Overnight Ictal Magnetoencephalography

PRACTICAL IMPLICATIONS

In patients with drug-resistant epilepsy for whom nocturnal seizures are recorded during an admission for EEG telemetry, overnight ictal MEG is a viable methodology that can help localize the seizure onset zone.

The added value of magnetoencephalography (MEG) in the presurgical evaluation for drug-resistant epilepsy is well-recognized.^1–4 However, MEG remains mostly limited to the analysis of interictal epileptic activity.^1,5 Seizures are uncommonly captured because of logistical considerations despite mounting evidence of the value of ictal MEG in localizing the seizure onset zone.^5–7 Here, we report the recording and analysis of ictal MEG recordings in 2 drug-resistant epilepsy presurgical candidates who spent a night sleeping in the MEG.

Case 1

An 8-year-old girl with drug-resistant epilepsy of suspected right frontotemporal origin was admitted for presurgical workup. Until 4 months before this admission, the patient only had nocturnal seizures. The frequency increased to ∼10 seizures/d, and she began having daytime events that the mother described as consisting of a blank stare and nonresponsiveness. Her first MRI revealed only asymmetry between bilateral temporal sulci. After routine admission for EEG telemetry during which frequent nighttime seizures were recorded, we performed a 4-hour overnight MEG recording during sleep. One brief seizure lasting ∼45 s was captured at 2:13 am (figure 1, A), which consisted of arousal, followed by forced head version to the left. Electrographic onset preceded the head turn by ∼6 s and was characterized by rhythmic alpha activity beginning at T4 on EEG and over right temporal sensors on MEG. MEG source imaging of alpha frequencies at seizure onset suggested a right anterior temporal generator (figure 1, B). Interictal MEG was concordant with seizure onset. A second 3T MRI including 3D T1 and fluid-attenuated inversion recovery sequences showed anterior temporal blurring of the gray-white matter junction (figure 1, C). Fluorodeoxyglucose PET and SPECT scans showed right temporal hypometabolism and hypoperfusion, respectively. She underwent a tailored resection including the anterior temporal cortex all the way to midposterior temporal cortex. Surgical pathology (figure 1, D–G) from the anterior temporal cortex, which included the MEG seizure onset zone, showed focal cortical dysplasia (FCD) type IIa, whereas the posterior aspect of the superior temporal gyrus contained rare dysmorphic neurons, but no frank FCD. The patient is currently seizure-free at 13 months follow-up.

Figure 1. Case 1 Ictal MEG and Pathology.

(A) Reduced montage (76 of 275 MEG channels displayed) showing the seizure onset, followed by movement artefact. (B) Magnetic source imaging of preop ictal MEG overlaid on postop MRI. (C) Preop T1 MRI showing blurring of the gray-white matter junction in an area colocalizing with the MEG seizure onset zone. (D–F) Surgical pathology consisting of cortex and subcortical white matter from the resected anterior aspect of the superior temporal gyrus showing FCD type IIa. (G) Surgical pathology consisting of cortex and subcortical white matter from the resected posterior aspect of the superior temporal gyrus showing rare dysmorphic neurons, but no frank FCD. (D) H&E 10x. (E) NeuN 10x. (F) SMI-32 20x. (G) H&E 20x. FCD = focal cortical dysplasia; MEG = magnetoencephalography.

Case 2

A 13-year-old boy with drug-resistant epilepsy who underwent a previous stereoelectroencephalography (SEEG) study that failed to localize the seizure onset zone was readmitted for presurgical evaluation. Since 2.5 years of age, multiple EEGs recorded nocturnal electrographic seizures localizing to the left posterior temporal-occipital region. In later EEGs, clinical seizures were additionally recorded in left frontal and frontopolar regions. Interictally, abundant spikes and continuous slow waves were recorded in the left temporal region. Before the present admission, the patient was having up to 20 seizures per night. Semiology was characterized by sudden arousal from sleep, confusion, agitation, some pelvic twisting and thrusting, and dystonic posturing of the right hand. We performed a 5-hour overnight MEG recording and recorded 14 clinical seizures. One representative seizure is shown in figure 2, B (onset for all seizures in figure 2, C) together with MEG source imaging of the low-voltage fast activity in figure 2, A. Interictal MEG was not clearly localizing. FDG PET also showed diffuse left hemispheric hypometabolism (figure 2, E). Given that most seizures on EEG and MEG localized to an area of left inferior parietal/posterior temporal cortex not covered by the previous SEEG study (figure 2, D) and suspected signal abnormalities in left temporal opercular/posterior insular cortex on 3T MRI, a second SEEG implantation with broader coverage was undertaken. Although abundant seizures were recorded, the study failed to localize a focal generator. Ictal spread to the same posterior perisylvian region that we had localized with ictal MEG preceded the stereotyped clinical onset, but this occurred late in the seizure. No surgery could be offered.

Figure 2. Case 2 Ictal MEG, FDG PET, and SEEG.

(A) MEG source imaging over 5 s windows of the seizure in (B) showing ictal spread beginning in the left supramarginal gyrus. (B) Reduced montage (76 of 275 MEG channels displayed) showing seizure onset and spread. (C) MEG source imaging of seizure onset in 9 independent seizures (4 L parietal, 3 L frontal, 1 L inferior temporal, 1 R occipital) captured during the overnight ictal MEG. Five seizures were excluded because of artefact or missed onset. Each color represents a different seizure. (D) Ictal MEG source imaging plotted over first SEEG postimplantation MRI. SEEG electrodes are shown in gray, with green channel labels. (E) ¹⁸F-FDG PET showing diffuse left hemispheric hypometabolism. MEG = magnetoencephalography.

Discussion

In many tertiary epilepsy centers, MEG is presently rarely used during nighttime off-hours. We argue that overcoming the logistical challenges of such recordings presents a significant opportunity for the acquisition of routine ictal MEG. In the 2 cases presented here, we demonstrate that recording natural sleep in unsedated children overnight is both achievable and can allow for routine ictal MEG recordings in well-selected patients, such as those with sleep-related seizures.

Practically, some suggestions from our experience include adding a comfortable mattress above the standard patient support and a thin pillow in the MEG helmet to prevent discomfort from the EEG electrodes, as well as placing a second bed in the magnetically shielded MEG room for a parent to sleep next to the child. In conclusion, overnight MEG recordings in well-selected candidates are proposed as a viable methodology to obtain routine ictal MEG in the presurgical evaluation for drug-resistant epilepsy.

Appendix. Authors

Study Funding

Funding for the overnight MEG scans and ancillary expenses was provided by the Foundation of the Department of Neurosurgery, Faculty of Medicine, McGill University as well as a “New Directions in Research” award from the Montreal Children's Hospital Foundation. J.T. Moreau. was supported by the Fonds de Recherche du Québec - Santé and the Foundation of Stars. S. Baillet was supported by a Discovery Grant from the Natural Science and Engineering Research Council of Canada (436355-13), the NIH (R01 EB026299), and a Tier-1 Canada Research Chair in Neural Dynamics of Brain Systems. This research was undertaken thanks in part to funding from the Canada First Research Excellence Fund, awarded to McGill University for the Healthy Brains, Healthy Lives initiative.

Disclosure

The authors report no disclosures relevant to the manuscript. Full disclosure form information provided by the authors is available with the full text of this article at Neurology.org/cp.