In general, substituting one antiepileptic drug for an other is a relatively slow process that is carried out over several weeks or months. The main reasons for this time frame are the fear of seizure exacerbation resulting from withdrawal of the first drug while the introduced drug is not (or not yet) protective, as well as the fear of side effects associated with rapid introduction of the new drug. However, there are reasons, besides avoidance of unnecessary delays, why it may be desirable to substitute antiepileptic drugs rapidly in some circumstances. One such circumstance is an acute reaction to a drug, such as a severe idiosyncratic adverse effect. In addition, in a patient who is already in the hospital for a seizure exacerbation, it might be desirable to complete a substitution rapidly before discharge and send the patient home on a new drug regimen at therapeutic levels. An additional reason might be the establishment of pharmacoresistance in a potential candidate for epilepsy surgery. The process of establishing pharmacoresistance should be as short as possible for several reasons. First, if the patient’s condition is not truly medically intractable, the successful therapeutic step will occur sooner. Second, in a patient who becomes seizure free after surgery, the overall long-term outcome is likely to be better if the surgery is performed earlier. Finally, a patient might already have been referred to a specialized center for epilepsy surgery but not yet meet acceptable criteria of medical intractability, and one additional adequate drug trial needs to be completed as rapidly as pos-sible.

Some clinical experience has been accumulated with the mutual substitutions of several established antiepileptic drugs, mostly in adult patients with focal onset seizures. In many instances, the substitution can be completed within about 5 days. This chapter describes the clinical experience with rapid introduction of carbamazepine, phenytoin, phenobarbital, and primidone, along with the concomitant rapid discontinuation of another drug within this group. Although the rapid replacement of valproate by carbamazepine, phenytoin, phenobarbital, or primidone is easy, the experience with rapid changes from one of these drugs to valproate monotherapy has not been favorable in patients with focal-onset seizures. Less experience has been acquired with the corresponding rapid changes in children, and no systematic observations have been reported about rapid changes involving the newer antiepileptic drugs.

When one antiepileptic drug is replaced by another within a short time, two issues require particular attention: (a) A therapeutic level of the drug to be introduced must be achieved as rapidly as possible and must then be maintained so as to allow rapid discontinuation of the previous drug, and (b) it must be anticipated that any pharmacokinetic interaction between the two drugs will be reversed after discontinuation of one drug. Therefore, rapid drug changes require an understanding of how rapidly a given drug can be introduced, as well as a qualitative and quantitative knowledge of the effect the drug being taken by the patient will have on the pharmacokinetics of the drug to be introduced. These rapid drug changes also require a good level of experience with antiepileptic therapy and appropriate supervision of the patient during the process.

When carbamazepine is to be introduced rapidly to replace another antiepileptic drug, three points need to be considered: (a) the fact that levels of carbamazepine in the usual therapeutic range are not tolerated during the first days of treatment, (b) the autoinduction of carbamazepine metabolism, and (c) the potential for induction of carbamazepine metabolism by other drugs. For example, if carbamazepine is to be introduced to replace phenytoin, which markedly accelerates the elimination of carbamazepine, the initial dose requirement of carbamazepine will be higher. However, after phenytoin has been discontinued, the dose requirements of carbamazepine decrease within 2 to 4 weeks, and toxic carbamazepine levels will occur if the dose is not reduced. The effects of existing antiepileptic drug regimens on carbamazepine dosage requirements are summarized in Table 1. The processes of heteroinduction and autoinduction of carbamazepine metabolism are all reversible, and the half-life of the induction and deinduction processes is about 4 to 6 days.^8,9 The initial dose for the introduction of carbamazepine is determined on the basis of the preexisting drug regimen.