Magnetic source imaging versus intracranial electroencephalogram: Neocortical versus temporolimbic epilepsy surgery

Knowlton and colleagues’ article provides welcome evidence supporting the notion that intracranial electroencephalogram (ICEEG) recording of ictal onsets may not always be necessary in patients with intractable epilepsy and nonlocalizing magnetic resonance imaging findings. For patients with neocortical epilepsy, these results are consistent with evidence indicating that interictal epileptiform abnormalities may be more predictive of surgical outcome than ictal ICEEG localizations, and as such, one should, indeed, consider the dispensability of ICEEG in this setting. However, one cannot go too far with the notion of ICEEG dispensability, and the title of Knowlton and colleagues’ article does not differentiate neocortical from temporolimbic epilepsy, although the magnetic resonance imaging entry criteria will have excluded most of the large number of patients with bilateral temporolimbic epilepsy from their study. For these patients, magnetic source imaging (MSI) of interictal epileptiform discharges offers nothing to replace the need for ICEEG recording of ictal onsets. Whereas the importance of interictal spikes may supersede that of intracranial ictal onsets in neocortical epilepsy, the converse is true with respect to temporolimbic epilepsy: The lack of prognostic value of interictal spike localizations in patients with ICEEG-confirmed temporolimbic seizure onset is well accepted. Moreover, patients with bilateral temporolimbic epilepsy will inevitably demonstrate bilateral, independent interictal spikes if a sufficient duration of recording is obtained. MSI of these spikes leaves one no further ahead in surgical planning. Relatedly, even for ipsilateral localization purposes, the discordant MSI and ICEEG findings in the three patients that Knowlton and colleagues discussed in detail highlight the need to consider temporolimbic epilepsy as an entity distinct from all neocortical epilepsies. The two patients with mesial temporal seizure onsets and remote areas of interictal spikes seemingly responded well with anteromesial temporal resections, whereas the patient with a large frontal resection sparing areas of posterior frontal and parietal spikes responded poorly. Knowlton and colleagues are to be congratulated for giving us evidence to begin to consider replacing ICEEG with MSI in many patients with intractable epilepsy, with the caveat that we acknowledge this advance does not pertain to the common situation of bilateral temporolimbic epilepsy, and the plea that we increasingly try to discuss neocortical and temporolimbic epilepsy as utterly distinct clinical entities. Succeeding in the latter hopefully should serve to strengthen the clarity of our thinking regarding the surgical management of epilepsy in the future.