Characterizing Depth Electrode Coverage in Stereoelectroencephalography on Seizure Onset Zone Localization and Seizure Outcomes

Abstract

Introduction: The number of intracranial depth electrodes implanted in stereoelectroencephalography (SEEG) investigations is primarily driven by the preimplantation hypothesis about SOZ location. Targeting is not standardized and highly variable between centers. Whether some of these electrodes may prove redundant, or target low-yield areas too frequently, is uncertain. Methods: We identified a retrospective multi-institutional cohort implanted with depth electrodes for iEEG monitoring between 2003 and 2022. We collected preoperative clinical features and iEEG investigation parameters, including the number of depth electrodes and contacts implanted. We built a propensity-matched cohort with respect to these covariates and evaluated outcomes, which included (1) the likelihood of SOZ localization, (2) complications, and (3) seizure-free outcomes as a function of electrode contact coverage. In addition, we aimed to identify brain regions commonly explored in conjunction with each other and identify the likelihood of a region being implicated in initial electroclinical seizure onset. Results: One hundred and sixty-seven patients were followed for a median of 3.8 (range 2, 18) years after SEEG. Propensity-matched cohorts demonstrated that a higher number of implanted contacts were associated with a greater likelihood of proceeding to treatment, but were not associated with SOZ localization, seizure freedom (Engel I), favorable seizure outcomes (Engel I/II), or complications, per Bayes factor analysis. Lateral orbitofrontal, supramarginal, posterior cingulate, inferior parietal, and inferior temporal areas were least likely to be implicated in initial electrographic onset, whereas hippocampus, caudal middle frontal, pericalcarine, and parahippocampal areas were most likely when controlling for electrode coverage. Conclusions: SEEG effectively localizes the SOZ in both lesional and non-lesional etiologies, and clinicians are generally optimizing the electrode coverage for hypothetical SOZ localization, leading to further therapeutic surgeries that may confer seizure freedom. Nevertheless, several areas are possibly being explored despite low likelihood (<2.5%) of participation within the SOZ.