Abstracts

Rationale: Electrical kindling in rats is a commonly used model for temporal lobe epilepsy and is highly predictive of human efficacy. However, the use of this model for identification of novel anticonvulsant drugs is limited by the amount of compound required, the cost, and the time- and labor- intensive nature of the model. Mice can be kindled via transcorneal stimulation; hence, this has been proposed as a cost-effective model for screening compounds for the treatment of partial epilepsy (Matagne and Klitgaard, Epil Res. 31; 59-71, 1998). In the present investigation, the corneal kindling model in mice was evaluated as a tool for screening known antiepileptic drugs (AEDs) including valproic acid (VPA), lamotrigine (LTG), phenytoin (PHT), carbamazepine (CBZ), and retigabine (RGB).Methods: Forty CF1 mice were kindled via twice daily corneal stimulation (3mA, 3 sec) until all mice displayed five consecutive stage five seizures (Racine RJ, Electroencephalogr Clin Neurophysiol. 32(3):281-94, 1972) according to the protocol established by Matagne and Klitgaard (1998). Following a 10-day stimulation-free period, groups of mice (n=8) received an i.p. injection of one of the test compounds. At the previously determined time to peak effect, motor function was evaluated with the rotorod test (Orlof et al., Proc. Soc. Exp. Biol. Med., 70:254-257, 1949), after which mice were immediately challenged with the kindling stimulus and their seizure score recorded. Animals displaying a seizure score of ≤ 3 were considered protected. The dose was increased or decreased between the limits of 0 and 100% protection until sufficient data had been obtained to calculate the median effective dose (i.e., ED50). A three-day washout period was allowed between each challenge dose. Results were compared to those obtained in the hippocampal kindled rat model of partial epilepsy. Results: With the exception of phenytoin, all of the AEDs tested displayed a dose-dependent protection against secondarily generalized seizures in the corneal kindled mouse. PHT was active at supratherapeutic and behaviorally toxic doses of 75 mg/kg. Overall, the rank order of potency was RGB≥LTG>CBZ>VPA. ED50’s and (95% confidence intervals) were calculated to be 7.3 (4.9-10.3), 9.9 (5.5-18.7), 33.9 (25.9-41), and 121 (84.7-169) mg/kg, respectively. Moreover, the relative order of potency compared favorably to results obtained previously in the hippocampal kindled rat.Conclusions: As suggested by Matagne and Klitgaard (1998), we found the corneal kindled mouse model to be a sensitive and efficient model for screening candidate compounds for the treatment of partial epilepsy. Results compare favorably and correlate very well with those obtained in an electrically kindled rat model of partial seizures. The mouse corneal model is superior in that it requires no surgery, requires significantly less test compound, and reduces the overall expense per animal while yielding comparable results. (Funding supported by: NO1-NS-4-2359 (HSW))