Abstracts

Rationale: Magnetoencephalography (MEG) is recognized as a valuable non-invasive tool for localizing the epileptogenic zone in pre-surgical epilepsy patients. The sensitivity of MEG for epileptic activity has been reported in a number of studies in the adult and pediatric population. Despite such growing literature, an important factor to consider is whether these standards can be extrapolated to any age group, namely children < 2 years. Technical issues such as head size and movement have recently been investigated, with compensation methods now validated for infant MEG recordings. There are, however, no reports on MEG experience in the infant group and its clinical application in the preoperative epilepsy surgery evaluation. Methods: We retrospectively reviewed MEG recordings performed at the Children’s Memorial Hermann Hospital in Houston between June 2012 and May 2017, and identified 23 infants (< 2 years) in whom MEG was performed as part of the presurgical workup for intractable epilepsy. We reviewed medical records with regards to seizure history, work-up procedures including brain MRI and scalp EEG, surgery notes, and follow-up outcomes. Results: Of 284 children undergoing MEG between June 2012 and May 2017, 23 were infants, with a mean age of 12.6 months at the time of MEG (range 3 - 23 months). The mean age at first seizure was 4.7 months. All but one of the infants had abnormal results. Although the EEG readings often showed patterns of generalized dysfunction, MEG was able to pinpoint foci of epileptic activity in 13 cases, 8 on a lobar level and 3 on an intralobar level. When interictal EEG results showed multifocal epileptiform discharges and MEG revealed a focal onset of epileptiform discharges (cases 16, 21 & 22), the surgery became a focal resection and achieved a favorable outcome in 2/3. MRI scans performed did not show lesions in 7/23 infants: in this MRI-negative group, 57% (4/7) had focal lateralized interictal spikes on MEG with good EEG correlates, while EEG results alone found multifocal or bi-focal epileptiform discharges in all cases. Of the 7 MRI-negative cases, 3 had surgery with a good outcome (Engel I-II). Among the 23 infants, 17 had surgery, with 11 becoming seizure free at follow-up (mean follow-up of 15 months). All but two infants who had a focal lateralized spike volume on MEG and had surgery became seizure free. When assessing MEG’s ability to accurately delineate seizure focus in the 17 patients who had surgery, we defined sensitivity as the probability that MEG could detect a focal spike volume if an infant’s post-surgery outcome was good. This amounted to 66%, comparable to previously cited values of 70% in children and adult populations. Conclusions: MEG recordings in our infant population were found to be as sensitive for identifying seizure focus as other age groups. MEG identified an isolated focus in infants whose brain MRI was negative or whose EEG findings suggested a generalized pattern, highlighting the value of performing MEG in the presurgical work-up of infants with intractable epilepsy.