reported a relationship between side of greatest MT activation and memory scores on the Wada test (Risse et al, 2007). Increased MT activation in the right hemisphere in epilepsy patients diagnosed with left mesial temporal sclerosis (LMTS) also has been reported using the same activation task (Ver Hoef et al., 2008). In the present study, we present a replication of Ver Hoef's findings, further validated by unilateral memory scores from the Wada test. Methods: The subjects were 14 epilepsy or brain tumor surgery candidates with either MRI diagnosis of LMTS (n=6) or no evidence of left temporal pathology (n=8). All patients were left hemisphere dominant for language based on the Wada test. The MEG unit consisted of a 148-channel Magnes 2500 WH system housed in a magnetically shielded room. The single equivalent current dipole (ECD) model across each whole hemisphere was used in the analysis. An auditory word recognition task was administered in which subjects identified target words repeated on successive trials. The hemisphere with the most MT activity sites (defined as dipoles within a one cm radius of the hippocampus) was identified for each patient. Wada memory scores were assessed with a maximum of 16 items presented prior to first motor return and calculated as a percent correct for each injection. Results: Four of 6 patients with LMTS had a preponderance of MT activation in the right hemisphere compared to 0 of 8 cases without left temporal pathology. Mean Wada memory performance by the LMTS group using the right hemisphere (left injection) was 72%, compared to a performance by the left hemisphere (right injection) of 50%. For patients without left temporal pathology, the Wada memory pattern was reversed; right hemisphere(left injection) score = 56% and the left hemisphere (right injection) score = 85%). Conclusions: These results support the likelihood that MT activation revealed with MSI in response to an auditory word recognition task, reflects functional verbal memory processing. When no temporal lobe pathology is present, this MT activation is noted predominately in the left hemisphere, while greater activity in the right hemisphere is seen in some patients with LMTS, supporting possible reorganization of verbal memory processing. This interpretation is consistent with patterns of Wada memory performance in the two groups. Current data are encouraging with respect to the development of a noninvasive procedure for assessing the integrity of recognition memory in a single hemisphere. When clinically validated, this and similar protocols may have the potential to replace the memory component of the Wada test. These results require further investigation with a larger sample size.