Many drug-drug interactions can accompany the use of anticonvulsants. As these agents are often used in combination, the addition of a second drug to a patient’s therapy may alter the plasma level of the initial treatment. This article discusses the principles underlying some important interactions
Other articles in this series include:
Antiarrhythmics (PJ, July 3, 1999, pp28-31)
Antihypertensives (PJ, April 17, pp547-51)
| Credit for learning: 1
This article will form the basis of questions under the PJ/College of Pharmacy Practice Credit for Learning Scheme |
Anticonvulsants have been associated with a wide range of drug-drug interactions in clinical use. Many drugs can affect the actions of anticonvulsants and vice versa. Interactions with anticonvulsants are usually pharmacokinetic in nature, although the precise mechanisms of many reported interactions are unclear. Anticonvulsants often have a narrow therapeutic index. An interacting drug which increases or decreases an anticonvulsant’s plasma level is likely to cause toxicity or loss of therapeutic control, respectively.
Because so many drug-drug interactions for anticonvulsants have been documented, it is not possible to discuss all of them within this article. Instead, the article focuses on the underlying principles of some important interactions encountered in practice, and summarises the implications of drug-drug interactions where two anticonvulsants are used together.
(a) Reduction of bioavailability A co-administered drug may reduce bioavailability of an anticonvulsant in several ways. It may affect the pH of the gastric or intestinal secretions such that the ionic characteristics of the drug alters. For example, oral phenytoin is less ionised and hence absorbed more effectively in an acid environment than in a neutral or alkaline one. Antacids have been shown to reduce the bioavailablity of phenytoin, perhaps by as much as 30 per cent.1 If an antacid is required by a patient taking phenytoin, it seems wise not to administer the agents within an hour of each other.
Some drugs may form insoluble complexes with an anticonvulsant, thereby reducing oral absorption. Acetazolamide has been reported to reduce the absorption of primidone, presumably by this mechanism.2
The absorption of phenytoin is known to be reduced when given to patients receiving enteral feeds. This is considered to be due to complex formation of phenytoin with constituents of the nutritional product, but patients receiving parenteral nutrition have also been shown to have reduced bioavailability of oral phenytoin. The mechanism for this latter phenomenon remains unclear.3
(b) Effects of drugs on the distribution characteristics of anticonvulsants Anticonvulsants vary considerably in relation to how they distribute throughout body fluids and tissues following absorption. Moreover, some agents may be highly bound to plasma proteins. There is potential for many co-administered drugs to interfere with these characteristics. Knowledge of the pharmacokinetic properties of both drugs is required to anticipate whether problems are likely when they are taken together. In general, an interaction involving alteration of an anticonvulsant’s distribution is only likely to be of clinical significance if there is also an effect on the drug’s elimination.
Displacement from plasma protein binding sites - Several anticonvulsants are extensively bound to plasma proteins, and are subject to displacement by other drugs which compete for their binding sites. Phenytoin, for example, is normally 90 per cent protein bound, mainly to serum albumin. Many acidic drugs (eg, salicylates, sodium valproate, some non-steroidal anti-inflammatory drugs, warfarin) are also strongly bound to albumin, and displacement of phenytoin can occur when such drugs are started in a patient maintained on phenytoin. The effect of a displacement interaction is normally a transient increase in concentration of free (unbound) phenytoin, which then becomes available to leave the bloodstream to distribute to peripheral tissue sites, and to the liver, which is the main organ of elimination of phenytoin. The net result of these changes is usually a reduction in the total phenytoin plasma concentration, although the free drug concentration (the therapeutically active moiety) is minimally affected.
The main clinical problem arising from this type of interaction is that the fall in measured phenytoin may be misinterpreted as a need to increase dosage, and thereby increase the risk of phenytoin toxicity. The process and its clinical implications are summarised in the case study (above).
(c) Interference with the hepatic metabolism of an anticonvulsant by a co-administered drug This type of interaction is one of the most commonly encountered in clinical practice. Many anticonvulsants are extensively metabolised by cytochrome P450 mediated oxidation pathways. The clearance of these drugs may be enhanced by agents which induce the synthesis of enzymes involved in these processes, or may be decreased by drugs which inhibit the activity of these enzymes. Some anticonvulsants are themselves enzyme inducing agents (phenytoin, carbamazepine) or enzyme inhibitors (sodium valproate). As anticonvulsants are frequently used in combination, one agent can often affect the plasma concentration of another. Some of the more recently introduced anticonvulsants (vigabatrin, gabapentin) are cleared by renal excretion, and hence are not generally susceptible to this type of interaction.
Interactions which result in an enhanced clearance of an anticonvulsant - There are several examples of this in the literature. Most documented interactions relate to situations where a second anticonvulsant is added to the therapy of a patient who has been taking a single agent for some time, the second drug being an inducing agent such as phenytoin, carbamazepine or phenobarbitone. It would be expected that the second drug would result in a fall in steady-state plasma concentration of the first, as the latter’s clearance would increase over time. For example, it has been shown that patients receiving carbamazepine with phenobarbitone showed a modest reduction in their steady-state carbamazepine concentrations. This was attributed to enzyme induction by phenobarbitone.4
Some cases, however, are complex in mechanism, especially where both drugs are inducing agents but may additionally compete for the same metabolic pathways. For example, when carbamazepine is given to a patient already receiving phenytoin, the effect is unpredictable, and may result in either a rise or a fall in phenytoin concentration.5 Generally, where concentrations of phenytoin have been altered by the co-administration of carbamazepine, this has not led to loss of seizure control or to toxicity.
Other drugs which are potent inducers of oxidative enzymes (eg, rifampicin) may also enhance clearance of phenytoin, requiring dosage adjustment of the latter during co-administration.
Lamotrigine is cleared by hepatic metabolism, but by glucuronidation rather than by oxidative pathways. Interestingly, paracetamol may induce glucuronidase enzymes involved in this mechanism, and may cause a reduction in lamotrigine plasma concentrations. The clinical significance of this interaction is unclear.6
Interactions which result in a reduced clearance of an anticonvulsant - Many drugs may inhibit the metabolism of anticonvulsants. This interaction leads to an increase in steady-state serum level which, if the anticonvulsant has a narrow therapeutic index, increases the risk of toxicity. A few examples of interactions with likely clinical significance are given here.
Phenytoin - This type of interaction is a particular problem with phentoin as the drug exhibits saturable metabolism even in the absence of an enzyme inhibitor drug. If phenytoin clearance is reduced sufficiently by an inhibitor such that its elimination rate ("rate out") becomes less than the dose taken ("rate in"), accumulation and toxicity will inevitably result. Serious toxicity has occurred when phenytoin has been given with a variety of enzyme inhibitors, including cimetidine, chloramphenicol and co-trimoxazole. Increases in phenytoin concentrations of less clinical significance have occurred with several drugs, including omeprazole, ceftriaxone, isoniazid and dextropropoxyphene.
Carbamazepine - As with phenytoin, a variety of drugs can inhibit this drug’s metabolism, increasing risk of accumulation. Erythromycin and, to a lesser extent, other macrolides, are well recognised to cause significant elevations of carbamazepine concentrations. Erythromycin, for example, has been reported to reduce carbamazepine clearance by 40 per cent, and to cause a three-fold increase in plasma concentration.7 Cimetidine, verapamil and several serotonin re-uptake inhibitor antidepressants (SSRIs) have all been associated with an increase in carbamazepine concentration and concomitant toxicity when co-administered.
Lamotrigine - As stated above, lamotrigine is cleared primarily by hepatic glucuronidation. Sodium valproate competitively inhibits glucuronidase enzymes and hence increases plasma lamotrigine concentrations. As this interaction may increase the risk of lamotrigine toxicity, the initial and maintenance doses of lamotrigine should be approximately halved for a patient receiving valproate.
As already stated, several anticonvulsants are either enzyme inducers or inhibitors and hence may affect the pharmacokinetics of many other drugs when co-administered. Some clinically important situations have been covered in a previous article (Interactions with warfarin, October 31, 1998, p704) and the reader is referred to this for details. This section deals with a number of significant interactions which pharmacists may encounter in clinical practice.
There are certain situations in which an anticonvulsant can cause a decreased effect of a co-administered drug. These interactions are usually due to the anticonvulsant’s effect as an inducing agent, reducing the circulating plasma concentration of the co-administered agent and hence its therapeutic effect. These interactions are generally common to phenytoin, carbamazepine and phenobarbitone/primidone.
Diuretics - The efficacy of frusemide may be reduced by 50 per cent when used with phenytoin. This may be partly due to enhanced metabolism of frusemide, but also to an impairment of the transport processes of frusemide to its site of action within the renal tubules, caused by phenytoin.8
Antiarrhythmic agents - Plasma concentrations of disopyramide and quinidine may be significantly reduced by phenytoin, presumably by enzyme induction. A reduction in the effectiveness of flecainide has been reported to occur when given with carbamazepine.9
Corticosteroids - Phenytoin, carbamazepine and phenobarbitone have been associated with a reduction in circulating plasma concentrations of dexamethasone. This may be clinically significant in two respects. First, patients requiring therapy with dexamethasone for the suppression of cerebral oedema are likely to need high doses (perhaps double the normal dose) for adequate response. Secondly, the use of this agent for the "dexamethasone suppression test" for the diagnosis of Cushing’s syndrome may give misleading results in a patient receiving phenytoin. Such patients are likely to show a less than normal suppression of corticosteroid secretion after a low dose of dexamethasone, but a "normal" suppression profile after a high dose, suggesting a positive diagnosis.
Neuromuscular blocking drugs - Patients receiving vecuronium and pancuronium may show a shorter recovery time from these agents if they have been receiving therapy with phenytoin prior to surgery. This may be partly due to enhanced clearance of the drugs caused by enzyme induction; alteration in sensitivity of receptors may also contribute to this effect.10
Opioids - Phenobarbitone has been reported to reduce plasma concentrations of methadone and to precipitate withdrawal symptoms in patients stabilised on this drug.11 Phenytoin and carbamazepine have also been associated with a reduction in methadone concentrations in chronic users. Although pethidine metabolism is induced by phenobarbitone, the enhanced production of the active metabolite norpethidine may lead to an increase in sedation and toxicity.12
Anticonvulsants are a heterogeneous group of drugs, many of which have both a narrow therapeutic index and the ability to interact with a wide range of other co-administered drugs, including other anticonvulsants. Most interactions, especially with the older agents, are pharmacokinetic in mechanism, and are predictable with a knowledge of the pharmacology of the drugs.
Case study: effect of protein binding displacement of phenytoin by sodium valproateA 25-year-old man with major tonic-clonic seizures, Mr P, is admitted for assessment of his epileptic control. He has been taking phenytoin 300mg daily as single agent therapy for several months, but has been experiencing an increase in his frequency of seizures over the past three weeks. The patient’s phenytoin plasma concentration measured on admission is 20μg/L (reference range 10-20μg/ml). It is decided to add sodium valproate 500mg tds to Mr P’s treatment, while maintaining his current dosage of phenytoin. His clinical control improves somewhat during the next few days. After a week, a second phenytoin plasma assay reveals a level of 16μg/ml. The registrar wonders if there is scope for an increase in Mr P’s phenytoin dosage to further improve his control. What would you advise ? COMMENT
YOUR RESPONSE
|
Mr Brown is chief pharmacist, Southport and Formby NHS Trust
|
Ahmad S. Reduced sensitivity to frusemide caused by chronic anticonvulsant therapy. BMJ 1974;3:657. 9. Spina E. Pisani F, Perruca E. Clinically significant pharmacokinetic drug interactions with carbamazepine. An update. Pharmacokinetics 1996;31:198-214. 10. Chen J, Kim YD, Dubois M, et al. The increased requirement of pancuronium in neurosurgical patients receiving phenytoin chronically. Anaesthesiology 1983;59:A288. 11. Lui SJ, Wang RIH. Case report of barbiturate-induced enhancement of methadone metabolism and withdrawal syndrome. Am J Psychiatry 1984:141:1287-8. 12. Staumbaugh JE, Wainer IW, Hemphill DM, et al. A potentially toxic drug interaction between pethidine (meperidine) and phenobarbitone. Lancet 1977;1:398-9. 13. Houghton GW, Richens A. Inhibition of phenytoin metabolism by other drugs used in epilepsy. Int J Clin Pharmacol, Therapy Toxicol 1975;12: 210. |