Lack of spontaneous typical seizures during intracranial monitoring with stereo-electroencephalography

Objective: In the presurgical evaluation of patients with drug-resistant epilepsy (DRE), occasionally, patients do not experience spontaneous typical seizures (STS) during a stereo-electroencephalography (SEEG) study, which limits its effectiveness. We sought to identify risk factors for patients who did not have STS during SEEG and to analyze the clinical outcomes for this particular set of patients.

Methods: We conducted a retrospective analysis of all patients with DRE who underwent depth electrode implantation and SEEG recordings between January 2013 and December 2018.

Results: SEEG was performed in 155 cases during this period. 11 (7.2%) did not experience any clinical seizures (non-STS group), while 143 experienced at least one patient-typical seizure during admission (STS group). No significant differences were found between STS and non-STS groups in terms of patient demographics, lesional/non-lesional epilepsy ratio, pre-SEEG seizure frequency, number of ASMs used, electrographic seizures or postoperative seizure outcome in those who underwent resective surgery. Statistically significant differences were found in the average number of electrodes implanted (7.0 in the non-STS group vs. 10.2 in STS), days in Epilepsy Monitoring Unit (21.8 vs. 12.8 days) and the number of cases that underwent resective surgery following SEEG (27.3% vs. 60.8%), respectively. The three non-STS patients (30%) who underwent surgery, all had their typical seizures triggered during ECS studies. Three cases were found to have psychogenic non-epileptic seizures. None of the patients in the non-STS group were offered neurostimulation devices. Five of the non-STS patients experienced transient seizure improvement following SEEG.

Significance: We were unable to identify any factors that predicted lack of seizures during SEEG recordings. Resective surgery was only offered in cases where ECS studies replicated patient-typical seizures. Larger datasets are required to be able to identify factors that predict which patients will fail to develop seizures during SEEG.