Rationale:
Sudden unexpected death in epilepsy (SUDEP) is a leading cause of mortality in people with epilepsy, with risk up to 27 times higher in young adults than in the general epilepsy population. Its unpredictable nature underscores the need for diverse translationally relevant models. Current rodent SUDEP models reflect genetic contributors (e.g., SCN1A or KCNQ1 knockout mice), but do not well reflect multifactorial pathophysiology or age of peak incidence (20-45 years old). Further, SUDEP in late-onset epilepsy (individuals over age 65) may be under reported and exacerbated by comorbid conditions, such as Alzheimer’s disease (AD; Zawar et al, Annals Neurol 2025). We consistently observe excessive premature mortality in both young and aged corneal kindled APP/PS1 mice (Del Pozo et al, Prog Neurobiol 2024; Knox et al, J Alz Dis 2025), a model of AD. We thus hypothesized that corneal kindled APP/PS1 mice may represent a novel SUDEP model and aimed to prospectively document spontaneous seizure-induced death using video-EEG (vEEG).
Methods:
Male and female APP/PS1 and transgene-negative (Tg-) mice were implanted with cortical vEEG electrodes at 4-5 weeks-old (7 males, 5 females). After one week recovery, mice were corneal kindled (3 sec, 60 Hz 1.6-2.0 mA; n=7) twice daily until fully kindled (five consecutive Racine stage 5 seizures; FK). Sham-kindled controls (n=5) received handling without corneal stimulation. vEEG was recorded during the dark cycle (8:00 pm-6:00 am). Simple linear regression assessed group differences in cumulative seizure burden, frequency and duration, as well as correlation between pre- and post-FK seizure metrics. Two-way ANOVA evaluated genotype and timepoint effects on average seizure burden, frequency, and daily seizure duration. All spontaneous recurrent seizure (SRS) metrics were obtained from a consecutive 10-day vEEG observation period-3 days before and 7 days after FK.
Results:
One kindled female APP/PS1 mouse died suddenly following a kindled stage 5 seizure just prior to meeting FK criterion; no other deaths occurred. No SRS events were observed in sham-kindled mice regardless of genotype across the 10-day observation period. APP/PS1 mice showed a significantly steeper increase in cumulative SRS burden and frequency versus Tg- mice (Fig 1A-B, non-overlapping 95% confidence intervals), while SRS duration showed a nonsignificant trend (Fig 1C). Mixed-effects modeling did not detect significant genotype × timepoint effects in daily SRS metrics (Fig 1D-F). Notably, pre-FK average SRS burden strongly predicted post-FK SRS burden across genotypes (Fig 2, r² = 0.99, p < 0.0001), whereas evoked seizure burden during kindling was not predictive of later SRS activity (Fig 2, r² = 0.04, p = 0.7).
Conclusions:
These data demonstrate the feasibility of implementing corneal kindled APP/PS1 mice as a novel multifactorial and inducible SUDEP model evoked in adult mice. The observed SUDEP-like mortality and progressive SRS phenotype, along with the predictive correlation in SRS burden, are being further investigated in additional cohorts with continuous vEEG and longitudinal survival analyses
Funding:
2023 American Epilepsy Society Postdoctoral Fellowship and NIA R01AG067788