F126. Focal seizures detected with intracranial electrodes after traumatic hemorrhage may appear as focal slowing on scalp EEG and be missed

Abstract

Traumatic intracranial hemorrhage (TIH) predisposes patients to clinical and subclinical seizures, which may contribute to worsened neurological outcomes. However, experimental data suggests that scalp EEG may be insensitive for detecting some nonconvulsive seizures, especially those originating from a deeper focus. We report the simultaneous scalp and intracranial EEG findings of a 32-year-old man following evacuation of a traumatic left epidural and subdural hematoma. The patient was enrolled in an exploratory study to obtain intracranial electroencephalography (IEEG) in adults undergoing emergent post-traumatic hemicraniectomy for TIH. Following hemicraniectomy and opening of the dura mater, a 4-contact depth electrode was placed in the subdural space overlying the ipsilateral frontal cortex adjacent to the evacuated hematoma. Intraparenchymal placement was avoided to prevent insertional damage in a friable brain. The leads were tunneled through the skin adjacent to the incision, and secured externally with a silk suture, under a single layer of oxidized regenerated cellulose sheet. Positioning of the electrodes was confirmed using routine postoperative noncontrast head CT. Scalp and intracranial EEG were recorded for 94 h. Scalp EEG demonstrated severe diffuse slowing (1–4 Hz) with overlying fast activity (12–20 Hz), consistent with anesthetic effect, and a left hemispheric breech rhythm. Focal slowing persisted over the left hemisphere despite anesthetic wean. Throughout the recording, IEEG showed intermittent rhythmic or sharp waves consistent with lateralized periodic discharges (1–2 Hz). At times these evolved sufficiently in frequency, morphology and amplitude to meet criteria for nonconvulsive seizures, and could be triggered consistently with stimulation. On scalp EEG however, these periodic discharges and seizures were not detected, although rhythmic or semi-rhythmic focal slowing (1–2 Hz) was concurrently seen on occasion. In response to the IEEG findings and given the persistence of clinical encephalopathy despite weaning of anesthesia, antiseizure treatment was increased until the seizures identified on IEEG resolved. The patient’s level of consciousness subsequently improved, suggesting that the nonconvulsive seizures may have contributed to a worsened encephalopathy. Following the recording, a rhythmicity spectrogram trends panel was developed for the intracranial electrode and consistently demonstrated low-frequency bands during seizure activity that could be used for rapid seizure identification. Following TIH, clinically relevant focal seizure activity may be entirely missed on scalp EEG or have subtle intermittent nonspecific correlate such as focal slowing. More research is needed to determine the frequency of this phenomenon and determine whether treatment of such seizures improves patient outcomes.