Abstract
Sleep‐related hypermotor epilepsy (SHE) is characterized by hyperkinetic seizures arising from sleep. Awake seizures occasionally occur and are associated with a worse prognosis, with important implications for driving and quality of life. We evaluated the clinical features and sleep/wakefulness distribution of seizures at onset and lifelong in a large cohort of clinical/confirmed SHE. Chi‐square test and a multivariate logistic regression model were used to identify predictors of awake seizures lifelong (primary endpoint). Positive and negative likelihood ratio (LR+, LR‐) were calculated. We included 165 patients (male/female: 105/60) with a 27.6‐year median follow‐up. Most (67.9%) presented with seizures exclusively from sleep; 32.1% presented with seizures both while asleep and while awake, or exclusively during wakefulness. Presentation with seizures in wakefulness shows a sensitivity of 62.5% and a specificity of 96.5% to predict the occurrence of awake seizures lifelong, with an LR + of 18 (95% confidence interval [CI] = 5.75‐55) and LR‐ of 0.39 (95% CI = 0.29‐0.52). On multivariate analysis, distribution of sleep/awake seizures at onset was confirmed as an independent risk factor of awake seizures lifelong (odds ratio = 56.7). Patients presenting with awake seizures have a 94% probability of awake seizures lifelong, whereas in those presenting with asleep seizures only, the percentage lowers to 27%. This aspect should be mentioned during physician‐to‐patient communication about prognosis.
OCT