Abstracts

RATIONALE: A few anticonvulsants have been reported to have persistent long-term effects despite their discontinuation. This effect goes beyond their normal pharmacokinetics. This has not been shown clinically for some of the newer anticonvulsants. We report on cases of persistent long-term anticonvulsant effect of levetiracetam following its cessation.
METHODS: This was a retrospective review of patients from Jan 2001 to March 2002 in the Epilepsy Monitoring Unit at the University of Colorado Hospital. The anticonvulsants of all the patients were stopped on the first day of hospitalization. Only the patients with temporal lobe epilepsy were included in the study, and patients who were on valproic acid, barbiturate or benzodiazepine were excluded. All the patients on levetiracetam monotherapy or polytherapy with other anticonvulsants were selected. These patients were then matched for age, sex and frequency of seizures with patients on other single or multiple anticonvulsants. The hospital day to onset of their first seizure was noted. If a patient did not have a seizure during their five-day hospitalization then they were assigned a value of day six to onset of seizure. The two groups were then compared using a two tail, independent variable T-test.
RESULTS: The two groups were matched for age, sex and frequency of seizures. In addition to levetiracetam, other anticonvulsants that patients were on included phenytoin, carbamazepine, topiramate, lamotrigine and zonisamide. There were equal number of patients on monotherapy and polytherapy in the two groups. The levetiracetam group (n = 7) had a mean of 4 days to onset of first seizure and a median of 6 days (range 2-6 days). The non-levetiracetam group (n = 7) had a mean of 2.1 days with a median of 2 days to onset of first seizure (range 1-3 days). Using a two tail T-test, this resulted in a p-value of 0.08. There were 3 patients in the levetiracetam group who had no seizures during the hospitalization, and all the patients in the non-levetiracetam group had seizures
CONCLUSIONS: Following the discontinuation of anticonvulsants, the levetiracetam group had a later day to onset of their first seizure than the non-levetiracetam group with 3 patients not having any seizures at all. Although the p-value of 0.08 was not statistically significant, it is trending towards significance. Also, the patients without seizures were assigned the most conservative value of day 6 to onset of seizure. This only occurred in patients in the levetiracetam group, and likely underestimated the actual day to onset of first seizure, thereby decreasing the significance of the difference between the two groups. Despite stopping the levetiracetam patients tended to remain seizure free for several days beyond. The exact mechanisms of levetiracetam are just beginning to be elucidated, but this study shows that the efficacy of levetiracetam appears to go beyond its pharmacokinetic effects.