17.1 Introduction

The use of electrodes to directly record from the cortex is intricately intertwined with the genesis of epilepsy surgery. The utility of electrocorticography (ECoG) for demarcating seizure-onset zones during chronic, extraoperative intracranial studies is the gold standard for localizing seizure onset, but the use of intraoperative ECoG—limited mostly to interictal epileptiform activity—to intuit epileptogenicity or make judgements about sufficiency of resection is nuanced. As with many aspects of neurosurgical technique, these can vary by institution and surgeon preference.

Whether the desired curative effects from surgery are mediated solely through removal of the offending substrate or rather through a more nuanced disruption of epileptogenic networks, is a subject of ongoing research. Regardless, seizures are an electrophysiologic phenomenon, and while macroscopic lesions are often their cause, electrographic study still offers the most accurate lens through which to delineate epileptiform activity. Intraoperative ECoG can be a useful tool for implicating areas of the brain as parts of the seizure network. It has been a critical part of the epilepsy surgeon’s armamentarium since Wilder Penfield’s pioneering work. The techniques have evolved to certain degree with technology, but they retain many of the same key aspects and tools that they had in the beginning. This chapter will detail some of the limitations, advantages, and considerations regarding intraoperative ECoG for epilepsy surgery.

17.2 Considerations and Limitations

A knowledge of the limitations of intraoperative ECoG is critical for deciding how best to use such techniques. One restriction to note is the limited window of recording offered during surgery. In contrast to chronic recordings made extraoperatively through implanted subdural electrodes, the time constraints inherent to surgery do not allow one to wait for a patient’s habitual seizures intraoperatively. Thus, inferences made in ECoG are drawn from interictal data. Interictal spikes and sharp waves do not necessarily correlate with ictal onset zone. Although likely representing an irritative zone broader and encompassing the ictal onset zone to some degree, they should not be conceptualized as directly representative of a focus of ictal epileptogenesis. ^{1 ,​ 2 ,​ 3 ,​ 4} This is a critical distinction to make, because it contextualizes intraoperative ECoG as a supplement to other extraoperative seizure-mapping studies, rather than a stand-alone diagnostic tool that can be the sole guide for resectional epilepsy surgery.

Another unique feature of intraoperative ECoG is the interplay of anesthesia with background activity and epileptiform discharges. ⁵ For example, opioid medications may enhance epileptiform activity in the cortex, increasing spike frequency such that areas with spike activity might be more easily recognized. ⁶ On the other hand, propofol, benzodiazepines, barbiturates, and etomidate are all reported to suppress epileptiform activity, which could interfere with the goals of ECoG. ⁷ Questions remain as to how well pharmacoactivated spikes correlate with native spike activity. ⁸ An example anesthesia protocol leading up to intraoperative ECoG might involve local anesthesia, nitrous oxide, narcotics, and dexmedetomidine. Anesthesiologists working on these cases should be familiar with the goals and phases of surgery so as to plan appropriately.

The importance of multidisciplinary interplay in planning for intraoperative ECoG cannot be overemphasized (Fig. 17‑1). The usefulness of intraoperative ECoG depends critically on being able to interpret the electrographic output of the studies in real time. As with other aspects of epilepsy surgery and decision-making at a large tertiary referral center for epilepsy, the input of specialized neurology colleagues is invaluable. A team of epileptologists trained in interpretation of electroencephalography (EEG) and ECoG should be available in the operating room to assist the surgeon in incorporating the data from ECoG into the surgical plan in real time (Fig. 17‑1).

17.3 Generalities of Technique

As discussed in the next section, the specifics of technique for ECoG depend on the question that the study is being used to ask. There are, however, some generalities that may be applied. Recording from the cortical surface can be carried out using either flexible strips or grids of electrodes that are composed of flat discs (the electrode contacts) of platinum, silver, or stainless steel imbedded in Teflon or silastic sheaths ⁹ (Fig. 17‑2). These are advantageous because of their flexibility and low profile. They can be slid into the subdural spaces beyond the margins of the craniotomy to record from areas more difficult to openly expose ¹⁰ (Fig. 17‑3). Moreover, these electrodes are usually designed to be used for chronic monitoring, so the decision to leave electrodes in for a long-term intracranial study can be made seamlessly during surgery. Alternatively, a fixed array of electrodes can be secured to the skull intraoperatively in a circular frame. Individual, rigid, wire electrodes can be positioned and repositioned strategically on the cortical surface, each of which are covered at the tip by a conductive material like carbon. ¹¹ This form of ECoG is only possible intraoperatively but is particularly useful for mapping of eloquent regions of cortex. The fixed array also has the advantage of being reusable.

Time is an important factor to take into consideration when performing ECoG, and standard surgical considerations regarding management of extended exposure of the cortex should be considered. There is no set time as to how long one must record from a specific area in order to characterize epileptiform activity. Immediate and early identification of epileptiform activity may immediately answer the question of whether spikes are generated in an area, but when ruling out the involvement of cortex, it may be more difficult to decide how long is sufficient. Studies using ECoG to weigh against the involvement of lateral temporal cortex in temporal lobe epilepsy (TLE) have used 5 minutes as a rough cut-off, and this seems to be a reasonable time-frame. ¹²