Abstracts

Rationale: This study examines sleep-seizure electroencephalography (EEG) associations as biomarkers to track severity of upcoming seizures including monitoring for death risk using a new mouse model of adult-onset, spontaneous epilepsy. The mice overexpress brain derived neurotrophic factor (BDNF) in the forebrain under the CAM kinase II alpha promoter (TgBDNF). TgBDNF mice develop convulsive seizures at ~4 months of age which worsen with each seizure episode. Seizures are elicited via tail lifts and cage agitation. Seizures become more severe with each additional episode as indicated by increased duration of postictal cortical activity suppression (i.e., EEG flat or EEG waves in delta frequency) associated with loss of posture/consciousness. In more advanced stages of the epilepsy, the mice expire following long period of postictal generalized EEG suppression culminating in cardiorespiratory arrest.

Methods: We surgically implanted subdural cortical recording electrodes on the TgBDNF mice skulls prior to the first seizure episode (3- channel EEG system, Pinnacle Technology, KS) to record seizure and sleep EEG for 7 weeks. The recordings targeted time period that spans moderate to severe phase of epilepsy. We used age- and litter-matched wild-type mice as controls. We hypothesized that as seizure severity progresses, detectable alterations in sleep structure will be identified that indicate life threatening potential of upcoming seizures. Seizures were dislodged by tail lifting/cage shaking induction at each episode. Twenty-four hours following seizure induction at each week, mice were connected to EEG apparatus for sleep recording during inactive phase. We analyzed early (first 1 hr) and late (last 1 hr) sleep separately to define the dynamic changes in rapid eye movement (REM) and slow-wave sleep (SWS) phases.

Results: Data were assessed by simple regression analyses between multiple seizure and sleep parameters. A shift in increased duration of SWS in late compared to early sleep was associated with a remarkable increase in duration of postictal generalized cortical suppression. Moreover, shorter delay to onset of SWS in late phase (last 1 hr) of sleep was associated with increased total seizure duration. Number of REM bout counts in early sleep was negatively associated with duration of postictal generalized cortical suppression (delta frequency), ictal duration and ictal spectral power. In addition, number of REM bout counts in the late phase of sleep was negatively associated with length of postictal generalized cortical suppression (delta frequency) and cumulative duration of spiking wave discharges observed in seizure recovery accompanied by behavioral freezing.

Conclusions: Our data indicate that dynamic alterations occur in sleep-seizure EEG measures indicative of progression and severity of epilepsy. Identifying and monitoring critical sleep-seizure EEG associations across disease progression may have predictive power for loss of consciousness and death risk in upcoming seizure episodes and, in turn may inform impact of such episodes on sleep architecture.

Funding: Please list any funding that was received in support of this abstract.: NIH/NINDS grant NS115049 funded to Dr. Isgor.