Abstracts

Rationale: Catamenial patterns of seizure exacerbation, in which seizure frequency and/or severity increase at certain points of the menstrual cycle, typically around ovulation or menstruation, are commonly observed in women with epilepsy. Current rodent models of catamenial epilepsy primarily focus on neurosteroid withdrawal and other treatments designed to mimic hormonal fluctuations associated with perimenstrual seizures. Less is known about whether seizure burden displays endogenous shifts across estrous cycle stages in rodent models of epilepsy. In addition, models of periovulatory catamenial epilepsy are lacking. We recently reported changes in circulating levels of the sex steroids estradiol and progesterone in female mice in the intrahippocampal kainate (IHKA) model of temporal lobe epilepsy (TLE) (1). In addition, these effects were different on diestrus compared with estrus, suggesting that there may be a dynamic relationship between seizure activity, steroid hormone levels, and cycle stage. Therefore, we examined whether seizure burden changes across the female reproductive estrous cycle in this model of TLE. Methods: Adult female mice on the C57BL/6J background were unilaterally injected with kainate (KA, 50 nl of 20 mM) and implanted with two twisted bipolar stainless-steel depth EEG electrodes in the dorsal hippocampus. 2 months after KA injection, continuous 24/7 local field potential (LFP) EEG and simultaneous video recordings were collected over three weeks, during which estrous cycle stage was assessed daily. Seizures were captured from EEG recordings using a custom machine learning model for seizure detection, with a 95% accuracy rate. EEG recordings were quantified for electrographic seizure frequency and duration, with minimum seizure duration set at 5 s. Power spectrum analysis across the 0.1-100 Hz frequency range was used to calculate total EEG power of seizure activity. Data collected on proestrus and estrus days were combined (“pro/estrus”) and compared with data collected on diestrus within the same mice using paired statistics. Results: Initial analyses were focused on 9 mice that displayed high seizure frequency (at least 20 seizures per hour). In these mice, seizure frequency was higher on pro/estrus (29.0 ± 4.3 seizures per hour) compared with diestrus (22.1 + 3.8, p = 0.005). Seizure duration was also longer on pro/estrus (17 + 0.9 s) compared with diestrus (16 + 1.2 s, p = 0.005). In addition, total seizure EEG power was 25% higher on pro/estrus compared with diestrus (p = 0.009). Conclusions: These initial results suggest that seizure burden fluctuates with estrous cycle stage in the IHKA mouse model of TLE, at least in mice that show a high seizure burden overall. Ongoing analyses will evaluate whether similar patterns are also seen in mice that display less severe seizure burdens. In the present dataset, proestrus and estrus stages, during which ovulation occurs and sexual receptivity is high, were associated with higher seizure burden than the diestrus stage. Therefore, the IHKA mouse model may be useful in modeling periovulatory patterns of catamenial seizure exacerbation. References: Li et al., 2018 eNeuro 5:e0273-18.2018 Funding: R01 NS105825, R03 NS103029