Summary: Generally, the efficacy of an anticonvulsant drug is a function of the intensity of the seizure process and the drug concentration in plasma and tissues. Because of the variability of the seizure process among individuals, a universally effective drug plasma level is not to be expected. Rather a 'desirable' plasma level range for each drug can be defined, the lower limits being those levels above which the majority of patients are well controlled, the upper limits being determined by the onset of drug intoxication. With the onset of administration of anticonvulsants in average doses, the maximum plasma levels and effect are reached only after several days or with some drugs, after several weeks of therapy. The lag time can be reduced with loading doses in the case of phenytoin or phenobarbitone (2 to 3 times the usual dose). For maintenance therapy phenytoin and phenobarbitone may be given in a single daily dose without marked fluctuations in the plasma levels. Some variations in the plasma levels relative to particular doses do occur, but in the majority of patients predictable ranges of plasma levels relative to doses can be defined. Marked deviations from these expected ranges are however, not infrequent. Low levels are most frequently caused by patients' failure to take the drug regularly. Rare causes for low plasma levels are malabsorption of the drug, or unusually rapid drug metabolism or elimination. High levels may be caused by ingestion of greater than the prescribed dose, by impaired drug metabolism or elimination (genetic or constitutional), or because of interactions with other drugs. Elevation of phenytoin plasma level to the toxic range has been caused by a number of drugs, most predictably with disulfiram, sulthiame and isoniazid. When the interfering drug cannot be discontinued, the phenytoin dosage must be reduced and a new dose titrated using the plasma level as the indicator. In the treatment of petit mal, the order of preference of drugs is based on efficacy and safety, with ethosuximide the drug of choice, followed in order by trimethadione, para-methadione, methsuximide and phensuximide. In grand mal and focal epilepsies, phenytoin and phenobarbitone are the best understood and successful drugs. Primidone is also very effective. Carbamazepine can be used as an alternative. A number of other agents are available for use in combination with the prime drugs in highly refractory cases, or alone when the major drugs are contra-indicated. ACTH, or if this fails, diazepam or nitrazepam, can be used to abolish seizures in infantile spasms. Diazepam is the drug of choice in status epilepticus; given intravenously in status grand mal, or intramuscularly or intravenously in status petit mal and status focal seizures. Adverse effects of anticonvulsants can be divided into three groups: those which are a result of intoxication and which are related to elevated plasma levels. These include nystagmus, sedation and unsteadiness, and are usually, but not inevitably, a consequence of overdosage. The second category includes the side-effects, which occur as by-products of the pharmacological action of the drugs when used in the usual doses and with plasma levels in the desirable range. They may be more severe with higher doses. Frequent among these are gum hypertrophy with phenytoin, photophobia with trimethadione and depression of folate with hydantoins and barbiturates. Idiosyncratic reactions comprise the third category of unwanted effects; these are unpredictable, are unrelated to dose or plasma level and presumably occur on an allergic basis. Most frequent are skin rashes, and most dangerous are blood dyscrasias. The dysmorphogenic effects, the effects on the pregnant mother and on the newborn are reviewed.

Single-dose kinetics of 0.1 mg/kg intravenous diazepam, 10 mg/kg oral antipyrine, and 300 mg oral phenytoin were followed in healthy subjects before and after 400 mg metronidazole twice a day for 5 days. When data before metronidazole dosing were compared with those after metronidazole dosing, there were no changes in total plasma clearance of diazepam (0.53 and 0.65 ml/min/kg), antipyrine (39.0 and 38.0 ml/min/kg), or phenytoin (0.56 and 0.55 ml/min/kg). Plasma t 1/2s and volumes of distribution of the three drugs tested were not affected by metronidazole, but urinary excretion of 4-hydroxyantipyrine decreased after metronidazole dosing. There was no change in the elimination of phenytoin as its hydroxylated metabolite after metronidazole. It is concluded that, at therapeutic concentrations, metronidazole does not significantly inhibit oxidative drug metabolism

Many drug interactions can be demonstrated, but only a few are so clinically significant that they necessitate adjusting drug dosages. The same drug combination may produce changes of variable extent or direction in different individuals. The reasons for this variability include genetic control of the rate and inducibility of drug metabolism, and environmental factors such as contact with chemicals. Among antimicrobial agents, chloramphenicol may cause accumulation of phenytoin (PHT) and phenobarbital (PB), and isoniazid may cause PHT, carbamazepine (CBZ), and primidone (PRM) to accumulate. Erythromycin may cause accumulation of CBZ. Among anti-ulcer agents, antacids may reduce PHT concentration while cimetidine may cause accumulation of PHT, CBZ, and diazepam (DZP). Salicylates displace strongly binding drugs such as PHT, DZP, or valproate (VPA) from the binding sites in plasma proteins, which may lead to some decline of the total plasma level with an increase in the unbound drug percentage. Conversely, anticonvulsants may influence the dosage requirements of oral anticoagulants by inducing their metabolism. Failures of oral contraceptives have been attributed to anticonvulsants in some patients. Probably the most predictable interaction that necessitates dosage adjustment is accumulation of PB caused by VPA. Intentional inhibition of PRM metabolism by nicotinamide serves as an example of attempts to utilize an interaction for improved therapeutic effect

Healthy volunteers received single doses of either phenytoin (300 mg IV), alprazolam (1 mg orally) or lorazepam (2 mg IV) on two occasions in random sequence. One of the two trials was a control; for the other trial, subjects ingested metronidazole, 250 mg three times daily beginning 4 days prior to and continuing for the duration of each kinetic study. Compared with control, metronidazole significantly prolonged phenytoin half-life (23 versus 16 hours, P less than .02) and reduced its clearance (.28 versus .33 mL/min/kg, P less than .005), known to depend on aromatic hydroxylation. However, metronidazole did not significantly alter kinetic variables for either alprazolam (metabolized by aliphatic hydroxylation) or lorazepam (metabolized by glucuronide conjugation). Thus, metronidazole has the capacity to impair the clearance of certain oxidatively metabolized drugs, but there is no apparent way to predict which drugs will be so influenced

Antiepileptic drug interactions represent a common clinical problem which has been compounded by the introduction of many new compounds in recent years. Most pharmacokinetic interactions involve the modification of drug metabolism; the propensity of antiepileptic drugs to interact depends on their metabolic characteristics and action on drug metabolic enzymes. Phenobarbital, phenytoin, primidone and carbamazepine are potent inducers of cytochrome P450 (CYP), epoxide hydrolase and uridine diphosphate glucuronosyltransferase (UDPGT) enzyme systems; oxcarbazepine is a weak inducer of CYP enzymes, probably acting on a few specific isoforms only. All stimulate the rate of metabolism and the clearance of the drugs which are catabolised by the induced enzymes. Valproic acid (valproate sodium) inhibits to different extents many hepatic enzyme system activities involved in drug metabolism and is able to significantly displace drugs from plasma albumin. Felbamate is an inhibitor of some specific CYP isoforms and mitochondrial beta-oxidation, whereas it is a weak inducer of other enzyme systems. Topiramate is an inducer of specific CYP isoforms and an inhibitor of other isoforms. Ethosuximide, vigabatrin, lamotrigine, gabapentin and possibly zonisamide and tiagabine have no significant effect on hepatic drug metabolism. Apart from vigabatrin and gabapentin, which are mainly eliminated unchanged by the renal route, all other antiepileptic drugs are metabolised wholly or in part by hepatic enzymes and their disposition may be altered by metabolic changes. Some interactions are clinically unremarkable and some need only careful clinical monitoring, but others require prompt dosage adjustment. From a practical point of view, if valproic acid is added to lamotrigine or phenobarbital therapy, or if felbamate is added to phenobarbital, phenytoin or valproic acid therapy, a significant rise in plasma concentrations of the first drug is expected with a corresponding increase in clinical effects. In these cases a concomitant reduction of the dosage of the first drug is recommended to avoid toxicity. Conversely, if a strong inducer is added to carbamazepine, lamotrigine, valproic acid or ethosuximide monotherapy, a significant decrease in their plasma concentrations is expected within days or weeks, with a possible reduction in efficacy. In these cases a dosage increase of the first drug may be required