Authors :
Presenting Author: Lucas Sainburg, BS – Vanderbilt University
Alexandra Roche, BA – Vanderbilt University Medical Center
Ghassan Makhoul, BS – Vanderbilt University Medical Center
Catie Chang, PhD – Vanderbilt University
Dario Englot, MD PhD – Vanderbilt University Medical Center
Victoria Morgan, PhD – Vanderbilt University Medical Center
Rationale:
Interictal epileptic spikes can help localize the epileptogenic zone (EZ) for surgical planning. These spikes increase functional MRI (fMRI) activity in the EZ and decrease activity in the default mode network (DMN)1, producing instances of negative functional connectivity (FC) between these regions. Here we applied single-timepoint dynamic FC (dFC) to detect interictal spike -like events in fMRI data of patients undergoing epilepsy surgery. We hypothesized that the number of interictal spike -like dFC events, which we term the dFC spike index, would be increased in the EZ. We tested this hypothesis by comparing the dFC spike index to surgical treatment locations in a cohort of patients with temporal lobe epilepsy (TLE), followed by case studies of frontal lobe epilepsy (FLE) and temporal encephaloceles.
Methods:
This study included 109 healthy controls, 49 patients with TLE who underwent surgical resection, and a case-series of one patient with FLE and two patients with temporal encephaloceles.
Resting-state fMRI data was collected before surgery (20 minutes, TR = 2s, 3x3x4mm3). Single-timepoint dFC timeseries between the DMN and each voxel were computed2. The dFC spike index was calculated as the number of negative dFC events averaged in each region, Z-scored to healthy controls (Fig 1).
To test if the locations of increased dFC spike index in TLE corresponded to the EZ, dFC spike index was compared between TLE and controls prior to Z-scoring at each region using Wilcoxon rank-sum tests with correction. Next, area under the ROC curve (AUC) analysis determined if patients who were free from disabling seizures 3 years after surgery (Engel I) had higher dFC spike index in their resections than patients who were not (Engel II-IV), since seizure freedom indicates that the EZ was resected. To test whether dFC spike index was increased in the EZ of the case studies, treatment locations were compared to dFC spike index maps.
Results:
Patients with TLE had increased dFC spike index in their ipsilateral anterior hippocampus, posterior hippocampus and temporal pole compared to controls (pFWE< 0.05), overlapping with the patients’ EZs. Patients with TLE had decreased dFC spike index in their contralateral medial precentral gyrus compared to controls (pFWE< 0.05). Patients with TLE with Engel I outcomes had higher dFC spike index in their resection than patients with Engel II-IV outcomes (AUC = 0.80, p = 0.0002; Fig 2A-B).