Abstracts

Rationale: Trial to trial variability in neural activity is prevalent, and is usually considered a nuisance that may decrease the fidelity of transmitted signals. However, stochastic processes may have important functional roles in the brain. The level of random background synaptic input modulates a neuron’s output gain – increasing input variance decreases output gain. Since seizures may be thought of as a derangement of gain control imparted by imbalance of excitation and inhibition, we study whether inter-trial variability of visual evoked responses predicts the occurrence of epileptiform activity in childhood absence epilepsy (CAE). Methods: Twenty-four participants aged 6-12 years and diagnosed with CAE were enrolled. Steady-state visual evoked potentials (SSVEP) were recorded using a high density EEG net. Visual stimuli consisted of a random checkerboard pattern flashing at a rate of 5.14 Hz. Stimulus contrast was swept from 0.75% to 40% during each 10 second epoch. Eighty epochs were recorded continuously with brief inter-stimulus intervals. We categorized each epoch in accordance with its temporal relationship to epileptiform discharges identified in the concurrent EEG. Epochs that overlapped epileptiform discharges were removed from analysis. Fourier analysis was used to extract the complex first harmonic SSVEP component. We quantified within-subject amplitude and phase variance using the T²_circ statistic (Mast & Victor, Electroencephalogr Clin Neurophysiol:78,1991,378-88) and the inter-trial phase coherence (ITPC), respectively. ITPC is an index of the dispersion of response phase across trials (zero when phases are randomly distributed; one when phases are perfectly aligned). Statistical significance was assessed using Wilcoxon rank sum test, controlled for multiple comparisons by false discovery rate. To avoid a confounder, comparison of ITPC was limited to conditions that did not differ in amplitude. We also performed subgroup analysis according to the number of epileptiform discharges recorded: none, rare (fewer than 5), or frequent (greater than 5). Results: In the cohort with rare discharges (N=8), the ITPC was significantly greater in those epochs immediately preceding and following a discharge than when the SSVEP was obtained away from a discharge (FDR < 0.03). This difference was evident at the lower contrasts. There was no significant difference in amplitude variance regardless of the temporal relationship to discharges. In the cohort with frequent discharges (N=5), no difference in ITPC or amplitude variance was observed. At a group level, there was no difference among those with rare, frequent, or none discharges. Conclusions: In select patients with CAE, increased response synchrony is temporally associated, though not necessarily coincident, with occurrence of epileptiform discharges. This finding suggests that response variability may play a homeostatic role to counter hyperexcitability. Funding: None