Abstracts

Rationale: Rolandic epilepsy (RE) belongs to a group of disorders in the epilepsy-aphasia spectrum characterized by a range of language processing and attention symptoms. Auditory mismatch negativity (MMN), the component of the event related potential elicited by a deviant sound after several standard stimuli, is postulated to reflect typical attention and language processing. In contrast to behavioral measures, MMN provides an objective and reproducible measure that can be obtained without a behavioral response or active attention. Given the cognitive features observed, we hypothesized that MMN would be reduced in RE compared to controls.


Methods: We prospectively recruited children with RE (N=10) and age-matched controls (N=17). All subjects underwent an MMN paradigm during a MEG recording with a 306-sensor system while watching a silent movie. Subjects then underwent a 3T structural brain MRI. The MMN paradigm consisted of 272 standard auditory stimuli interrupted by 68-oddball stimuli at randomized intervals. The standard stimulus was a complex tone consisting of ten tones between 500 Hz – 700 Hz. The oddball stimulus was the summation of each of the ten standard stimuli with frequencies increased by 30%. The stimulus duration and interstimulus interval were 50 ms and 900 ms, respectively. Data were Maxwell filtered with continuous HPI correction and bandpass filtered (0.2Hz-40Hz). Artifactual channels were removed. Trials were epoched (-200 ms to 500 ms from time of stimulus) and rejected if any amplitude exceeded 5% of the highest values from all epochs. 68 standard and 68 oddball trials were averaged. MEG source estimates were obtained using the DSPM algorithm in MNE software. Source data from the averaged trials were extracted for each subject from the vertex with the maximal negativity in the auditory cortex between 90-110 ms post stimulus (N100 response). Subjects were excluded from analysis if an N100 evoked response was not elicited (N=1 controls, N=3 RE). Within and between groups, evoked MMN responses were compared between standard and oddball stimuli using 2 tailed t-tests at each time point (uncorrected).


Results: Preliminary results from 6 controls (3F, 12.8± 2.2 years) and 5 RE subjects (4F, 13.3 ± 2.5 years) are reported. On group average, a MMN response was observed in the control group from 257-373 ms (p< 0.05), peaking at 293 ms (p< 0.001). On group average, no evidence of an MMN response was observed in the RE group at any time point following the N100 (p>0.7). The MMN was larger in the control group than the RE group between 252 ms and 323 ms (p< 0.05), peaking at 290 ms (p=0.002).


Conclusions: We identified group differences in the MMN responses in the RE group compared to controls. This finding suggests abnormal processing of auditory input. Future work could explore relationships between the MMN response and attention and language processing symptoms.


Funding: NIH NINDS R01NS115868