Abstracts

Rationale: Cognitive impairment is a common comorbidity in medically refractory epilepsy. While it is clear that etiology and recurrent seizures often underlie the type and severity of permanent impairment, the possible role of interictal spikes (IISs) in transient cognitive deficits remains obscure. Studies have demonstrated that IISs result in general transient cognitive impairment as well as specific behavioral deficits (Holmes and Lenck-Santini Epilepsy & Behavior 2006). However, these studies have predominantly used scalp EEG in humans (Binne et al. Epilepsy Res 1987) or intracranial EEG in animal models (Kleen et al. Ann Neurol 2010). Scalp EEG limits the ability to localize IISs inducing cognitive disruption, while animal models are unable to illustrate more subtle impairments. In this study we characterized the effect of IISs on short-term visual memory using human intracranial EEG data, thereby providing the groundwork for developing therapies. Methods: Fourteen epileptic patients implanted with subdural and depth electrodes for seizure localization were tested on one of two multi-item short-term memory tasks. Patients were presented with a series of images on a computer screen. After a short delay (<1s), the patients were instructed to recall and report the image order using the buttons on a keypad. IISs were detected semi-automatically using a pattern-adapted wavelet-based IIS detection algorithm. Only channels with IIS rates >1/minute were included in the study. Results: The most frequently sampled regions included the inferior temporal gyrus (n=10), superior temporal gyrus (n=10), temporal pole (n=5), middle temporal gyrus (n=4), parahippocampal gyrus (n=4). Patients that did not learn the task (recall accuracy <=60%, n=5) had significantly higher IIS rates compared to patients who learned the task (accuracy>60%, n=9; p<0.001, Permutation test,), and recall performance was negatively correlated with IIS rate (r=-0.41, p<1e-6, 14 patients). Furthermore, the rate of IISs during incorrect trials was significantly higher than in correct trials (p<0.001, Permutation test), particularly in electrodes in the superior temporal and middle temporal gyri. To further evaluate the effect of IIS on recall performance, we compared the IIS rates during three task-relevant epochs: pre-trial baseline, the image presentation, and recall. IIS rates were significantly correlated with recall behavior in the recall and image presentation epochs but not in the pre-trial baseline epoch. Conclusions: Our results provide evidence that IISs are associated with transient disruption in short-term visual memory and IIS may contribute to cognitive impairment in epilepsy depending upon when and where they occur. These results suggest that therapies targeting IIS may be effective in increasing the quality of life for epilepsy patients.