Subclinical Seizures on Stereotactic EEG: Characteristics and Prognostic Value
Abstract number :
2.027
Submission category :
3. Neurophysiology / 3C. Other Clinical EEG
Year :
2021
Submission ID :
1825503
Source :
www.aesnet.org
Presentation date :
12/5/2021 12:00:00 PM
Published date :
Nov 22, 2021, 06:44 AM
Authors :
Benjamin Cox, MD - Mayo Clinic Rochester; Jamal Khattak, MD - Neurology - Henry Ford Hospital; Keith Starnes, MD - Pediatric Neurology - Mayo Clinic Rochester; Benjamin Brinkmann, PhD - Neurology - Mayo Clinic Rochester; Jamie Van Gompel, MD - Neurosurgery - Mayo Clinic Rochester; Kai Miller, MD, PhD - Neurosurgery - Mayo Clinic Rochester; W. Richard Marsh, MD - Neurosurgery - Mayo Clinic Rochester; Lily Wong-Kisiel, MD - Pediatric Neurology - Mayo Clinic Rochester; Elson So, MD - Neurology - Mayo Clinic Rochester; David Burkholder, MD - Neurology - Mayo Clinic Rochester
Rationale: Although stereotactic EEG (sEEG) has become a widely used intracranial EEG technique, reports of the significance of sEEG recorded subclinical seizures (SCS) and surgical outcome are limited.
Methods: We reviewed all patients who had undergone intracranial monitoring with sEEG at our institution through December 2020 for the purpose of localizing drug resistant epilepsy (n=169). For each patient the presence of SCS was noted, as was their occurrence in sleep, wakefulness, or both. The concordance between subclinical and clinical seizure onset zone (SOZ) was classified as complete (same electrodes involved), partial (the subclinical SOZ was within clinical SOZ), overlapping (some, but not all of the subclinical SOZ electrodes were within clinical SOZ), or no concordance (completely separate electrodes involved). Surgeries included resection, laser interstitial thermal ablation (LITT) and focal disconnection. The location of subclinical SOZ in relation to the surgical site was recorded as complete, partial, or not treated. Pathology was recorded for patients with resection. Outcomes were classified using the ILAE system, with class 1 and 2 considered “excellent” outcomes.
Results: SCS were observed during sEEG monitoring in 84 of 169 patients (50%, Table 1). There was no difference in the prevalence of SCS based on imaging abnormalities, temporal vs extratemporal SOZ, SCS occurrence in sleep, number of electrodes, pathology, or lobar location. SCS were more common in female patients than male (62% vs 40%, p=0.0054). SCS had complete concordance with clinical SOZ in 40% of patients, partial concordance in 29%, overlapping in 19%, and discordant in 12%. Eighty-one patients had surgery, 42 of whom had SCS (resection 31, LITT 11, disconnection 2). There was no difference in excellent outcome based on the presence of SCS or on the concordance with clinical SOZ, however, there were improved outcomes in patients with destruction of the subclinical SOZ compared with patients who did not have SCS treated (HR 2.90, p=0.034, 95% CI 1.05-8.03)(Fig. 1).
Conclusions: These findings demonstrate that SCS are commonly seen on sEEG and tend to colocalize with the clinical SOZ in most patients. Discordance with clinical SOZ is not necessarily a predictor of poor surgical outcome; rather, resection of the subclinical SOZ seems to correlate with excellent surgical outcome. For unclear reasons, subclinical seizures occur more commonly in females than males.
Funding: Please list any funding that was received in support of this abstract.: None.
Neurophysiology