Abstracts

Rationale: Hippocampal interictal epileptiform discharges (IEDs) can disrupt memory task performance. While a negative impact of IEDs during memory retrieval has been demonstrated in multiple studies, the effect of IEDs during encoding is unclear. IED effects may be task-dependent, and traditional list learning paradigms may be limited by poor performance. The present study examines the effects of IEDs during the Visual Paired Comparison Task (VPCT). This task is based on novelty preference, which reflects the natural tendency to fixate on novel items as opposed to previously viewed items. A greater percentage of time viewing a novel item compared to a previously studied item is an implicit indication of recognition memory for the “old” item, providing a continuous outcome measure of memory performance.

Methods: Subjects with medically refractory focal-onset epilepsy completed the VPCT during intracranial EEG recordings with left hippocampal depth electrodes placed for clinical purposes. The task consisted of an encoding phase in which identical pictures were displayed side-by-side for 5 seconds, followed by a 2- or 30-second delay, after which a recognition trial was administered (Figure 1). This sequence was repeated 20 or 40 times with different stimuli, depending on subjects’ fatigue and time constraints. Using infrared-based eye tracking, a “novelty preference score” was calculated (fixation time_new/fixation time_new+old). IEDs were identified by an automated algorithm (Janca et al., Brain Topogr, 2015; 28:172–83). The relationships between IED counts during encoding, maintenance, and recognition phases and novelty preference scores were determined using an autoregressive generalized estimating equation (GEE). A secondary model assessed the effect of IED timing within the encoding period, by dividing the encoding period into one-second duration bins.

Results: Three adults participated in the study (2 male, 1 female, right-handed), all with seizure onset zones involving the left temporal lobe. Mean novelty preference was 65.4% (sd 24%). A total of 100 trials were completed across subjects; 34, 68, and 37 trials had IEDs during encoding, maintenance, and recognition periods, respectively. IEDs during encoding had a negative impact on novelty preference score (β=-0.016, p=0.003), while no significant effects of IEDs during maintenance or recognition were evident. Significant negative associations with novelty preference score were seen for IEDs at time bins 0-1, 2-3, 3-4, and 4-5 seconds during encoding (p-values< 0.001).