• antiepileptic drugs;
  • drug interactions;
  • enzyme induction;
  • enzyme inhibition;
  • epilepsy;
  • review


  1. Top of page
  2. Abstract
  3. Introduction
  4. Mechanisms of AED interactions
  5. Interactions between AEDs
  6. Interactions between AEDs and other drugs
  7. Conclusions
  8. References

Some patients with difficult-to-treat epilepsy benefit from combination therapy with two or more antiepileptic drugs (AEDs). Additionally, virtually all epilepsy patients will receive, at some time in their lives, other medications for the management of associated conditions. In these situations, clinically important drug interactions may occur. Carbamazepine, phenytoin, phenobarbital and primidone induce many cytochrome P450 (CYP) and glucuronyl transferase (GT) enzymes, and can reduce drastically the serum concentration of associated drugs which are substrates of the same enzymes. Examples of agents whose serum levels are decreased markedly by enzyme-inducing AEDs, include lamotrigine, tiagabine, several steroidal drugs, cyclosporin A, oral anticoagulants and many cardiovascular, antineoplastic and psychotropic drugs. Valproic acid is not enzyme inducer, but it may cause clinically relevant drug interactions by inhibiting the metabolism of selected substrates, most notably phenobarbital and lamotrigine. Compared with older generation agents, most of the recently developed AEDs are less likely to induce or inhibit the activity of CYP or GT enzymes. However, they may be a target for metabolically mediated drug interactions, and oxcarbazepine, lamotrigine, felbamate and, at high dosages, topiramate may stimulate the metabolism of oral contraceptive steroids. Levetiracetam, gabapentin and pregabalin have not been reported to cause or be a target for clinically relevant pharmacokinetic drug interactions. Pharmacodynamic interactions involving AEDs have not been well characterized, but their understanding is important for a more rational approach to combination therapy. In particular, neurotoxic effects appear to be more likely with coprescription of AEDs sharing the same primary mechanism of action.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Mechanisms of AED interactions
  5. Interactions between AEDs
  6. Interactions between AEDs and other drugs
  7. Conclusions
  8. References

Although monotherapy remains the mainstay for the treatment of epilepsy, combinations of antiepileptic drugs (AEDs) are used frequently in patients not responding to a single medication. AEDs may also be combined with drugs used to treat intercurrent or associated conditions. When multiple drug therapy is used, there is a possibility of clinically relevant drug interactions, which in patients with epilepsy are particularly common for a variety of reasons: (i) AEDs are administered for prolonged periods, often for a lifetime, thereby increasing the probability of coprescription; (ii) most AEDs have a narrow therapeutic index, and even relatively modest alterations in their pharmacokinetics can result in loss of response or toxic effects; (iii) the most widely used AEDs (carbamazepine, valproic acid, phenytoin and phenobarbital) have prominent effects on the activity of enzymes which metabolize the majority of existing medication; (iv) most of the old and new generation AEDs are substrates of the same enzymes [1, 2].

Comprehensive reviews of interactions involving AEDs have been published recently [1–5]. The purpose of the present review is to highlight those which, because of their frequency or magnitude, are especially likely to have adverse clinical consequences.

Mechanisms of AED interactions

  1. Top of page
  2. Abstract
  3. Introduction
  4. Mechanisms of AED interactions
  5. Interactions between AEDs
  6. Interactions between AEDs and other drugs
  7. Conclusions
  8. References

The vast majority of clinically important AED interactions result from induction or inhibition of drug metabolizing enzymes. However, other mechanisms, including pharmacodynamic interactions, may be occasionally at play.

Enzyme induction

Carbamazepine, phenytoin, phenobarbital and primidone (henceforth referred to collectively as enzyme-inducing AEDs) stimulate the activity of a variety of cytochrome P450 (CYP) enzymes, including CYP1A2, CYP2C9, CYP2C19 and CYP3A4, as well as glucuronyl transferases (GT) and epoxide hydrolase [6–9]. Because these enzymes are involved in the biotransformation of the majority of therapeutic agents, patients taking enzyme inducing AEDs metabolize at a faster rate a wide range of concomitantly administered medications, whose dosage requirements may be consequently increased. For drugs which are converted to active or toxic metabolites, conversely, enzyme induction may result in enhancement of the activity of the affected drug: one example is represented by the induction of primidone metabolism by phenytoin, which results in increased serum concentrations of the active metabolite phenobarbital, with the attendant risk of phenobarbital-related adverse effects [1, 10].

None of the newer AEDs shares the broad spectrum enzyme-inducing activity of older generation agents [11]. However, oxcarbazepine, lamotrigine, felbamate and, at dosages ≥ 200 mg day−1, topiramate stimulate the metabolism of oral contraceptive steroids, possibly by tissue-selective stimulation of CYP3A4 [12, 13], and oxcarbazepine has also a stimulating effect on the GT-mediated lamotrigine metabolism [14] and, to a lesser extent, the CYP3A4-mediated oxidation of felodipine [15]. In addition, most new generation AEDs are cleared fully or partly by inducible enzymes (Table 1), and they are therefore a target for interactions mediated by enzyme induction.

Table 1.  Main routes of elimination of antiepileptic drugs (AEDs)
DrugMain route(s) of eliminationMain enzyme system involved
  • 1

    Oxcarbazepine is a prodrug, virtually entirely converted by cytosolic aryl-ketone-reductase to the active metabolite monohydroxycarbazepine (MHD). The indicated routes of elimination and enzymes involved refer to MHD.

Old generation AEDs
CarbamazepineOxidationCYP3A4 (active 10,11-epoxide metabolite cleared by epoxide hydrolase)
PhenobarbitalOxidation + N-glucosidation (75% of the dose) and renal excretion (25%)CYP2C9 and CYP2C19
PhenytoinOxidationCYP2C9 and CYP2C19
Valproic acidOxidation (>50%) and glucuronide conjugation (30–40%)Mitochondrial oxidases, CYPs and glucuronyl transferases
New generation AEDs
FelbamateOxidation (>50%) and renal excretion (>30%)Inducible CYP isoforms
GabapentinRenal excretionNone
LamotrigineGlucuronide conjugationGlucuronyl transferase type 1A4
LevetiracetamRenal excretion (75%) and hydrolysis (25%)Hydrolase
Oxcarbazepine1Glucuronide conjugation (>50%) and renal excretion (<30%)Glucuronyl transferases
PregabalinRenal excretionNone
TopiramateOxidation (20–60%) and renal excretion (40–80%)Inducible CYP isoforms
S-VigabatrinRenal excretionNone
ZonisamideOxidation + reduction + N-acetylation (>50%) and renal excretion (30%)CYP3A4 and N-acetyl-transferases

Enzyme inhibition

Valproic acid differs from other older generation AEDs in being an inhibitor rather than an inducer of drug metabolizing enzymes, including those involved in the oxidation of phenobarbital, the glucuronidation of lamotrigine and the conversion of carbamazepine-10,11-epoxide to the corresponding diol (epoxide hydrolase) [1]. Oxcarbazepine is a weak inhibitor of CYP2C19, and may increase by this mechanism the plasma levels of phenytoin and, to a lesser extent, phenobarbital [1]. Felbamate, an AED rarely used because of serious haematological and hepatic toxicity, is a more potent metabolic inhibitor, and it causes important elevations in the serum levels of phenytoin, phenobarbital, valproic acid, carbamazepine-10,11-epoxide, and N-desmethylclobazam [1, 11]. Other clinically important interactions mediated by enzyme inhibition are those whereby the metabolism of an AED is inhibited by drugs used for other indications.

Other pharmacokinetic mechanisms

Clinically important AED interactions mediated by inhibition of gastrointestinal absorption have been reported rarely. In recent years, however, evidence has been provided that medications which induce or inhibit CYP enzymes may also modulate the expression of drug transporters, including P-glycoprotein (P-gp) and multiple drug resistance proteins 2 and 3 (MRP2 and MRP3), in the gastrointestinal tract, in the kidney and in other tissues [16, 17]. These observations raise the possibility that some AED interactions currently ascribed to enzyme induction may in fact be mediated by reduced gastrointestinal absorption or enhanced renal elimination of the affected drug, as shown in a recent elegant study of the interaction between carbamazepine and the β1-blocker talinolol [18].

Valproic acid, phenytoin and tiagabine are highly bound to plasma proteins and may be involved in displacement from protein binding sites. The most common of these interactions is the displacement of plasma protein-bound phenytoin by valproic acid [19, 20]. Unless additional mechanisms are at work, these interactions are not clinically important, because the displaced drug is diluted into a large volume of distribution and/or is rapidly cleared, resulting in a new state in which the total serum concentration of the affected drug is decreased but the unbound (pharmacologically active) concentration is unchanged [21]. Nevertheless, clinicians must be aware of these interactions when interpreting serum drug concentration data. In particular, in patients taking phenytoin in combination with valproate, therapeutic and toxic effects will occur at total serum phenytoin concentrations which are lower than those required to produce equivalent effects in patients not taking valproate [19].

Pharmacodynamic interactions

Pharmacodynamic interactions are usually inferred by default when a change in clinical response apparently due to a drug interaction is not reflected in any identifiable pharmacokinetic change [1]. Pharmacodynamic interactions involving AEDs are difficult to document objectively, given the complexity of quantifying dose–response relationships in the clinical setting. Clinical observations, however, suggest that these interactions may be more common than previously thought, and their characterization would be important for a more rational approach to the use of AED combinations in epilepsy [22].

Interactions between AEDs

  1. Top of page
  2. Abstract
  3. Introduction
  4. Mechanisms of AED interactions
  5. Interactions between AEDs
  6. Interactions between AEDs and other drugs
  7. Conclusions
  8. References

Interactions resulting in decreased concentration of the affected drug

The four major enzyme-inducing AEDs (carbamazepine, phenytoin, phenobarbital and primidone) stimulate the metabolism and reduce the serum concentration of most other concurrently administered AEDs, most notably valproic acid [23, 24], tiagabine [25], ethosuximide [26], lamotrigine [27, 28], topiramate [29], oxcarbazepine and its active monohydroxy-derivative (MHD) [30], zonisamide [31], felbamate [32] and many benzodiazepine drugs [33, 34]. The metabolism of carbamazepine itself is subject not only to autoinduction, but also to heteroinduction by phenytoin and barbiturates [35].

The clinical significance of these interactions is usually relatively modest, because the partial loss of efficacy resulting from the decreased serum concentration of the affected drug tends to be compensated for by the antiepileptic effect of the added drug. However, when the decrease in AED concentration is particularly prominent, the interaction may result in worsened seizure control, unless the dosage of the affected drug is adjusted accordingly [36]. AEDs whose dosage requirements are particularly increased in the presence of enzyme-inducing comedication include valproic acid, carbamazepine, lamotrigine and tiagabine. For example, the plasma concentration of valproic acid can be reduced on average by 50–75% in patients comedicated with enzyme inducers [24], and the concentration of lamotrigine is also reduced by over 50%[28]. Conversely, the plasma concentrations of levetiracetam, gabapentin, pregabalin and vigabatrin are not affected to any important extent by comedication with other AEDs [11].

Since enzyme induction is a reversible phenomenon, particular caution is required when an enzyme-inducing agent is discontinued, because the serum concentration of concurrently administered AEDs may increase to potentially toxic levels. For example, in patients comedicated with valproic acid, discontinuation of carbamazepine, or substitution of carbamazepine with another AED devoid of enzyme-inducing activity, may result in a prominent rebound increase of serum valproic acid concentrations, with the attendant risk of toxicity [37]. Likewise, clinically important increases in the serum concentration of tiagabine and lamotrigine may be observed after withdrawal of enzyme-inducing comedication, or substitution of the latter with other AEDs which do not have enzyme-inducing effects.

Interactions resulting in increased concentration of the affected drug

Two clinically important interactions stand out in this category: the inhibition of the metabolism of lamotrigine and phenobarbital by valproic acid. Inhibition of lamotrigine metabolism is already maximal at dosages of valproate around 500 mg day−1, and results in an approximate twofold increase in serum lamotrigine levels [38–40]. Because the risk of lamotrigine-induced skin rashes is dependent on the rate of rise of serum lamotrigine concentration, in patients comedicated with valproate lamotrigine should be initiated at reduced dosages (in adults, 25 mg on alternate days) and titrated more slowly to target dosages which are lower than those used in patients not taking valproate. Although there is no risk of rash when valproate is added on in a patient already stabilized on lamotrigine, neurotoxic effects may occur if the dosage of the latter is not reduced by about 50% as soon as the dosage of valproate reaches, in an adult, about 250–500 mg day−1.

The second clinically important interaction caused by valproic acid is an elevation in serum phenobarbital levels, which is probably secondary to inhibition of CYP2C9 and/or CYP2C19 [41]. Although today phenobarbital is used infrequently in Europe and the USA, it remains the most commonly prescribed AED in developing countries, and therefore its interaction with valproate is particularly important. After adding valproate, serum phenobarbital concentrations increase over several weeks by about 30–50% in most patients, but individual patients may show a greater interaction. A reduction in phenobarbital (or primidone) dosage by up to 80% may be required to avoid adverse effects.

Less important interactions mediated by metabolic inhibition include an increase in serum phenytoin (by up to 40%) after administration of oxcarbazepine [42], and an inconsistent increase in serum phenytoin after addition of topiramate [43]. In carbamazepine-treated patients started on valproate comedication, neurotoxic signs may occasionally be caused by an increase in serum carbamazepine-10,11 epoxide levels, secondary to inhibition of epoxide hydrolase by valproic acid [44]. Interactions caused by sulthiame and felbamate, while clinically relevant (Table 2), are uncommon because these agents are rarely used in the current management of epilepsy.

Table 2.  Drugs which have been found to increase the serum concentration of antiepileptic drugs, presumably by inhibiting their metabolism
Affected drug Interfering drug
  1. The list should not be regarded as exhaustive. For further information and a list of references, see Patsalos and Perucca [1, 2]. 1These drugs increase the concentration of the active metabolite carbamazepine-10,11-epoxide, the effect being most clinically relevant with valpromide. The concentration of carbamazepine is not affected by valproic acid and valpromide, ands it is decreased by felbamate. 2Interaction inconsistent and limited to an increase in unbound phenytoin concentration. Total serum phenytoin concentration usually decreases due to displacement from plasma protein binding sites.

CarbamazepineAntiepileptic drugs:Felbamate1, valproic acid1, valpromide1
Antidepressants:Fluoxetine, fluvoxamine, nefazodone, trazodone, viloxazine
Antimicrobials:Clarithromycin, erythromycin, fluconazole, isoniazid, ketoconazole, metronidazole, ritonavir, troleandomycin
Miscellaneous:Cimetidine, danazol, dextropropoxyphene, diltiazem, risperidone, quetiapine1, ticlopidine, verapamil
LamotrigineAntiepileptic drugs:Valproic acid
PhenobarbitalAntiepileptic drugs:Felbamate, phenytoin, sulthiame, valproic acid
PhenytoinAntiepileptic drugs:Felbamate, oxcarbazepine, sulthiame, valproic acid2
Antidepressants:Fluoxetine, fluvoxamine, imipramine, sertraline, trazodone, viloxazine
Antimicrobials:Chloramphenicol, fluconazole, isoniazid, miconazole, sulfaphenazole
Antineoplastic drugs:Doxifluridine, fluorouracil, tamoxifen, tegafur, UFT
Miscellaneous:Allopurinol, amiodarone, azapropazone, cimetidine, chlorpheniramine, dextropropoxyphene, diltiazem, disulfiram omeprazole, phenylbutazone, sulfinpyrazone tacrolimus, ticlopidine, tolbutamide
Valproic acidAntiepileptic drugs:Felbamate

Pharmacodynamic interactions

Clinical observations suggest that certain AED combinations may be associated with adverse or beneficial interactions at pharmacodynamic level [22]. In particular, it has been shown repeatedly that the combination of valproate with lamotrigine [45], or valproate with ethosuximide [46], may produce seizure control in patients who did not respond to the highest tolerated dose of either drug given alone: this may be explained by a pharmacodynamic interaction resulting in synergistic efficacy and/or infra-additive neurotoxicity. Conversely, the combinations of lamotrigine with carbamazepine [47], or oxcarbazepine with carbamazepine [42], have been more commonly associated with neurotoxic effects compared with combinations of the same drugs with other agents, an observation which may be explained by the common action of carbamazepine, lamotrigine and oxcarbazepine in blocking voltage-dependent sodium channels. Indeed, combinations of drugs acting by different mechanisms would be expected to be more beneficial than combinations of drugs sharing the same mode of action, even though current knowledge of the pharmacology of AEDs is still insufficient to allow a mechanistic approach to AED therapy [22].

Interactions between AEDs and other drugs

  1. Top of page
  2. Abstract
  3. Introduction
  4. Mechanisms of AED interactions
  5. Interactions between AEDs
  6. Interactions between AEDs and other drugs
  7. Conclusions
  8. References

Interactions resulting in decreased AED concentration

Serum lamotrigine levels are decreased by about 50% of oral contraceptive steroids, an interaction which is likely to be caused by stimulation of uridine GT type 1A4 (UGT1A4) activity by the steroids [48]. This interaction can result in reduced seizure control in some women [49]. Interestingly, the interaction follows a cyclic pattern, with a marked decrease in serum lamotrigine levels during the 21 days of intake of the oestro-progestinic pill, and a twofold rebound increase in AED concentration during the pill-free week [13].

Other interactions resulting in a clinically important decrease in serum AED concentration have been rarely reported, but they can be important in individual cases. Notable examples include the marked inhibition of the gastrointestinal absorption of phenytoin given concurrently with some nasogastric feeds [50, 51], the decrease in serum phenytoin concentrations caused by cisplatin and some other antineoplastic drugs [52], and the dramatic fall in serum valproic acid concentration after addition of some antibiotics of the carbapenem class [53, 54].

Interactions resulting in increased AED concentration

Only extensively metabolized AEDs are affected by these interactions, which are mediated by metabolic inhibition [2]. Interactions resulting in elevated serum AED concentrations have been reported mostly with carbamazepine, phenytoin and phenobarbital, while new AEDs appear to be rarely affected (Table 2). This is partly explained by the fact that many new AEDs undergo little or no biotransformation (Table 1), but it is also possible that interactions affecting new AEDs are under-recognized due to limited clinical experience and lack of routine application of therapeutic drug monitoring [11].

Theoretically, any of the interfering drugs listed in Table 2 can precipitate clinical signs of intoxication with the affected drug, although the magnitude of interaction varies from drug to drug and from patient to patient. Potentially serious adverse consequences can be minimized by careful clinical observation and monitoring of the serum concentration of the potentially affected drug. Most interactions mediated by metabolic inhibition can be predicted based on the contribution of specific isozymes to the clearance of the affected drug (Table 1) and knowledge of the influence of the comedication of interest on these isozymes [55]. For example, precipitation of serious carbamazepine toxicity by clarithromycin, erythromycin and troleandomycin can be predicted by the ability of the latter to inhibit CYP3A4 activity [54–57]. Azithromycin, dirithromycin, rokitamycin and spiramycin are macrolide antibiotics which do not inhibit CYP3A4 and therefore are not expected to affect carbamazepine metabolism.

Interactions resulting in decreased concentration of other drugs

A large proportion of patients with epilepsy are treated with enzyme-inducing AEDs, and therefore they represent a unique population in terms of vulnerability to interactions mediated by enzyme induction [11]. The list of such interactions is impressive (Table 3), and most are clinically significant. In terms of magnitude, the interaction is greatest with drugs subject to extensive first-pass metabolism such as itraconazole [58], praziquantel [59], indinavir [60–62], and most dihydropyridine calcium antagonists [63–65], whose plasma concentration may be decreased over five- to 10-fold in enzyme induced patients. Because of practical difficulties in compensating for such interactions, effective use of some of these drugs may not feasible in enzyme-induced patients. Apart from the magnitude of the pharmacokinetic change, the clinical implications also vary depending on the therapeutic index of the affected drug. Serious and even irreversible consequences may be observed, as in the case of oral contraceptives (Box 1), oral anticoagulants (Box 2), immunosuppressants and chemotherapeutic agents. In some categories of patients, e.g. those requiring anticancer therapy, the use of AEDs devoid of enzyme-inducing properties is clearly preferred [2, 52].

Table 3.  Drugs whose serum concentration has been reported to be decreased by coadministration of enzyme-inducing antiepileptic drugs (AEDs) (carbamazepine, phenobarbital, phenytoin and primidone)
  1. These interactions have not necessarily been shown with all enzyme inducers, and there can be differences in the enzyme-inducing and inhibiting spectrum of carbamazepine, phenobarbital, phenytoin and primidone. The list should not be regarded as exhaustive. For further information and a list of references, see Patsalos and Perucca [2] and Vecht et al.[52]. 1Some of these drugs (for example, bupropion, procarbazine, ifosfamide, amiodarone, disopyramide) have active metabolites. Therefore, a reduced concentration of parent drug does not necessarily imply a reduced pharmacological effect. 2Interaction reported with phenobarbital and probably due to impaired griseofulvin absorption. 3Interaction likely to be extended to other CYP3A4 substrates such as nevirapine, efavirenz, delavirdine, ritonavir and saquinavir. 4Interaction only relevant after oral administration of verapamil. 5Phenytoin may cause an initial decrease in anticoagulant effect, followed by an increase in warfarin concentration. 6There is suggestive evidence that oxcarbazepine may also decrease serum cyclosporin A levels. 7Other AEDs which have been found to decrease the concentration of the oestrogen and/or progestagen component of oral contraceptive steroids include felbamate, lamotrigine, oxcarbazepine and, at dosages > 200 mg day−1, topiramate.

Antidepressants1Amitriptyline, bupropion, citalopram, clomipramine, desipramine, desmethylclomipramine, doxepin, imipramine, mianserin, mirtazepine, nefazodone, nortriptyline, paroxetine, protriptyline
AntimicrobialsAlbendazole, doxycycline, griseofulvin2, indinavir3, itraconazole, metronidazole, praziquantel
Antineoplastic drugs19-aminocampthotecin, busulfan, cyclophosphamide, etoposide, ifosfamide, irinotecan, methotrexate, nitrosureas, paclitaxel, procarbazine, tamoxifen, teniposide, thiotepa, topotecan, vinca alkaloids
Antipsychotic drugsChlorpromazine, clozapine, haloperidol, mesoridazine (active metabolite of thioridazine), olanzapine, quetiapine, risperidone, ziprasidone
BenzodiazepinesAlprazolam, clobazam, clonazepam, desmethyldiazepam, diazepam, midazolam
Cardiovascular drugs1Alprenolol, amiodarone, atorvastatin, dicoumarol, digoxin, disopyramide, felodipine, metoprolol, mexiletine, nifedipine, nimodipine, nisoldipine, propranolol, quinidine, simvastatin, verapamil4, warfarin5
ImmunosuppressantsCyclosporin A6, sirolimus, tacrolimus
SteroidsCortisol, dexamethasone, hydrocortisone, methylprednisolone, prednisone, prednisolone, steroid oral contraceptives7
MiscellaneousFentanyl, metadone, metyrapone, misonidazole, paracetamol, pethidine, theophylline, thyroxine, vecuronium (and some other nondepolarizing neuromuscular blocking agents)

Interactions resulting in increased concentration of other drugs

Interactions whereby AEDs increase the serum concentrations of drugs used for other indications are observed rarely. By virtue of its enzyme-inhibiting properties, valproic acid may increase the plasma levels of a variety of drugs, including zidovudine, lorazepam, nimodipine, paroxetine, amitryptiline, nortriptyline, nitrosureas and etoposide [2]. At least for amitriptyline and nortriptyline [66, 67], and for some antineoplastic drugs [52], these interactions may cause signs of toxicity.

Box 1 Antiepileptic drugs and oral contraceptives

Despite the fact that prescription of oral contraceptives in women with epilepsy is relatively common, knowledge of the interactions occurring between these agents and antiepileptic drugs (AEDs) is unsatisfactory. In a 1996 survey conducted in the USA, only 4% of neurologists and none of the obstetricians were correct in identifying the interactions between the combined oral contraceptive pill and six major older generation AEDs [69]. It is no surprise,therefore, that surveys also found women to be poorly informed on this issue [70].

The following AEDs have been shown to induce the metabolism of the oestrogen and/or progestagen components of the contraceptive pill [2, 12]:

  • • 
  • • 
  • • 
  • • 
  • • 
  • • 
  • • 
  • • 

In women taking the above AEDs, the efficacy of the pill may be reduced [12]. The magnitude of interaction may vary with type and dose of AED. In particular, topiramate does not affect serum norethisterone and ethinylestradiol levels at dosages up to 100 mg day−1, has little or no effects at 200 mg day−1, and decreases serum norethisterone consistently at higher doses [71, 72]. Lamotrigine 300 mg day−1 has been found to cause only a modest reduction (19%) in levonorgestrel levels [13]. In comparison, the reduction in ethinylestradiol and progestagen levels with carbamazepine 600 mg day−1 is in the order of 50%[12, 71, 73]. Gabapentin, levetiracetam, pregabalin, tiagabine, valproate, vigabatrin and zonisamide have been reported not to interact with steroid oral contraceptives. Benzodiazepines and ethosuximide are also considered not to interact [2, 12].

In general, AEDs not interacting with oral contraceptives are a preferable treatment choice for women with epilepsy in whom contraception is contemplated. If oral contraceptives are combined with AEDs which induce their metabolism, ethinylestradiol dose should be increased from between 20 and 35 µg to 50 µg. If breakthrough bleeding occurs, some authors recommend increasing ethinylestradiol dose further to 75 or 100 µg [12]. Although even high-dosage pills may not provide full protection, pregnancy rates are still much lower compared with barrier methods; moreover, a spermicidal gel or a barrier method could be used in addition to the pill to increase the level of protection. With respect to alternative methods of contraception, intrauterine contraceptives releasing levonorgestrel directly into the uterine cavity are considered to act mainly through a local effect and their efficacy is considered not be be affected in a major way by enzyme-inducing AEDs [12], even though more data are needed [74].

Reciprocal interactions may occur whereby oral contraceptives affect serum AED concentrations. Administration of the combined contraceptive pill causes a decrease in serum lamotrigine levels by about 50%[48], which may lead to loss of seizure control in some women [49]. Conversely, a rebound increase in serum lamotrigine levels with possible signs of toxicity may be observed when the contraceptive pill is discontinued. This interaction follows a cyclic pattern, with a marked decrease in serum lamotrigine levels during the 21 days of intake of the pill, and an increase in lamotrigine concentration during the pill-free week [13]. A similar cyclic interaction resulting in markedly decreased serum AED concentration during the 21 days of pill intake has been described for valproate in a single case report, and requires confirmation [75].

Pharmacodynamic interactions

Pharmacodynamic interactions between AEDs and other drugs are poorly characterized. A well-documented example is the potentiation of the effects of nondepolarizing neuromuscular blockers (NDNMBs) by acutely administered AEDs [68]. Conversely, chronic AED therapy may cause resistance to NDNMBs, due to a combination of effects such as enzyme induction and upregulation of acetylcholine receptors.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Mechanisms of AED interactions
  5. Interactions between AEDs
  6. Interactions between AEDs and other drugs
  7. Conclusions
  8. References

Clinically important AED interactions are frequently observed in clinical practice, and often they can be anticipated by knowledge of the underlying mechanism. Whenever possible, these interactions should be prevented by avoiding the unnecessary use of polytherapy, and by selecting comedications which are less likely to interact. If the use of potentially interacting drugs cannot be avoided, adverse clinical consequences may be minimized, as appropriate, by individualized dose adjustments guided by careful monitoring of clinical response and measurement of serum drug concentrations.

Competing interests: None declared.

Box 2 Antiepileptic drugs and oral anticoagulants

Oral anticoagulants have an important role in the prevention and treatment of thromboembolic disorders. These drugs, however, have a narrow therapeutic window, and interactions affecting their pharmacokinetics or pharmacodynamics may result in serious complications [76].

There is a large body of evidence indicating that older generation enzyme-inducing AEDs, most notably barbiturates and carbamazepine, stimulate the metabolism of warfarin and other coumarin drugs, thereby increasing their dosage requirements [76, 77]. The size of the required increase in anticoagulant dosage varies from one patient to another: for warfarin, it may range from virtually nil to as much as 10-fold [78], and repeated International Normalized Ratio (INR) determinations are necessary to tailor dose adjustments to individual needs. A most dangerous situation comes when barbiturates or carbamazepine are discontinued, or substituted with other AEDs devoid of enzyme-inducing properties. Under these circumstances, the metabolism of the antocoagulant will slow and, if dosage is not reduced, there is a serious risk of massive haemorrhage [79].

Phenytoin's interactions with warfarin are more complex and unpredictable, and may involve a decrease or an increase in anticoagulant activity, or even biphasic responses [76, 77]. As a result of this, achieving a stable level of anticoagulation may be difficult when warfarin and phenytoin are combined [80]. Valproic acid may also increase serum warfarin concentrations, and facilitate bleeding by interfering directly with platelet function and coagulation processes [81].

Of the newer AEDs, oxcarbazepine 900 mg day−1 has not been found to affect the anticoagulant effect of warfarin, but the possibility of interaction at higher dosages has not been tested [2]. Levetiracetam 2000 mg day−1 and tiagabine 12 mg day−1 have also been reported not to interfere with the anticoagulant effect or the pharmacokinetics of warfarin [2].

Overall, newer generation AEDs with low interaction potential appear to be a preferable choice in patients who take oral anticoagulants. In any case, careful monitoring of the INR is recommended whenever comedication is changed in these patients.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Mechanisms of AED interactions
  5. Interactions between AEDs
  6. Interactions between AEDs and other drugs
  7. Conclusions
  8. References
  • 1
    Patsalos PN, Perucca E. Clinically important drug interactions in epilepsy: general features and interactions between antiepileptic drugs. Lancet Neurol 2003; 2: 34756.
  • 2
    Patsalos PN, Perucca E. Clinically important drug interactions in epilepsy: interactions between antiepileptic drugs and other drugs. Lancet Neurol 2003; 2: 47381.
  • 3
    Patsalos PN, Froscher W, Pisani F, Van Rijn C. The importance of drug interactions in epilepsy therapy. Epilepsia 2002; 43: 36585.
  • 4
    Hachad H, Ragueneau-Majlessi I, Levy RH. New antiepileptic drugs: review on drug interactions. Ther Drug Monit 2002; 24: 91103.
  • 5
    Majkowski J, Bourgeois B, Patsalos PN, Mattson RH, eds. Antiepileptic Drugs: Combination Therapy and Interactions. Cambridge: Cambridge University Press 2005.
  • 6
    Perucca E, Hedges A, Makki KA, Ruprah M, Wilson JF, Richens A. A comparative study of the enzyme inducing properties of anticonvulsant drugs in epileptic patients. Br J Clin Pharmacol 1984; 18: 40110.
  • 7
    Strolin Benedetti M, Bani M. Metabolism-based interactions involving oral azole antifungals in humans. Drug Metab Rev 1999; 31: 665717.
  • 8
    Anderson GD. A mechanistic approach to antiepileptic drug interactions. Ann Pharmacother 1998; 32: 55463.
  • 9
    Spina E, Perucca E, Levy RH. Predictability of metabolic antiepileptic drug interactions. In Antiepileptic Drugs: Combination Therapy and Interactions, eds MajkowskiJ, BourgeoisB, PatsalosPN, MattsonRH. Cambridge: Cambridge University Press 2005.
  • 10
    Perucca E. Clinical implications of hepatic microsomal enzyme induction by antiepileptic drugs. Pharmacol Ther 1987; 33: 13944.
  • 11
    Perucca E. The clinical pharmacology and therapeutic use of the new antiepileptic drugs. Fund Clin Pharmacol 2001; 15: 40517.
  • 12
    Crawford P. Interactions between antiepileptic drugs and hormonal contraception. CNS Drugs 2002; 16: 26372.
  • 13
    Sidhu J, Bulsara S, Job S, Philipson R. A bidirectional pharmacokinetic interaction study of lamotrigine and the combined oral contraceptive pill in healthy subjects. Epilepsia 2004; 45 (Suppl. 7): 330.
  • 14
    May TW, Rambeck B, Jurgens U. Serum concentrations of lamotrigine in epileptic patients: the influence of dose and comedication. Ther Drug Monit 1996; 18: 52331.
  • 15
    Baruzzi A, Albani F, Riva R. Oxcarbazepine: pharmacokinetic interactions and their clinical relevance. Epilepsia 1994; 35 (Suppl. 3): S14S19.
  • 16
    Kim RB. Drugs as P-glycoprotein substrates, inhibitors, and inducers. Drug Metab Rev 2002; 34: 4754.
  • 17
    Synold TW, Dussault I, Forman BM. The orphan nuclear receptor SXR coordinatively regulates drug metabolism and efflux. Nat Med 2001; 7: 58490.
  • 18
    Giessmann T, May K, Modess C, Wegner D, Hecker U, Zschiesche M, Dazert P, Grube M, Schroeder E, Warzok R, Cascorbi I, Kroemer HK, Siegmund W. Carbamazepine regulates intestinal P-glycoprotein and multidrug resistance protein MRP2 and influences disposition of talinolol in humans. Clin Pharmacol Ther 2004; 76: 192200.
  • 19
    Perucca E, Hebdige S, Frigo GM, Gatti G, Lecchini S, Crema A. Interaction between phenytoin and valproic acid: plasma protein binding and metabolic effects. Clin Pharmacol Ther 1980; 28: 77989.
  • 20
    Tsanaclis LM, Allen J, Perucca E, Routledge PA, Richens A. Effect of valproate on free plasma phenytoin concentrations. Br J Clin Pharmacol 1984; 18: 1720.
  • 21
    Benet LZ, Hoener BA. Changes in plasma protein binding have little clinical relevance. Clin Pharmacol Ther 2002; 71: 11521.
  • 22
    Perucca E, Levy RH. Combination therapy and drug interactions. In Antiepileptic Drugs, 5th edn, eds LevyRH, MattsonRH, MeldrumBS, PeruccaE. Philadelphia: Lippincott Williams & Wilkins 2002; 96102.
  • 23
    Perucca E, Gatti G, Frigo GM, Crema A, Calzetti S, Visintini D. Disposition of sodium valproate in epileptic patients. Br J Clin Pharmacol 1978; 5: 4959.
  • 24
    May T, Rambeck B. Serum concentrations of valproic acid: influence of dose and comedication. Ther Drug Monit 1985; 7: 38790.
  • 25
    Samara EE, Gustavson LE, El-Shourbagy T, Locke C, Granneman GR, Sommerville KW. Population analysis of the pharmacokinetics of tiagabine in patients with epilepsy. Epilepsia 1998; 39: 86873.
  • 26
    Giaccone M, Bartoli A, Gatti G, Marchiselli R, Pisani F, Latella MA, Perucca E. Effects of enzyme inducing anticonvulsants on ethosuximide pharmacokinetics in epileptic patients. Br J Clin Pharmacol 1996; 41: 5759.
  • 27
    Bartoli A, Guerrini R, Belmonte A, Alessandri MG, Gatti G, Perucca E. The influence of dosage, age, and comedication on steady state plasma lamotrigine concentrations in epileptic children: a prospective study with preliminary assessment of correlation with clinical response. Ther Drug Monit 1997; 19: 25260.
  • 28
    Armijo JA, Bravo J, Cuadrado A, Herranz JL. Lamotrigine serum concentration-to-dose ratio: influence of age and concomitant antiepileptic drugs and dosage implications. Ther Drug Monit 1999; 21: 18290.
  • 29
    Britzi M, Perucca E, Soback S, Levy RH, Fattore C, Crema F, Gatti G, Doose DR, Maryanoff BE, Bialer M. Pharmacokinetic and metabolic investigation of topiramate disposition in healthy subjects in the presence and absence of enzyme induction by carbamazepine. Epilepsia 2005; 47: 17.
  • 30
    Tartara A, Galimberti CA, Manni R, Morini R, Limido G, Gatti G, Bartoli A, Strada G, Perucca E. The pharmacokinetics of oxcarbazepine and its active metabolite 10-hydroxycarbazepine in normal subjects and in epileptic patients treated with phenobarbitone or valproic acid. Br J Clin Pharmacol 1993; 36: 3668.
  • 31
    Ojemann LM, Shastri RA, Wilenski AJ, Friel PN, Levy RH, McLean JR, Buchanan RA. Comparative pharmacokinetics of zonisamide (CI-912) in epileptic patients on carbamazepine or phenytoin monotherapy. Ther Drug Monit 1986; 8: 2936.
  • 32
    Wagner M, Graves NM, Marineau K, Holmes GB, Remmel RP, Leppik IE. Discontinuation of phenytoin and carbamazepine in patients receiving felbamate. Epilepsia 1991; 32: 398406.
  • 33
    Khoo KC, Mendels J, Rothbert M, Garland WA, Colburn WA, Min BH, Lucek R, Carbone JJ, Boxenbaum HG, Kaplan SA. I nfluence of phenytoin and phenobarbital on the disposition of a single oral dose of clonazepam. Clin Pharmacol Ther 1980; 28: 36875.
  • 34
    Sennoune S, Mesdjian E, Bonneton J, Genton P, Dravet C, Roger J. Interactions between clobazam and standard antiepileptic drugs in patients with epilepsy. Ther Drug Monit 1992; 14: 26974.
  • 35
    Christiansen J, Dam M. Influence of phenobarbital and diphenylhydantoin on plasma carbamazepine levels in patients with epilepsy. Acta Neurol Scand 1973; 49: 5436.
  • 36
    Perucca E, Dulac O, Shorvon S, Tomson T. Harnessing the clinical potential of antiepileptic drug therapy: dosage optimisation. CNS Drugs 2001; 15: 60921.
  • 37
    Battino D, Croci D, Granata T, Bernardi G, Monza G. Changes in unbound and total valproic acid concentrations after replacement of carbamazepine with oxcarbazepine. Ther Drug Monit 1992; 14: 3769.
  • 38
    Yuen AWC, Land G, Weatherley B, Peck AW. Sodium valproate inhibits lamotrigine metabolism. Br J Clin Pharmacol 1992; 33: 5113.
  • 39
    Binnie CD, Van Emde Boas W, Kasteleijn Nolste Trenite DG, De Korte RA, Meijer JWA, Meinardi H, Miller AA, Overweg J, Peck AW, Van Wieringeen A, Yuen WC. Acute effects of lamotrigine (BW430C) in persons with epilepsy. Epilepsia 1986; 27: 24854.
  • 40
    Gidal BE, Anderson GD, Rutecki PR, Shaw R, Lanning A. Lack of an effect of valproate concentration on lamotrigine pharmacokinetics in developmentally disabled patients with epilepsy. Epilepsy Res 2000; 42: 2331.
  • 41
    Kapetanovic IM, Kupferberg HJ, Porter RJ, Theodore W, Schulman E, Penry JK. Mechanism of valproate–phenobarbital interaction in epileptic patients. Clin Pharmacol Ther 1981; 29: 4806.
  • 42
    Barcs G, Walker EB, Elger CE, Scaramelli A, Stefan H, Sturm Y, Moore A, Flesch G, Kramer L, D’Souza J. Oxcarbazepine placebo-controlled, dose-ranging trial in refractory partial epilepsy. Epilepsia 2000; 41: 1597607.
  • 43
    Bialer M, Doose DR, Murthy B, Curtin C, Wang SS, Twyman RE, Schwabe S. Pharmacokinetic interactions of topiramate. Clin Pharmacokinet 2004; 12: 76380.
  • 44
    Pisani F, Fazio A, Oteri G, Ruello C, Gitto C, Russo F, Perucca E. Sodium valproate and valpromide: differential interactions with carbamazepine in epileptic patients. Epilepsia 1986; 27: 54852.
  • 45
    Pisani F, Oteri G, Russo MF, Di Perri R, Perucca E, Richens A. The efficacy of valproate–lamotrigine comedication in refractory complex partial seizures: evidence for a pharmacodynamic interaction. Epilepsia 1999; 40: 11416.
  • 46
    Rowan AJ, Meijer JWA, De Beer-Pawlikowski N, Van Der Geest P, Meinardi H. Valproate–ethosuximide combination therapy for refractory absence seizures. Arch Neurol 1983; 40: 797802.
  • 47
    Besag FMC, Berry DJ, Pool F, Newbery JE, Subel B. Carbamazepine toxicity with lamotrigine: pharmacokinetic or pharmacodynamic interaction? Epilepsia 1998; 39: 1837.
  • 48
    Sabers A, Ohman I, Christensen J, Tomson T. Oral contraceptives reduce lamotrigine plasma levels. Neurology 2003; 61: 5701.
  • 49
    Sabers A, Buchholt JM, Uldall P, Hansen EL. Lamotrigine plasma levels reduced by oral contraceptives. Epilepsy Res 2001; 47: 1514.
  • 50
    Bauer LA. Interference of oral phenytoin absorption by continuous nasogastric feeding. Neurology 1982; 32: 5702.
  • 51
    Worden JP Jr, Wood CA Jr, Workman CH. Phenytoin and nasogastric feedings. Neurology 1984; 34: 132.
  • 52
    Vecht CJ, Wagner GL, Wilms EB. Interactions between antiepileptic and chemotherapeutic drugs. Lancet Neurol 2003; 2: 4049.
  • 53
    De Turck BJ, Diltoer MW, Corneli PJ, Cornelis PJ, Maes V, Spapen HD, Camu F, Huyghens LP. Lowering of plasma valproic acid concentrations during concomitant therapy with meropenem and amikacin. J Antimicrob Chemother 1998; 42: 5634.
  • 54
    Sander JW, Perucca E. Epilepsy and comorbidity: infections and antimicrobials usage in relation to epilepsy management. Acta Neurol Scand 2003; 108 (Suppl. 180): 1622.
  • 55
    Von Moltke LL, Greenblatt DJ, Schmider J, Wright CE, Harmatz J, Schader RI. In vitro approaches to predicting drug interactions in vivo. Biochem Pharmacol 1998; 55: 11322.
  • 56
    Pauwels O. Factors contributing to carbamazepine–macrolide interactions. Pharmacol Res 2002; 45: 2918.
  • 57
    Van Rosensteil NA, Adam D. Macrolides antibacterials. Drug interactions of clinical significance. Drug Safety 1995; 13: 10522.
  • 58
    Ducharme MP, Slaughter RL, Warbasse LH, Chandrasekar PH, Van de Velde V, Mannens G, Edwards DJ. Itraconazole and hydroxyitraconazole serum concentrations are reduced more than ten-fold by phenytoin. Clin Pharmacol Ther 1995; 58: 61724.
  • 59
    Bittencourt PRM, Gracia CM, Martins R, Fernandes AG, Diekmann HW, Jung W. Phenytoin and carbamazepine decrease oral bioavailability or praziquantel. Neurology 1992; 42: 4926.
  • 60
    Romanelli F, Jennings HR, Nath A, Ryan M, Berger J. Therapeutic dilemma: the use of anticonvulsants in HIV-positive individuals. Neurology 2000; 54: 14047.
  • 61
    Hugen PW, Burger DM, Brinkman K, Ter Hofstede HJ, Schuurman R, Koopmans PP, Hekster YA. Carbamazepine–indinavir interaction causes antiretroviral therapy failure. Ann Pharmacother 2000; 34: 46570.
  • 62
    Liedtke MD, Lockart SM, Rathbun RC. Anticonvulsant and antiretroviral interactions. Ann Pharmacother 2004; 38: 4829.
  • 63
    Flockart DA, Tanus-Santos JE. Implications of cytochrome P450 interactions when prescribing medication for hypertension. Arch Intern Med 2002; 162: 40512.
  • 64
    Tartara A, Galimberti CA, Manni R, Parietti L, Zucca C, Baasch H, Caresia L, Muck W, Barzaghi N, Gatti G, Perucca E. Differential effects of valproic acid and enzyme-inducing anticonvulsants on nimodipine bioavailability in epileptic patients. Br J Clin Pharmacol 1991; 32: 33540.
  • 65
    Michelucci R, Cipolla G, Passarelli D, Gatti G, Ochan M, Heinig R, Tassinari CA, Perucca E. Reduced plasma nisoldipine concentrations in phenytoin-treated patients with epilepsy. Epilepsia 1996; 37: 110710.
  • 66
    Spina E, Perucca E. Clinical significance of pharmacokinetic interactions between antiepileptic and psychotropic drugs. Epilepsia 2002; 43 (Suppl. 2): 3744.
  • 67
    DeToledo JC, Haddad H, Ramsey SE. Status epilepticus associated with the combination of valproic acid and clomipramine. Ther Drug Monit 1997; 19: 713.
  • 68
    Soriano SG, Martyn JA. Antiepileptic-induced resistance to neuromuscular blockers: mechanisms and clinical significance. Clin Pharmacokinet 2004; 43: 7181.
  • 69
    Krauss GL, Brandt J, Campbell M, Plate C, Summerfield M. Antiepileptic medications and oral contraceptive interactions. A national survey of neurologists and obstetricians. Neurology 1996; 46: 15349.
  • 70
    Crawford P, Lee P. Gender difference in the management of epilepsy—what women are hearing. Seizure 1999; 8: 1359.
  • 71
    Doose DR, Wang SS, Padmanabham M, Schwabe S, Jacobs D, Bialer M. The effect of topiramate or carbamazepine on the pharmacokinetics of an oral contraceptive containing norethisterone and ethinyl estradiol in healthy obese and nonobese female subjects. Epilepsia 2003; 44: 5409.
  • 72
    Rosenfeld WE, Doose DR, Walker SA, Nayak RK. Effect of topiramate on the pharmacokinetics of an oral contraceptive containing norethisterone and ethinyl estradiol in patients with epilepsy. Epilepsia 1997; 38: 31723.
  • 73
    Crawford P, Chadwick DJ, Maryin C, Tjia J, Back DJ, Orme M. The interaction of phenytoin and carbamazepine with combined oral contraceptive steroids. Br J Clin Pharmacol 1990; 30: 8926.
  • 74
    Bounds B, Guillebaud. Observational series on women using the contraceptive Mirena concurrently with anti-epileptic and other enzyme-inducing drugs. J Fam Plann Reprod Health Care 2002; 28: 7880.
  • 75
    Herzog AG, Farina EL, Blum AS. Serum valproate levels with oral contraceptive use. Epilepsia 2005; 46: 9701.
  • 76
    Harder S, Thurmann P. Clinically important drug interactions with anticoagulants. An update. Clin Pharmacokin 1996; 30: 41644.
  • 77
    Cropp JS, Bussey HI. A review of enzyme induction of warfarin metabolism with recommendations for patient management. Pharmacotherapy 1997; 17: 91728.
  • 78
    Breckenridge A. Drug interactions with oral anticoagulants. BMJ 1974; 2: 397400.
  • 79
    McDonald MG, Robinson DS. Clinical observations of possible barbiturate interference with anticoagulation. J Am Med Ass 1968; 204: 959.
  • 80
    Hassan Y, Awaisu A, Aziz NA, Ismail O. The complexity of achieving anticoagulation control in the face of warfarin–phenytoin interaction: an Asian case report. Pharm World Sci 2005; 27: 169.
  • 81
    Stephen LJ. Drug treatment of epilepsy in elderly people: focus on valproic acid. Drugs Aging 2003; 20: 14152.