Responsive neural stimulation for epilepsy

Scenario

A 27-year-old man who had had seizures for over 15 years was being evaluated for possible invasive monitoring to define his epileptic focus in a more localized fashion. His hope, of course, was that a resectable focus could be found, and he was looking forward to proceeding with that surgery in the near future after several years of deliberation over this option. He had had several discussions with his parents and his epileptologists (two different neurologists over the past 10 years) regarding potential risks and benefits of intracranial epilepsy surgery. One reason he decided to move forward now was that he had recently had an increase in his seizure frequency, despite recent pharmacological attempts to regain some modicum of control over them again. For a few years in his early twenties, he had only one or two seizures per year. However, more recently, seizures had increased in frequency, and in the last 6 months, they had become now almost one per week, making his ability to maintain the employment he had in a start-up company as a programmer very difficult. Although supportive, his superiors indicated they hoped he would be able to return to work within a 6-month timeframe if they gave him time off to try to resolve things better with his epilepsy.

Another aspect of his circumstances, given he had an increase in seizure frequency, was that he now was aware that even if a resectable focus was not able to be found, a new device could be implanted that would simultaneously record brain activity and then, if a seizure was noted to be starting, stimulate in an appropriate manner to minimize or stop the seizure from progressing and propagating. This was known as a responsive neural stimulation (RNS) system. This gave him an extra shot at a reasonably good outcome, although the likelihood of becoming seizure free was still fairly low without a resectable focus per se.

Although he had had several extracranial monitoring sessions through the epilepsy center and two magnetic resonance imaging (MRI) scans over the most recent 5 years (one with a specific epilepsy protocol), no lesion (such as a region of cortical dysplasia) had been found nor had a clearly reliable onset location of his seizures been located. One problem was he had two seizure types: an infrequent generalized tonic-clonic pattern, which he had n’t had in over 2 years at this point, and the second (now increasing in frequency) which typically began with an interruption of attention or speech if talking. These would resolve within 10–15 seconds most of the time but would progress with eventual head turning and arm rising on the right and lip smacking, with occasional vocalizations in less common incarnations. Some of these seizures occurred during sleep, although it was unclear how often he had nocturnal seizures. Monitoring had localized onsets to the right side but could not isolate temporal onset possibilities (either mesial or extramesial with rapid mesial spread) from potential frontal onsets, which had been more apparent in some of the recorded seizures at times. An intracranial study would be needed to better determine the actual onset, if possible, and discussion site centered on the plan for electrode placement in that initial operation.

Part of this plan included the realization that electrodes would be needed over both temporal and frontal areas and a larger grid would be necessary for more definitive mapping of cortical regions, as well as a depth electrode in the both the right and left hippocampus. As such, the surgical plan would require a craniotomy instead of just bur holes or small craniectomies to place strip electrodes and a few depth electrodes. If RNS was to be considered in the future, the ability to implant the device in the remaining area of the skull might need to be modified. In addition, a sodium amytal test was performed, which allowed the consideration of temporal and mesial temporal resection if necessary. A 256-channel high-density extracranial monitoring session was performed as well, and the more recent 3T MRI was rereviewed.

• Variation : Owing to previous surgery, there is not enough room to place the responsive neural stimulation device on the same side of the head -- It may be clear that because of difficulties healing from prior surgery or the location of the previous incisions, the new incision to place the RNS device cannot be made reasonably on the same side of the scalp. Incisions may compromise blood supplies, compromising healing, or the curvature of the open areas on the scalp for an incision and the curvature of the device itself do not match well enough. Typically, there is only a region of the parietal area that fits well enough in this regard. Sometimes the region slightly more anterior on the skull can be used as well, spanning from parietal to frontal regions. In these cases, plans must be made to accommodate the device on the opposite side of the head. Placement of the head in the Mayfield head holder and its positioning in the operating room (OR) for surgery needs to be carefully considered in such cases, allowing access for the electrode placements and their tunneling to the device. An alternative option to be considered in certain cases as well is to place the device wholly within a prior craniotomy, the opening for the tray being well within the margins of the prior bone flap size. This option works out well when it can be accomplished.