JLE

Epileptic Disorders

MENU

Cortical surface intracranial electrodes identify clinically relevant seizures missed on scalp EEG after traumatic intracranial hemorrhage Volume 20, numéro 6, December 2018

Illustrations


  • Figure 1

  • Figure 2

All intracranial hemorrhages may provoke seizures (Young et al., 1996; Annegers et al., 1998; Claassen et al., 2003, 2007; Rudzinski et al., 2011). The factors which lead to acute seizures or chronic epilepsy are an active field of study (Joseph et al., 2016; Pollandt et al., 2017). Traumatic brain injury (TBI) severity and spontaneous hemorrhage expansion have been shown to correlate with the risk of both acute seizures and epilepsy (Annegers et al., 1998; Claassen et al., 2003, 2007; Frey, 2003; Temkin, 2003; Rabinstein et al., 2010; Rudzinski et al., 2011;). Subdural hematoma (SDH) and epidural hematoma are common causes of death and disability, and delayed complications often present a major challenge. Epileptic complications are common after acute SDH evacuation, with seizures or epileptiform changes occurring in 25% of patients. Published data suggest that this is frequently overlooked in the care of these patients (Rabinstein et al., 2010). Seizures and epileptiform changes following SDH evacuation have a strong association with lower Glasgow Coma Scale scores after surgery and are more common after hematoma evacuation by craniotomy (Rabinstein et al., 2010).

A growing and rapidly evolving body of evidence suggests that a significant proportion of encephalopathy seen in patients after intracerebral and subarachnoid hemorrhages are due to clinically undetected ictal activity (Claassen et al., 2003, 2007, 2013a; Fabricius et al., 2008; Mikell et al., 2016). This has led to recommendations to consider EEG monitoring in comatose intracranial hemorrhage patients (Brophy et al., 2012; Claassen et al., 2013b). Yet such patients are not universally monitored with EEG in the absence of convulsive seizures, and even when monitored, it appears that a proportion of seizures go undetected on scalp EEG. Waziri et al. demonstrated this using bedside placement post-TBI of transcortical EEG (depth) electrodes, in which eight of 14 patients had electrographic seizures not seen on scalp EEG (Waziri et al., 2009). Discrepancies between scalp and intracranial EEG (IEEG) are well known (Tao et al., 2005; Brophy et al., 2012) in patients undergoing epilepsy surgery evaluation with implanted electrodes. Comatose or otherwise encephalopathic patients after evacuation of SDH often have abnormal scalp EEG but no detected seizures despite prolonged monitoring (Rudzinski et al., 2011).

We report the case of a comatose patient who underwent evacuation and decompressive hemicraniectomy for a traumatic epidural and subdural hematoma, whose IEEG showed frequent non-convulsive seizures that were not detected on concurrent scalp EEG. The patient had improvement in his state of consciousness after treatment with antiseizure medications aborted the seizures.

Methods

This is a case report of a single patient that is part of an ongoing research project. To test our hypothesis that seizures and other electrophysiological disturbances are an under-recognized contributor to coma, focal deficits, and prolonged encephalopathy following SDH evacuation, we are performing an exploratory study to obtain IEEG on patients requiring emergent decompressive post-traumatic hemicraniectomy. The goal of this research is to detect and characterize these electrophysiological disturbances. The study protocol was approved by the Mayo Clinic Institutional Review Board.

Informed consent was obtained from the next of kin simultaneously with the initial portions of the emergency surgery. After the hemicraniectomy is drilled and the dura mater is opened widely in the standard fashion, a 1-lead 4-contact cylindrical electrode with platinum contacts (Ad-Tech Medical, Racine, Wisconsin, USA) is laid over each of the ipsilateral frontal cortex and motor strip, based on the anatomical estimation of the surgeon (supplementary figure 1). Practical concerns limit the number of cortical regions that can be monitored in this protocol, and the frontal and motor regions are generally visualized in craniotomies performed for trauma and are of high clinical yield. Regions of resected, cauterized, severely contused or friable brain are generally avoided (analogously to subdural surgical drain placement), but otherwise there is no specific targeting of injured anatomy. They are tunneled through the skin in a pre-sterilized area adjacent to the incision. As the dura is left open in decompressive hemicraniectomy to control intracranial pressure, the practice at our institution is to lay a single layer of thin oxidized regenerated cellulose sheets (Surgicel Original Absorbable Hemostat; Ethicon Inc., Somerville, New Jersey, USA) over the exposed cortex to improve hemostasis and provide an additional barrier between scalp and brain to facilitate cranioplasty. The IEEG leads are placed under this layer, which helps secure them in place during closure. A tin disk reference is placed on the scalp at the CPz position and EEG data is acquired in reference through an XLTEK EMU40X amplifier and displayed electronically using Natus NeuroWorks Software (Natus Medical Incorporated, Pleasanton, California, USA). EEG recordings from the scalp and electrodes are obtained simultaneously.

Following wound closure, the patient is transferred to the neurosciences ICU. A post-operative non-contrast computed tomography (CT) scan of the head is obtained, after which scalp macroelectrodes are placed in conventional fashion to provide correlation with the cortical EEG and to aid treatment of any macroscopic seizures. The imaging data is collected to confirm the cortical location of the IEEG leads. Both the IEEG and external EEG leads are maintained for a maximum of five days or until the patient is conscious, without focal neurological deficits, and not overtly encephalopathic, at which point the leads are removed percutaneously without requiring further surgery.

Case study

A 32-year-old man with no neurologic history fell off a moving vehicle at high speed and struck his head on the pavement, and ultimately required surgical evacuation and decompression due to large epidural and subdural hematomas (figure 1A-D). He underwent emergent surgical evacuation with IEEG lead placement under the study protocol. A post-operative CT showed a decreased mass effect and adequate placement of the cortical electrodes over the frontal and parietal lobes (figure 1E-G).

Scalp and intracranial EEG were recorded for 94 hours starting shortly after completion of surgery. Scalp EEG initially demonstrated severe diffuse slowing (1-4 Hz) with overlying fast activity (12-20 Hz), consistent with an anesthetic effect, and a left hemispheric breach rhythm. After discontinuation of propofol on the third post-operative day, left hemispheric focal slowing persisted (figure 2).

From the onset of IEEG recording, the frontal electrode recorded successfully. The parietal electrode did not record (likely due to mechanical damage to its delicate leads) and its channels were hidden in the review montage. The frontal subdural electrode showed intermittent, waxing and waning, sharply-contoured rhythmic delta activity, as well as periodic sharp waves, consistent with lateralized periodic discharges. At times, the rhythmic delta activity evolved in frequency, location, and morphology, persisting for at least ten seconds, thus meeting the criteria for electrographic seizures used by Claassen et al. (2004) in their study of seizure detection by continuous EEG in critically ill patients. On scalp EEG, however, this ictal activity was not detected, although intermittent rhythmic or quasi-rhythmic delta slowing (1-2 Hz) without evolution was concurrently seen on occasion (figure 2). The rhythmic activity could also be triggered consistently with stimulation (e.g. nursing care), suggesting stimulus-induced rhythmic, periodic, or ictal discharges (SIRPIDs) (Hirsch et al., 2004). There were no convulsive seizures.

In response to the IEEG findings and given persistence of coma despite anesthesia wean, antiseizure treatment with levetiracetam was increased until the electrographic seizures identified on IEEG resolved at a daily total dose of 2,000 mg. The patient's level of consciousness improved shortly after the seizures were controlled on the fourth post-operative day, suggesting that the electrographic seizures may have contributed to a worsened encephalopathy. During the night and morning prior to his improvement in consciousness, the rhythmic delta scalp correlate and IEEG lateralized periodic discharges became less frequent.

The patient improved and was extubated on the fifth post-operative day. He was without a focal neurological deficit. Antiseizure medications were discontinued three months post-operatively without issue, over eight months of follow-up.

Discussion

IEEG findings for seizures and SIRPIDs were identified following extra-axial hematoma evacuation for this patient using our method of cortical surface monitoring and were not identified on simultaneous scalp EEG. The patient was comatose while these findings were present and he regained alertness shortly after seizures were successfully treated, and although these were temporally related events, the correlation is uncertain. This case illustrates that, after evacuation of an extra-axial post-traumatic hematoma, clinically relevant focal subclinical seizure activity may be missed on scalp EEG or have subtle intermittent non-specific correlate not meeting criteria for seizures.

Our patient did not experience any adverse events related to the study. We chose electrodes that are widely used in elective epilepsy monitoring surgery as “depth” electrodes, and are FDA approved for IEEG monitoring. They allow for percutaneous removal in stereotactic EEG, but are used strictly over the cortex in our study (without penetrating the parenchyma) and were chosen for their ability to be removed percutaneously at the bedside without subjecting the patient to an additional operation. Despite the established safety of intraparenchymal placement seen in prior TBI IEEG studies (Waziri et al., 2009) and elective epilepsy surgery, this protocol minimizes any potential risk of iatrogenic hemorrhage and trauma to the penumbra, watershed or eloquent cortical tissue most pertinent for monitoring to avoid further injury to the friable traumatically injured brain in our experimental study protocol.

Future research into the risk factors, diagnosis, and best treatment of subclinical seizures after traumatic brain injury is warranted. Continued development of consensus-based guidelines for monitoring and treatment in cases of reduced consciousness after TBI may be useful. It is unclear if routine placement of minimally invasive IEEG leads after surgery for TBI is warranted, and this may be a topic of future outcomes-based research. The development of more effective percutaneously removable monitoring electrodes, perhaps integrated into surgical drains, or intracranial pressure or tissue oxygenation monitors, will increase their utility in this cohort, and continued research into IEEG in trauma patients may enhance our understanding of both epileptogenesis and non-convulsive seizures in this population.

Supplementary data.

Summary didactic slides and a supplementary figure are available on the www.epilepticdisorders.com website.

Acknowledgements and disclosures

Funding was obtained for the study using institutional funding sources and does not pose any conflicts of interest relevant to this study.

None of the authors have any conflict of interest to declare.