The Current State and Future Promise of Magnetoencephalography (MEG) in Epilepsy and Beyond

Richard C. Burgess, MD, PhD

To really understand normal brain function and how to provide optimal treatment when brain function goes awry, we need to know what is happening dynamically in time and space: What kind of activity is occurring and when—on a millisecond by millisecond basis—does it occur? Where is the activity taking place, and what are the spatiotemporal relationships and interactions of the involved structures? This information about both normal functioning and abnormal brain processes is needed for clinicians treating patients and scientists investigating how the brain operates.

Answering precisely these questions is where Magnetoencephalography (MEG) excels. Functional imaging modalities such as fMRI, PET, SPECT, etc have poor time resolution and are only indirect measures of neuronal function—unlike MEG. Both EEG and MEG measure the electrical activity of neurons (by picking up the electric potentials and the corresponding magnetic field respectively), directly recording the communications between neurons.

Andreas V Alexopoulos, MD, MPH

Despite the development, commercialization, and now widespread prescription of many new antiseizure medications in recent decades, 30% to 40% of epilepsy patients exhibit drug-resistance and remain vulnerable to continuing morbidity—and even mortality. Within this group, resective or ablative procedures can offer a cure in a subset of patients with drug-resistant focal epilepsy. The key to successful outcome—both in terms of seizure freedom and minimal neurological deficit post-operatively—is precise localization of the epileptogenic zone, along with the nearby eloquent cortex. The core of the presurgical evaluation is noninvasive and consists of inpatient VEEG monitoring over multiple days for characterization and localization of the patient’s spells, anatomical and functional MRI to identify structural lesions and areas of physiological activation, nuclear medicine scans (PET, ictal SPECT, etc) to reveal areas of abnormal function in the interictal and ictal state, as well as neuropsychological testing.

Cleveland Clinic, as a National Association of Epilepsy Centers level 4 institution (the most comprehensive level), not only provides a range of therapeutic options (including epilepsy surgery and neurostimulation), but also utilizes long-term monitoring with intracranial electrodes when appropriate, to pinpoint in three dimensions the area responsible for the patient’s seizures. At Cleveland Clinic, MEG (the newest noninvasive diagnostic technique) and stereo-EEG (SEEG, the most advanced and accurate intracranial diagnostic technique) were launched at approximately the same time in our Epilepsy Center. In parallel with advances in intracranial monitoring, MEG has become an important part of our standard of care, especially to identify the targets for implantation of SEEG electrodes. Though potential hardware and software advancements may make MEG more accessible in the future, MEG is currently not widely available, and only a small proportion of the most experienced epilepsy centers offer this highly advanced technology.

As evidenced by the growth of MEG in epilepsy programs both within and outside the United States, MEG has found its most fruitful clinical application in epilepsy. We have carried out more than 3,000 MEG studies since the MEG lab’s inception in 2008, and currently perform about 30 studies per month (with 70% as outpatients including outside referrals, and 30% as inpatients). Patients who have had MEG studies range in age from 6 months to 78 years (35% pediatric, 65% adult). About 35% of patients have had previous neurosurgical procedures.

At our center, MEG for guidance of SEEG implantation has become an indispensable precursor, which is then incorporated into the planning along with all of the other localizing data using sophisticated multi-modality image integration and processing. We have conclusively demonstrated that the systematic use of MEG improves the yield of complex and costly intracranial evaluations and the outcomes following surgical intervention, and as such provides an avenue for cost-effective use of presurgical and surgical tools.¹ It is also important to note that the improved and expanded use of MEG in complicated epilepsy cases provides a higher diagnostic yield and more precise localization of epileptic activity than does scalp EEG, even in patients who do not go on to intracranial evaluation.

MEG is a noninvasive and passive diagnostic test, which is simple and painless for the patient. Like EEG, MEG offers a very high temporal resolution but has higher spatial resolution because it measures the intracerebral currents without any disruption from the high resistivity of the skull, which is not the case with EEG. Another of the advantages of MEG is that there is no exposure to radiation, external magnetic fields, needles or other unpleasant procedures—the MEG apparatus is passively "listening” to the usual operation or state of commotion within the brain. By presenting tasks or stimuli to the patient, we can also investigate the time-course and location of sensory, motor and cognitive processing—helpful for presurgical mapping to identify eloquent cortex adjacent to brain tumors or other lesions. Many of our patients come to us with implanted devices, both in the brain (such as DBS, RNS, VNS, shunts), as well as elsewhere in the body (such as pacemakers, AIDs, loop recorders, pumps, orthopedic or dental implants). And many pediatric or cognitively impaired patients are unable to remain perfectly still during MEG recording. Because of the noise-reduction and movement-compensation techniques employed at our center, these obstacles can almost always be overcome to obtain a satisfactory recording.

MEG is an ideal tool for studying brain connectivity in health and disease. Sophisticated approaches to signal analysis have led to advancements in the study of brain networks and their dynamic interactions. Work is underway for application of MEG to other neurological disorders, such as neurodegenerative diseases (including Alzheimer disease and Parkinson disease), autism, traumatic brain injury and psychiatric conditions.

Our ongoing research in patients with focal epilepsy has found that "resting state” MEG activities could be used to localize the epileptic region, even in the absence of any identifiable epileptic discharge during the MEG recording. In a separate line of work we have been able to show the anomalous processing of sensory stimuli in patients with chronic pain syndromes using MEG.Cleveland Clinic is planning to construct a new Neurological Institute building, which will enable more innovative, multidisciplinary research and even better integration—both in terms of physical space and neuroscience disciplines. We hope to be able to track brain development as children mature, and brain deterioration as patients age, and to also track improvement as these patients are treated. MEG could potentially reduce therapeutic trial and error, by predicting a patient’s suitability for a given treatment, and thereafter by titrating medication dosages or neurostimulation parameters.

At Cleveland Clinic, our facility with both MEG and SEEG developed in parallel, and our capability has been expanded by fusing these diagnostic techniques with advanced multimodality image integration.SEEG has reduced the invasiveness of our intracranial evaluations, while providing even better localization information, but is costly, labor-intensive, and carries some risk for complications. Perhaps in a subset of patients, MEG will soon eliminate the need for any invasive intracranial recordings. More broadly, MEG is uniquely positioned to provide an ever-expanding window into our healthy brain and its afflictions.

REFERENCE
1. Murakami H, Wang ZI, Marashly A, et al. Correlating magnetoencephalography to stero-electroencephalography in patients undergoing epilepsy surgery. Brain. 2017;140(3):e20