OBJECTIVE: To describe the primary types of antiepileptic drug (AED) interactions by using a mechanistic approach. DATA SOURCES: A literature search was performed using MEDLINE and bibliographies of recent review articles and published abstracts. DISCUSSION: AEDs are associated with a wide range of drug interactions, including hepatic enzyme induction and inhibition and protein-binding displacement. Hepatic induction by AEDs affects the metabolism of a limited number of drugs with low therapeutic indices. Anticipation of induction interactions and careful clinical monitoring may alleviate potential problems. Most commonly used AEDs are eliminated through hepatic metabolism catalyzed by the cytochrome P450 (CYP) and uridine diphosphate glucuronosyltransferase (UGT) enzymes. Phenytoin, phenobarbital, and carbamazepine induce CYP and UGT enzymes. Lamotrigine is a weak inducer of UGT. Valproate is a broad-spectrum inhibitor of UGT enzymes, epoxide hydrolase, and CYP2C enzymes. Felbamate induces CYP3A4, but inhibits CYP2C19 substrates. Topiramate inhibits only CYP2C19 substrates. Ethosuximide, gabapentin, tiagabine, and vigabatrin are neither inducers nor inhibitors of drug metabolism. Hepatic enzyme inhibition usually occurs because of competition at the enzyme site. Knowledge of the specific metabolic enzymes involved in the metabolism of AEDs allows clinicians to predict potential interactions. CONCLUSIONS: By understanding the mechanisms of drug interactions, the pharmacist can play a key role in patient care by anticipating and preventing AED drug interactions

Coumarin derivatives combine 3 unfavorable properties which make them prone to potentially life threatening drug-drug interactions: (i) high protein binding; (ii) cytochrome P450 dependent metabolism; and (iii) a narrow therapeutic range. An entire list of drugs which are supposed to interact with coumarins (mostly with warfarin) comprises about 250 different compounds. Noteworthy are the interactions with cardiovascular or antilipidaemic drugs which are often coadministered with coumarins: amiodarone, propafenone and fibrates. Cardiovascular drugs which are obviously devoid or proven to be devoid of an interaction are angiotensin converting enzyme (ACE) inhibitors, calcium antagonists, beta-blockers and cardiac glycosides. There are several other drugs which enhance the hypoprothrombinaemic response to coumarins by various mechanisms: inhibitors of the elimination of the eutomer S-(-)-warfarin (e.g. miconazole, phenylbutazone), combined with protein binding displacement (e.g., sulfinpyrazone, phenylbutazone), synergistic hypoprothrombinaemia (e.g. cefazoline). Furthermore, bleeding complications may occur with drugs affecting platelet function [aspirin (acetylsalicylic acid) and several nonsteroidal anti- inflammatories (NSAIDs)]. Strong inducers of coumarin metabolism are rifampicin (rifampin) and carbamazepine. Biphasic interactions may occur where a drug first enhances the hypoprothrombinaemic response to a coumarin but has a sustained inducing effect on coumarin metabolism (e.g. phenytoin or sulfinpyrazone). The complex response of coumarins to concomitant drug therapy makes it difficult to predict the occurrence and degree of a deterioration of anticoagulant control in individual patients. For clinical practice, it seems advisable that one should monitor for changes in prothrombin time when adding or deleting any newly approved drug or any drug suspected (e.g. on the basis of this review) to cause an interaction to patients on coumarin therapy. The onset of the adverse prothrombin time response might be from between 1 to 2 days up to 3 weeks (in case of phenprocoumon) after starting a concomitant drug regimen. With amiodarone, an adverse prothrombin time response might occur up to 2 months after initiating therapy. For heparins, only a drug interaction with aspirin or nitroglycerin seems clinically relevant due to the possibility of coadministration during acute cardiac events. Both drugs are shown to enhance the activated partial thromboplastin time response to heparin

Several drugs have recently been shown to increase the activity of drug-metabolizing enzymes in the liver microsomes. Phenobarbital is particularly able to stimulate an increase in the synthesis of a variety of drug-metabolizing enzymes. The activity of liver microsomal enzymes that metabolize diphenylhydantoin, griseofulvin, zoxazolamine, strychnine and coumarins has been shown to be increased by phenobarbital.The effect of barbiturates on the activity of coumarins has been reported in animals, and in single-dose acute experiments in man. A single patient, receiving bishydroxycoumarin (Dicumarol) over a prolonged period, has been reported to have a substantial decrease in plasma anticoagulant level and activity with the administration of phenobarbital. This study was undertaken to evaluate the effect of phenobarbital on the anticoagulant requirements of a group of patients receiving prolonged anticoagulant therapy.

Ten male volunteers received phenobarbital, glutethimide, chloral betaine and a placebo at bedtime for 3 week periods. Plasma warfarin levels and prothombin times were determined after a single oral dose of warfarin prior to and immediately following placebo or hypnotic drug treatment. Glutethimide and phenobarbital lowered the plasma warfarin concentration and reduced the half-life of warfarin by nearly 50 per cent. Chloral betaine had less effect but differed significantly from control values. Phenobarbital and glutethimide interfered with the hypoprothrombinemic effect of warfarin; chloral betaine and placebo did not. The findings suggested increased metabolic inactivation of warfarin by induction of liver microsomal enzymes.