The aim of the present study was to investigate in vivo cortical excitability in the human brain. We studied 45 consecutive patients with refractory epilepsy in whom subdural or intracerebral electrodes were implanted for assessment prior to epilepsy surgery. We compared cortical responses to single pulse stimulation (up to 8 mA, 1 ms duration) in areas where seizure onset occurred, with responses recorded elsewhere. Two main types of responses were seen: (i) 'early responses', spikes and/or slow waves starting within 100 ms after the stimulus which were observed in most regions in all patients; and (ii) 'delayed responses', spikes or sharp waves occurring between 100 ms and 1 s after stimulation which were seen in some regions in 27 patients. The distributions of early and delayed responses were compared with the topography of seizure onset. Whereas early responses were seen in most regions and seem to be a normal response of the cortex to single pulse stimulation, the distributions of delayed responses were significantly associated with the regions where seizure onset occurred. We conclude that the presence of delayed responses can identify regions of hyperexcitable cortex in the human brain. The study of delayed responses may improve our understanding of the physiology and dynamics of neuronal circuits in epileptic tissue and may have an immediate clinical application in assessment of candidates for surgical treatment of epilepsy.

Although acute electrocorticography (ECoG) is routinely performed during epilepsy surgery there is little evidence that the extent of the discharging regions is a useful guide to tailoring the resection or that the findings are predictive of outcome or pathology. Patterns of discharge propagation have, however, rarely been considered in assessing the ECoG. We hypothesize that regions where discharges show earliest peaks ('leading regions') are located in the epileptogenic zone, whereas sites in which late, secondary, propagated activity occurs have less epileptogenic potential and do not need to be excised. To allow intraoperative topographic ECoG analysis, a computer program has been developed to identify leading regions and the sites showing greatest rates or amplitudes of spikes. Their topography has been compared retrospectively with pathology and seizure control in 42 consecutive patients following temporal lobe surgery. Leading regions were most often found in the hippocampus, the subtemporal cortex and the superior temporal gyrus. The most common propagation patterns were from hippocampus to subtemporal cortex and vice versa. There was no association between seizure outcome and the location of regions with greatest incidence or amplitude of spikes or location of leading regions. There was, however, a strong and significant association between poor outcome and non-removal of leading regions other than those in the posterior subtemporal cortex. All leading regions (other than posterior subtemporal) were resected in 27 patients of whom 25 had a favourable outcome. Leading regions (other than posterior subtemporal) remained in 14 patients of whom only four had a good outcome. One patient had no epileptiform activity in the ECoG and good outcome. Persistent posterior subtemporal leading regions remained in nine subjects; all had favourable outcome (Grades I or II) but only three were seizure free. These results suggest that: (i) interictal epileptiform discharges may originate from a complex interaction between separate regions, resulting in propagation and recruitment of neuronal activity along specific neural pathways; (ii) removal of all discharging areas appears unnecessary to achieve seizure control provided that leading regions (other than posterior subtemporal) are removed; and (iii) identification of such leading regions could be used to tailor resections in order to improve seizure control and reduce neurological, neuropsychological and psychiatric post-surgical morbidity.

No abstract

Connections between human temporal and frontal cortices were investigated by intracranial electroencephalographic responses to electrical stimulation with 1-ms single pulses in 51 patients assessed for surgery for treatment of epilepsy. The areas studied were medial temporal, entorhinal, lateral temporal, medial frontal, lateral frontal and orbital frontal cortices. Findings were assumed to be representative of human brain as no differences were found between epileptogenic and non-epileptogenic hemispheres. Connections between intralobar temporal and frontal regions were common (43-95%). Connections from temporal to ipsilateral frontal regions were relatively uncommon (seen in 0-25% of hemispheres). Connections from frontal to ipsilateral temporal cortices were more common, particularly from orbital to ipsilateral medial temporal regions (40%). Contralateral temporal connections were rare (< 9%) whereas contralateral frontal connections were frequent and faster, particularly from medial frontal to contralateral medial frontal (61%) and orbital frontal cortices (57%), and between both orbital cortices (67%). Orbital cortex receives profuse connections from the ipsilateral medial (78%) and lateral (88%) frontal cortices, and from the contralateral medial (57%) and orbital (67%) frontal cortices. The high incidence of intralobar temporal connections supports the presence of temporal reverberating circuits. Frontal cortex projects within the lobe and beyond, to ipsilateral and contralateral structures.

No abstract