SEIZURES ASSOCIATED WITH ANTIDEPRESSANTS - A REVIEW

Background: Seizures are uncommon, but serious, adverse effects of ant idepressant drugs. A better understanding of drug-related seizure risk , its predictors, and its neurophysiologic basis might help clinicians avoid this adverse event. A better understanding of the factors invol ved in the determination of seizure risk would be helpful for interpre tation of seizure rates reported. Method. The authors review case repo rts, series of cases, and information from clinical trials of antidepr essants to determine antidepressant-related seizure risk. Predisposing factors are identified. Effects of dose, blood levels, and duration o f treatment on seizure risk are examined. Electrophysiologic and in vi tro models of drug-related seizure induction are discussed. Results: A significant proportion of drug-related seizures occurs in individuals with an identifiable predisposition, such as previous seizures, sedat ive or alcohol withdrawal, and multiple concomitant medications. Seizu re risk for most antidepressants increases with dose (or blood level), and comparisons between drugs should consider seizure rates at the ef fective dose (or blood level) for each drug. For imipramine, the most frequently studied tricyclic, the literature indicates a seizure rate between 0.3% and 0.6% at effective doses. In unselected patients and a t higher doses, these rates may be higher. Fluoxetine, sertraline, flu voxamine, trazodone, nomifensine, and the monoamine oxidase inhibitors have a lower seizure risk. Estimates for recently marketed antidepres sants with intermediate seizure risk are complicated by the fact that effective doses and blood levels are not well established. Conclusion: Assessment of seizure risk in individuals involves consideration of p redisposing factors, the antidepressant selected, and the bioavailabil ity of the drug. Future studies of seizure risk would benefit from the use of specified criteria for determination of probable seizure event s, a priori definition of predisposing exclusions, samples sufficientl y large to provide adequate power, blood level monitoring, and inclusi on of duration of drug treatment in the calculation of risk.