Authors :
Presenting Author: Ryuzaburo Kochi, MD, PhD – Wayne State University
Zhong Irene Wang, PhD – Epilepsy Center, Neurological Institute, Cleveland Clinic
Spencer Morris, MS – Epilepsy Center, Neurological Institute, Cleveland Clinic
Balu Krishnan, PhD – Cleveland Clinic Foundation
Demitre Serletis, MD, PhD – Cleveland Clinic
Imad Najm, MD – Cleveland Clinic
Juan Bulacio, MD – Cleveland Clinic
Andreas V. Alexopoulos, MD – Epilepsy Center, Neurological Institute, Cleveland Clinic
Richard Burgess, MD, PhD – Epilepsy Center, Neurological Institute, Cleveland Clinic
Hiroatsu Murakami, MD, PhD – Epilepsy Center, Neurological Institute, Cleveland Clinic
Rationale: Magnetoencephalography (MEG) is a powerful tool for the presurgical evaluation of medically intractable epilepsy.
Utilizing MEG dipole cluster findings to inform intracranial EEG and subsequent resections has been shown to increase the likelihood of achieving seizure-free outcomes.
However, not all MEG clusters are the same. Some patients became seizure-free without including MEG clusters in the resection. This phenomenon may indicate that certain MEG clusters may represent only part of a broader epileptic network.
In this study, in a large cohort of medically intractable focal epilepsy patients where gold standard epileptogenic zone localization was available, we aimed to examine the correlation between MEG cluster characteristics and the epileptogenic zone.
Methods: This retrospective study included 101 patients who achieved at least one year of seizure freedom after surgery. All patients underwent preoperative MEG, intracranial EEG evaluation, and resective/ablative surgery. MEG data were collected using a 306-channel whole-head MEG system. SECD model was used for source localization. Dipole clusters were classified based on anatomical location, tightness (tight or loose), and orientation (stable or variable). Clusters were defined as “matched” if ≥70% of dipoles were within the surgery cavity, and “mismatched” otherwise (Figure 1).
Intracranial EEG findings in the mismatched clusters were examined. Statistical analyses were performed to compare the characteristics between matched and mismatched MEG clusters.
Results: Sixty-four patients had single clusters, and 37 had multiple clusters. There was no significant difference in detecting matched clusters between single-cluster and multiple-cluster groups. However, the rate of matched clusters was significantly lower in frontal (p=0.004) and parietal (p=0.017) resections compared to temporal resections. Among multiple-cluster patients, those with unilateral clusters trended towards a higher likelihood of matched clusters compared to those with bilateral clusters (92% vs. 63%, p=0.066). Patients without matched clusters had longer epilepsy durations (p=0.045). Tight clusters were more likely to be matched (p=0.028). Clusters with stable orientations were also more likely to be matched (p=0.041). The combination of cluster tightness and stable orientation significantly correlated with concordance to the resected area (p=0.048). SEEG sampling of 49 mismatched clusters revealed that 92% indeed exhibited interictal discharges, and 63% were involved in seizure propagation, indicating that the MEG clusters represented the irritative zone and/or propagation. Figure 2 illustrates a “connectome” of resected region and the location of mismatched clusters. This tool may be used prospectively to predict the potential origins of mismatched clusters, particularly when they do not align with the clinical hypothesis.
Conclusions: This study underscores the clinical significance of MEG cluster characteristics in identifying the epileptogenic zone. Tight and stable orientation clusters remain robust indicators of correct identification of the epileptogenic zone leading to seizure freedom.
Funding: No funding.