Differential relational memory impairment in temporal lobe epilepsy.

OBJECTIVE: Temporal lobe epilepsy (TLE) is typically associated with pathology of the hippocampus, a key structure involved in relational memory, including episodic, semantic, and spatial memory processes. While it is widely accepted that TLE-associated hippocampal alterations underlie memory deficits, it remains unclear whether impairments relate to a specific cognitive domain or multiple ones. METHODS: We administered a recently validated task paradigm to evaluate episodic, semantic, and spatial memory in 24 pharmacoresistant TLE patients and 50 age- and sex-matched healthy controls. We carried out two-way analyses of variance to identify memory deficits in individuals with TLE relative to controls across different relational memory domains, and used partial least squares correlation to identify factors contributing to variations in relational memory performance across both cohorts. RESULTS: Compared to controls, TLE patients showed marked impairments in episodic and spatial memory, with mixed findings in semantic memory. Even when additionally controlling for age, sex, and overall cognitive function, between-group differences persisted along episodic and spatial domains. Moreover, age, diagnostic group, and hippocampal volume were all associated with relational memory behavioral phenotypes. SIGNIFICANCE: Our behavioral findings show graded deficits across relational memory domains in people with TLE, which provides further insights into the complex pattern of cognitive impairment in the condition.

Atypical connectome topography and signal flow in temporal lobe epilepsy.

Temporal lobe epilepsy (TLE) is the most common pharmaco-resistant epilepsy in adults. While primarily associated with mesiotemporal pathology, recent evidence suggests that brain alterations in TLE extend beyond the paralimbic epicenter and impact macroscale function and cognitive functions, particularly memory. Using connectome-wide manifold learning and generative models of effective connectivity, we examined functional topography and directional signal flow patterns between large-scale neural circuits in TLE at rest. Studying a multisite cohort of 95 patients with TLE and 95 healthy controls, we observed atypical functional topographies in the former group, characterized by reduced differentiation between sensory and transmodal association cortices, with most marked effects in bilateral temporo-limbic and ventromedial prefrontal cortices. These findings were consistent across all study sites, present in left and right lateralized patients, and validated in a subgroup of patients with histopathological validation of mesiotemporal sclerosis and post-surgical seizure freedom. Moreover, they were replicated in an independent cohort of 30 TLE patients and 40 healthy controls. Further analyses demonstrated that reduced differentiation related to decreased functional signal flow into and out of temporolimbic cortical systems and other brain networks. Parallel analyses of structural and diffusion-weighted MRI data revealed that topographic alterations were independent of TLE-related cortical thinning but partially mediated by white matter microstructural changes that radiated away from paralimbic circuits. Finally, we found a strong association between the degree of functional alterations and behavioral markers of memory dysfunction. Our work illustrates the complex landscape of macroscale functional imbalances in TLE, which can serve as intermediate markers bridging microstructural changes and cognitive impairment.

Electrodiagnostic artifacts due to neurostimulation devices for drug resistant epilepsy.

Background: Neurostimulation devices including vagus nerve stimulation (VNS), deep brain stimulation (DBS), and responsive neurostimulation (RNS) are approved therapeutic options for drug resistant epilepsy (DRE). As these devices are increasingly used in clinical practice, it is of importance to recognize their artifacts in electrodiagnostic studies. Methods: This is a retrospective study of all adult DRE patients treated with neuromodulation devices for epilepsy at our center between 2012 and 2021. Available EEGs were reviewed for neurostimulator-related artifacts. Results: Fifty-two patients were included. 37% of patients had neurostimulation related electrophysiological artifacts (20% of VNS, 75% of DBS, all patients with dual VNS-DBS treatment, and in the single patient with RNS). Artifacts were intermittent, appearing most commonly simultaenously in the EEG and ECG. VNS artifacts were monomorphic appearing mostly in the lower temporal EEG electrodes, whereas DBS artifacts were with variable morphology, amplitude, and scalp distribution. At times, the artifacts resembled electrographic seizures in the EEG and mimicked extrasystole or asystole in the ECG. Conclusions: With the increasing use of neurostimulation treatments for DRE, and the need for frequent electrodiagnostic studies in this patient population, it is important clinicians recognize these electrophysiological findings as artifacts, to avoid misdiagnosis and facilitate accurate interpretation.