Abstracts

RATIONALE: Accumulating data suggests that status epilepticus (SE) causes neuronal loss in affected brain areas and may be a key determining factor in subsequent epileptogenesis. The development and targeting of effective therapeutic intervention would be greatly facilitated by a biomarker providing a (preferably peripheral) readout of neuronal degeneration and predicting the likelihood of progression to epilepsy. Serum neuron specific enolase (sNSE) may be such a marker. Levels are reported to rise following an episode of SE in humans and peak concentrations to correlate with long-term outcome (DeGiorgio et al., Neurology, 1995, 45:1134-1137). Although elevated sNSE has been observed following pilocarpine-induced SE in rats (Sankar et al., Epilepsy Res., 1997, 28:129-136), data from animal models of SE is generally limited. These experiments were designed to evaluate sNSE levels in a new mouse model of kainic acid (KA)-induced SE.
METHODS: Conscious male CD-1 mice (20-30 g, Charles River) were injected intracerebroventricularly (i.c.v.) with saline or 0.19 nmol KA in a dose volume of 5 [mu]l and seizure-related behaviour categorized according to the following grades: (1) staring, head-nodding, Straub tail, hunched posture; (2) forelimb clonus, salivation; (3) jumping, rearing, unsteady posture; (4) tonic clonic seizure, vocalisation, rearing and falling, barrel rolling, circling. SE was defined as the constant repetition of seizures for at least 10 min. Mice were culled at 24, 48, or 96 h following injection and blood taken by cardiac puncture. Serum NSE was quantified by radioimmunoassay (Pharmacia, Sweden).
RESULTS: None of the saline-treated animals demonstrated any seizure-related behaviour (n = 8). KA administration induced seizure activity in 20 out of 22 mice and SE was observed in 15 mice (68%). Median (and interquartile range) sNSE concentration in control animals 24 h after i.c.v. saline administration was 15.3 (3.4) ng.ml^-1 (n = 7). There was no significant difference between sNSE concentrations in groups of animals exhibiting SE 24, 48, or 96 h post-KA injection or in controls (Kruskal-Wallis all samples comparison). However, analysis of the data according to seizure severity revealed a weak correlation between sNSE levels and maximum seizure grade (R = 0.38, p = 0.048, n = 28). Median (and interquartile range) sNSE level in all animals exhibiting grade 4 seizures was 21.0 (5.2) ng.ml^-1 (n = 5), and that of seizure-free animals was 15.7 (2.8) ng.ml^-1 (n = 8) (p = 0.06, Mann Whitney U test).
CONCLUSIONS: Seizure severity and sNSE levels in KA-treated mice were weakly correlated in this preliminary study supporting further characterization of this mouse model of SE and the utility of this biomarker.