Abstracts

Rationale: Surgical treatment remains the most efficient way of achieving seizure freedom for patients with temporal lobe epilepsy. However, there is a subgroup of patients with drug-refractory epilepsy that do not fulfill the candidacy criteria for resective surgery. Since its approval in the US in 2010, laser interstitial thermal therapy (LiTT) emerged as a minimally invasive option for surgical treatment of refractory epilepsy. There is a multitude of factors responsible for poor outcomes, involving different stages and levels of both the electro-clinico-anatomical assessment of the seizure onset zone. In this study, we focus exclusively on identifying electrocorticographic (ECoG) biomarkers that distinguish MTL patients with favorable outcome from those with poor outcome after hippocampal laser ablation. Methods: We searched EMU database for patients fulfilling the following criteria: 1) they had stereotactic electroencephalography (sEEG), 2) sEEG electrode coverage included at least one lead in the hippocampus 3) they were treated with LiTT in the hippocampus and 4) they had at least 12 months of post-operative follow-up. A board-certified epileptologist (N.Z.) and an epilepsy neurophysiologist (V.K.) reviewed all ictal iEEGs of the selected patients. Ictal iEEG signature evaluation was focused in the first 5 seconds after the seizure onset and the emergence of delta (1-3 Hz), theta (4-7Hz), low beta (8-13Hz), high beta (14-30Hz) and gamma (>30Hz) frequency bands was registered for each event. In addition, the site of most prominent ictal activity within ablation site (anterior vs posterior hippocampus) was noted for every patient. Good outcome was considered as Engel class 1. Results: We identified eight patients (mean age 38.1 ± 12, one male) that underwent LITT therapy of the hippocampus following intracranial EEG monitoring between 2011-2018. None of these patients had prior intracranial surgeries. Six patients lacked any structural abnormalities on MRI, and two were found to have hippocampal sclerosis. A total of 67 seizures were analyzed. The average number of electrodes implanted in the hippocampus (site of ablation) was 4.3. Ablation was performed on left hippocampus in three patients, and on the right side in remaining patients. Average follow up after surgery was 20.7 months, ranging from 14 to 27 months. Findings summarized in Table 1. Four patients had ictal onset only involving mesial electrodes with frequencies exceeding 14 Hz. Out of those four patients, three had predominant frequencies exceeding 30 Hz in addition to (14-30 Hz) activity. Interestingly, these were the only patients in our study with a good outcome. Ictal onset with frequencies between 14-30 Hz found both at ablation and non-ablation sites was seen in two subjects. Two patient were found to have ictal onset in non-ablation site exceeding 14 Hz. None of these patients were seizure free following laser ablation.There were two patients with ictal activity prominent over the posterior hippocampus. Both patients continued to have seizures following LiTT therapy. There were recognizable spectral features in patients who were seizure free. A single declining phase indicated that the frequency of ictal activity was the highest at ictal onset, with a decline in frequencies as seizure further evolved. This is in contrast to patients who weren’t seizure free where the most prominent frequency at ictal onset tends to increase in the initial few seconds following the onset (indicated by the arrows in Fig 1), before decreasing again creating a distinct “crescendo-decrescendo” pattern in the time-frequency plot. Conclusions: We identified biomarkers on intracranial EEG that distinguished patients based on their outcome following LiTT therapy. Presence of 30+Hz at onset seen exclusively in hippocampus, with no extramesial activity at onset predicted good outcome (Engel class I). Presence of ictal activity exceeding 14 Hz in extramesial structures regardless of hippocampal ictal activity at onset, and predominant ictal activity in posterior hippocampus were both predictors of poor outcome. Funding: No funding