Prolonged ambulatory EEG findings and risk of seizure recurrence in adults with generalized tonic-clonic seizures alone: A multicentre study

Epilepsy with generalized tonic-clonic seizures alone (GTCA) is a distinct subtype of idiopathic generalized epilepsy (IGE), yet its prognostic markers remain poorly defined. This study aimed to evaluate whether prolonged ambulatory EEG (paEEG), when integrated with clinical variables, improves the identification of patients at increased risk of seizure recurrence. We retrospectively analyzed 32 paEEG recordings from 27 patients with GTCA, classifying them in two groups according to seizure recurrence within 1 year after paEEG (seizure recurrence vs no seizure recurrence). We included only paEEG recordings of patients who had maintained the same antiseizure medication (ASM) regimen (dosage and type) throughout the whole period from the EEG recording up to the outcome assessment (defined as seizure recurrence within 1 year or 1-year seizure freedom). EEG features and interictal epileptiform discharges (EDs), were quantified in both wakefulness and sleep. Predictors were evaluated through univariable analysis and multivariable logistic regression. Most paEEG parameters examined-including total frequency, density, and duration of EDs did not differ significantly between recordings with and without subsequent seizure recurrence during either wakefulness or sleep. In contrast, several wake-onset markers (ED frequency, spike density, and duration) were significantly different between the two groups (p < 0.05). In multivariable analysis, the use of valproate showed a protective effect, while longer ED duration at wake onset increased the likelihood of relapse, both remaining independent predictors. The combined model achieved good apparent discriminative performance (AUC = 0.879). Sensitivity analyses at the patient level and a GEE model accounting for within-subject correlation yielded consistent results, supporting the robustness of wake-onset EDs as predictors. These findings support the integration of paEEG and detailed sleep-wake transitional analysis into prognostic assessment in GTCA, although further validation in larger cohorts is required.