Appendix e-5.
The first explanation is supported by postoperative seizures with the same preoperative semiology and frequency and by abolition of seizures after the frontal resection. The second explanation is supported by SEEG onset in the temporal lobe 10 seconds before frontal seizure activity and later propagation to the frontal depth electrodes and to the additional ICEEG subdural strip placed over the MEG-indicated frontal spike zone and thus no indication of an extratemporal source by standard criteria. 35 It is also supported by gliosis in the resected temporal specimen, which has a high correlation with an independent zone of seizure origin.
If the first, more parsimonious explanation were correct, there appear two main explanations, neither of which are cogent at this time without further data. First, the strip may not have been directly over the frontal seizure zone and thus a grid would have been essential to detect the focal frontal onset before propagation to the temporal lobe. Second, the larger surface area of extracortical grid electrodes, compared to the smaller surface area of intracortical depth electrodes, would be less sensitive to detect fine fast activity at the onset of a seizure, which would then be seen first in the temporal depth before the frontal strip. The first explanation is problematic. It suggests that the electrographic seizure would not propagate first by U fibers to surrounding gyri but rather first by the uncinate fasciculus to the temporal depth. It also suggests that the seizure would not be detected first on the orbitofrontal depth electrodes, which sampled both deep and superficial frontal cortex, before going to the temporal depth. The second explanation is also not satisfying, since again one would expect propagation to the orbitofrontal, cingulate or supplementary motor depth electrodes first before the temporal depth.
Usually our protocol has been to place an isolated strip over the MEG spike zone when the patient has proceeded to depth electrodes, to minimize mass effect and only add a burr hole. In the future we may modify this to place multiple strips through the burr hole for more complete coverage over the MEG spike zone.