Abstracts

Rationale: Rolandic epilepsy (RE, also called self-limited epilepsy with centrotemporal spikes, SeLECTS) is the most common focal epilepsy syndrome, characterized by seizures and epileptiform activity from the perisylvian cortex. We have recently identified a focal spindle deficit in children with RE, revealing thalamocortical circuit dysfunction. Sleep spindles are generated by the thalamic reticular nucleus (TRN), a key structure that also gates thalamocortical information. We hypothesize that children with RE would have impaired sensory gating that could predict attentional difficulties common to this disorder.

Methods: We prospectively recruited children with RE (N=10, 6F, 12.3 ± 2.3 years) and age-matched controls (N=14, 6F, 13.6 ± 3.4 years). RE patients were classified as active (remission) based on either (a) a clinical definition of being seizure-free < 1 year (≥1 year); or (b) an EEG definition based on the presence (absence) of epileptiform discharges during sleep. All subjects had MEG recordings with a 306-sensor system while passively listening to ~100 paired auditory stimuli. This included an initial auditory click (stimulus 1, S1) followed by an identical second click (stimulus 2, S2) after an interstimulus interval of 500 ms with intertrial intervals of 8-10 s. Then, subjects took Connors’ Continuous Performance Test (CCPT) and had a high-resolution anatomical MRI. For CCPT, subjects were shown 360 letters (duration 250 ms, intertrial interval of 1, 2 or 4 s) and told to press a spacebar for every non-X letter. MEG data was bandpass filtered (5-50 Hz), bad channels removed, then epoched into single trials with epochs rejected if the peak-to-peak amplitude exceeded 1000 fT/cm in MEG gradiometers and 10000 fT in MEG magnetometers. MEG source estimates were obtained using MNE software. In source space, the P50 response was the maximal positive cortical response near the auditory cortex between 50 and 100 seconds after S1 and S2 at the same vertex. A difference in the sensory gating ratio (S2/S1) between groups was tested with a linear mixed effects model, where increased values indicate worse thalamic gating. Commission errors, omission errors, average reaction time, and variability from CCPT were compared between groups using linear models. The relationship between SG ratio and attentional measures was evaluated within each group using linear models with age and gender included as covariates.

Results: Children with active and resolved RE, using either clinical (p=0.029, p=0.010) or EEG definitions (p=0.019, p=0.024), have impaired sensory gating compared to controls (Figure 1A-B). There was no difference in attention measure between groups. Among children with RE, shortened reaction times and higher commission errors correlated with increased sensory gating deficits (Table 1).

Conclusions: Children with RE have a sensory gating deficit that persists beyond seizure resolution. This finding implicates TRN dysfunction and may contribute to the attentional deficits observed in this developmental epilepsy.

Funding: NIH NINDS R01NS11586