Abstracts

Rationale: High frequency oscillations (HFOs), reflecting the occurrence of ripples (80-250Hz) and fast ripples (>250Hz), are surrogate markers of the epileptogenic zone (Frauscher et al., 2017). Previous work has linked these markers, recorded using intracranial EEG, to MRI-based measures of hippocampal atrophy in patients with temporal lobe epilepsy, suggesting that hippocampal cell loss and synaptic reorganization may contribute to abnormal electrophysiological activity in focal epilepsy (Ogren et al., 2009). In the current work, we explored whether the association between local atrophy and HFO generalizes to the neocortex in patients with drug-resistant extratemporal epilepsy. Methods: We studied 5 women with drug-resistant extratemporal epilepsy (mean age=28.2; SD=4.5). Patients had completed high-resolution T1-weighted (T1w) MRI at the Montreal Neurological Institute and had subsequently undergone stereo-EEG implantations as part of their pre-surgical workup. Cortical thickness was measured across the cortical mantle using automated procedures, registered to a surface-based template (FWHM=10), and z-scored relative to a cohort of 43 healthy controls evaluated on the same MRI system (14 women; mean age=30.3; SD=6.4; Fig1A,C). Electrode contact locations were co-registered with each patient's native T1w scan, mapped to the cortical surface, and subsequently registered to the same surface-based template as the cortical thickness data (Fig1B,C). Results: Patients showed diverse patterns of cortical thickness alterations relative to controls. When averaging across patients, tendencies for more prominent thinning were seen in prefrontal regions, while temporal and parietal areas showed more thickening. This variability in patterns of cortical thickness alterations was likely associated with the heterogeneity of our patient sample, which presented with different lesion types and epileptogenic zone locations. Ripple (mean=8.1; SD=7.3) and fast ripple rates (mean=2.6; SD=7.1) per minute at each contact were automatically detected within a 10-minute interictal recording sample, and were also variable across patients and contacts. To account for inter-patient heterogeneity, we harnessed linear mixed-effects models that included patient identity as a random effect and HFO rates as fixed effects with cortical thickness as a dependent variable. We observed a significant association between cortical thinning and increased HFO rate. Results were consistent when using only ripple rate, fast ripple rate, or a composite measure of ripples and fast ripples consisting of the average Z-score between both markers at each contact (Fig 1D). Findings were consistent when restricting the control cohort to women only, to dispel potential effects due to sex mismatching between patients and controls. Conclusions: While based on a small sample and awaiting further replication, our exploratory results suggest a noteworthy association between MRI-derived measures of neocortical thinning and HFOs recorded with intracranial EEG. These findings suggest that cellular and synaptic reorganization processes captured by measures of cortical thinning may serve as a potential marker of epileptogenicity. Funding: Canadian Institutes of Health ResearchCanadian Open Neuroscience Platform