Authors :
Presenting Author: Graham Johnson, PhD – Vanderbilt University
Derek Doss, BE – Vanderbilt University; Danika Paulo, MD – Vanderbilt University Medical Center; Jared Shless, BS – Vanderbilt University Medical Center; Abhijeet Gummadavelli, MD – Vanderbilt University Medical Center; Hakmook Kang, PhD – Vanderbilt University; Shilpa Reddy, MD – Vanderbilt University; Robert Naftel, MD – Vanderbilt University Medical Center; Shawniqua Williams Roberson, MD – Vanderbilt University Medical Center; Benoit Dawant, PhD – Vanderbilt University; Sarah Bick, MD – Vanderbilt University Medical Center; Victoria Morgan, PhD – Vanderbilt University; Dario Englot, MD, PhD – Vanderbilt University Medical Center
Rationale:
It has been postulated that seizure onset zones (SOZs) in focal epilepsy are actively suppressed by the rest of the brain between seizures. This is termed the Interictal Suppression Hypothesis (ISH). Evidence for the ISH includes increased inward connectivity and decreased outward connectivity to the SOZs during interictal stereotactic-electroencephalography (SEEG) recordings. However, is this reciprocally imbalanced architecture of the resting network affected during seizures? We sought to characterize the ictal SEEG network to observe perturbations from the interictal state and determine if the suppressive paradigm is altered during ictal dynamics.
Methods:
For 81 patients undergoing presurgical evaluation for drug resistant focal epilepsy at a single center, we collected the entire SEEG recordings and annotated all seizure events – 680 total seizure events captured. We then calculated the partial directed coherence (PDC) for the SOZs, propagation zones (PZs) and non-involved zones (NIZs) during a sliding window with a width of five seconds and stride of one second for the following epochs: 5 minute period four hours before seizure onset (interictal), five minutes before seizure (preictal), during seizure (ictal), five minutes after seizure (postictal). We averaged the dynamic connectivity for each epoch and conducted a two-way repeated measures analysis of variance (RM-ANOVA) to elucidate temporal and group effects. Finally, to account for power changes during the seizure, we reconducted the analysis controlling for signal power.
Results:
There is a global ictal decrease in NIZ inward and outward connectivity - exemplified in example seizure connectivity timeline (Figure 1A). However, the SOZs exhibited a dramatic increase in inward connectivity and decrease in outward connectivity during the seizure (Figure 1B). This global decrease in connectivity with paradoxical increase in inward connectivity for SOZs is demonstrated in group level analysis and was unchanged between raw and power-normalized analyses across all 680 seizures. (RM-ANOVA group p-value 2.22e-11 and 1.86e-11 respectively, Figure 2). Interestingly, the postictal connectivity architecture collapsed with SOZ/PZ/NIZs demonstrating no significant difference. This large decrease in both outward and inward connectivity in the postictal state to all regions is consistent with postictal impairment of normal brain networks.
Conclusions:
The global ictal decrease in connectivity of NIZs and dramatic increase in ictal inward-outward SOZ connectivity could indicate a focused and increased suppression of SOZs by the rest of the brain network during seizures and be characterized by an extension of the ISH to include ictal dynamics.
Funding: R01NS112252, F31NS120401, R01NS075270