Fiorenzo Albani PharmD*
Roberto Riva MD**
Agostino Baruzzi MD***
* Research Assistant, Department of Neurological Sciences, University of Bologna, Bologna, Italy
** Senior Researcher, Department of Neurological Sciences, University of Bologna, Bologna, Italy
*** Professor of Neurology, Department of Neurological Sciences, University of Bologna, Bologna, Italy
Antiepileptic drug (AED) interactions are relatively common and represent a frequent clinical problem during epilepsy treatment (1,2). Oxcarbazepine (OXC) is an AED chemically related to carbamazepine (CBZ), but shows a completely different metabolic profile. After oral administration in humans, OXC undergoes rapid and almost quantitative enzymatic reduction to form its active metabolite, 10,11-dihydro-10-hydroxycarbazepine (monohydroxy derivative, MHD). Only minimal amounts of the parent drug are found in peripheral blood (3). Elimination of MHD occurs through direct renal excretion, glucuronidation, and, marginally, hydroxylation to a dihydroxy derivative (DHD) (3). Only the latter reaction depends on microsomal cytochrome P450 (CYP450) enzymes, suggesting that interference with oxidative metabolism, usually the most common source of pharmacokinetic interactions, should be of minor relevance. Studies on the enzyme-inducing properties of OXC treatment, tested in most cases using antipyrine as a probe, have been reviewed previously (4). In brief, OXC seems to have only a modest inducing action, possibly more evident at high doses or related to induction of specific isoforms of CYP450 enzymes (2, 3, 4). Another common source of pharmacokinetic interactions for AEDs is at the level of competitive plasma protein binding (1,2). These interactions, however, seldom are clinically relevant, and in the case of OXC are unlikely to occur to any significant extent because the main active substance (MHD) is only approximately 40% plasma protein bound (3).
To define better the clinical relevance of OXC drug interactions, some reference to other AEDs is made thoroughout this review.
EFFECTS OF OXCARBAZEPINE ON ANTIEPILEPTIC DRUGS
Information about OXC effects on the kinetics of other AEDs derives from a few specific studies and from AED plasma concentration data collected during polytherapy clinical studies. Houtkooper et al. (5) reported that the substitution of OXC for CBZ given in association with valproic acid (VPA) or phenytoin (PHT), at constant dosages, increased VPA and PHT plasma concentrations by 20% to 30%. Battino et al. (6) reported similar data on total and free VPA concentrations in young epileptic patients. Total or partial deinduction is the probable cause of these effects. Therefore, when OXC is substituted for an inducing drug in polytherapy, the clinical status of the patient and the plasma concentrations of other AEDs should be closely monitored.
McKee et al. (7) studied the effect of OXC on the steady-state concentrations of CBZ, VPA, and PHT in 3 groups of 12 chronically treated patients, both after single-dose administration (OXC 600 mg) and after 1-week treatment (OXC 300 mg three times daily), compared with placebo. Median areas under the conentration curve (AUCs) for CBZ, VPA, and PHT during a dosage interval (12 hours) did not differ significantly after treatment with OXC and placebo, although a trend for reduced CBZ AUC and increased PHT AUC was observed.
Analyzing data collected in two large, controlled clinical trials, Hossain et al. (8) reported that plasma concentrations of phenobarbital (PB) increased (+14%) and those of CBZ were reduced (-15%) during OXC coadministration. More clinically relevant is the observation of a mean 40% increase in PHT concentrations for MHD plasma levels above 43 µg/mL (8).
In a series of 14 patients treated with OXC and lamotrigine (LTG), the LTG level-to-dose ratio (LDR) was 33% less with respect to LTG monotherapy (9). Given LTG's pharmacokinetic characteristics (2), an involvement of enzymes responsible for LTG glucuroconjugation may be suspected. For comparison, the LTG LDR in patients cotreated with CBZ was approximately 50% less.
EFFECTS OF OXCARBAZEPINE ON OTHER DRUGS
The interaction of AEDs with oral anticoagulants often is of clinical importance (1). The influence of OXC on the anticoagulant effect of warfarin was investigated in 10 healthy volunteers (10). A dose of 900 mg administered daily for a week, after a warfarin titrating period of 3 weeks, did not significantly modify warfarin action as measured by prothrombin time (mean Quick values 36.6% at baseline, 38.1% after OXC). The authors concluded that when coadministration of warfarin is required, OXC offers a clinical advantage over other AEDs, such as PB, PHT, and CBZ. However, no data in patients are available.
The clinical significance of AED interaction with oral contraceptives also is well known (1,11), and the potential effects of OXC have been investigated. In a preliminary study (12), OXC 900 mg/day added for a month to a stable oral contraceptive regimen (low-dose, triphasic) reduced the bioavailability of ethinylestradiol (EE) and levonorgestrel (LN) by 48% and 32%, respectively, in the 10 healthy women who completed the protocol. These data were confirmed in a crossover, placebo-controlled study in 22 healthy women receiving a high-dose, fixed-ratio combination of EE (50 µg) and LN (250 µg) (13), where AUCS0-24 of EE and LN were decreased by 47% during OXC treatment at 1,200 mg/day. It has been suggested that OXC selectively induces the CYP3A-mediated metabolism responsible for the major EE and LN metabolic pathways (10, 11, 12). In the study by Klosterskov Jensen et al. (12), breakthrough bleeding, a clinical consequence of reduced hormone bioavailability and an indication of diminished contraceptive efficacy, showed an incidence of 15%, which was considerably higher than that observed when the same oral contraceptives were used without associated drugs. Similarly, of six women receiving OXC in combination with an oral low-dose contraceptive containing 30 µg of EE, four had breakthrough bleeding with OXC (14). In 59 patients treated with CBZ, 37 (63%) had the same adverse effect (14). Therefore, the same cautions exercised with enzyme-inducing AEDs possibly apply to OXC as well.
Felodipine, a calcium antagonist, undergoes extensive first-pass oxidative hepatic metabolism and normally has an oral absolute bioavailability of 15%. Drug-inducing AEDs reduce this bioavailability to less than 1% (15). A study performed in eight volunteers, however, reported a 28% relative reduction in felodipine bioavailability (absolute value ~10%), when OXC (900 mg/day) was coadministered for 1 week (16). The clinical relevance of this interaction in patients is not defined.
EFFECTS OF ANTIEPILEPTIC DRUGS ON OXCARBAZEPINE
An early retrospective study showed that enzyme-inducing drugs such as PB and PHT do not induce MHD formation but can increase its oxidative conversion to DHD. This has been considered to be clinically insignificant and to represent a minor pathway in OXC metabolism (17). The ratio between MHD concentrations (milligrams per liter) and the OXC oral doses (milligrams per kilogram), however, was lower in a group of adult patients receiving comedication with enzyme-inducing drugs (mean value, 0.74) compared with patients receiving OXC alone (mean value, 0.94), indicating that some interaction with enzyme-inducing comedication may occur (18). When OXC was combined with VPA, the ratio was 0.93, similar to that found in monotherapy (18).
These preliminary data have been substantially confirmed by specific studies. Tartara et al. (19) reported a comparison of OXC and MHD kinetics after acute administration in normal subjects and in patients treated with other AEDs. Three groups of eight subjects (drug-free healthy control subjects or epileptic patients on chronic treatment with either PB or VPA alone) each received a single oral dose of OXC 600 mg. Plasma concentrations of OXC and MHD were followed for up to 48 hours. The AUCs of OXC and MHD were significantly lower in patients receiving PB than in control subjects, whereas no differences were found between patients receiving VPA and control subjects.
In the study already mentioned, McKee et al. (7) reported the effect of CBZ, VPA, and PHT on OXC pharmacokinetics in 3 groups of 12 patients each, both after single-dose administration (OXC 600 mg) and after 1-week treatment (OXC 300 mg three times daily), compared with placebo. A group of seven otherwise untreated patients served as control, receiving active treatment (OXC) only. MHD AUCs at steady state were lower in PHT- and CBZ-treated patients with respect to control subjects (approximately one-third less). In CBZ-treated patients the difference reached statistical significance (p < .05). Possible explanations were a small induction effect on MHD metabolism or the induction of an alternative pathway of OXC biodegradation. Values for patients on VPA cotherapy were similar to those for control subjects.
In a small series of patients, May et al. (20) reported comparable mean values for MHD LDR after treatment with OXC in monotherapy or with VPA cotreatment. The MHD free fraction during VPA cotherapy was slightly but significantly higher (64% versus 56.7% in monotherapy). Interestingly, two patients receiving methsuximide had the lowest MHD LDR: <50% with respect to monotherapy.
Overall, these data suggest that enzyme-inducing AEDs such as PB, PHT, and CBZ cause only modest modifications of OXC pharmacokinetics in most cases; the interaction, however, may be clinically relevant in some patients. The association with VPA should not require any adjustment of OXC dosages.
The effect of felbamate (FBM) on OXC kinetics was assessed in 18 healthy volunteers (21). Subjects received OXC, 1,200 mg/day, on an open basis in combination with placebo or FBM, 2,400 mg/day, for two 10-day treatments periods. FBM had no significant effects on MHD plasma or urine pharmacokinetics compared with placebo. However, DHD maximum plasma concentration and AUC0-12 values as well as DHD urinary excretion (free and total) were significantly increased, probably as a consequence of the induction of oxidative metabolism of MHD. The authors considered the effects of FBM on OXC pharmacokinetics not clinically relevant.
EFFECTS OF OTHER DRUGS ON OXCARBAZEPINE
Most data on the effects of other, non-AED drugs on OXC derive from preclinical studies in healthy volunteers. Verapamil inhibits the metabolism of CBZ to an extent that can produce clinical manifestations of CBZ neurotoxicity (1). The potential interaction of verapamil and OXC was studied in 10 healthy volunteers (22). After titration of OXC to 900 mg/day, verapamil (240 mg/day) was administered for 1 week. The main result was a 20% decrease of the MHD AUC compared with baseline, an effect that remains to be explained. This interaction, however, is likely to be clinically negligible in most cases. Cimetidine interacts with many drugs (23), inhibiting their metabolism and causing clinical toxicity in some cases. The cimetidine-OXC interaction was studied in eight healthy volunteers: cimetidine treatment 800 mg/day for 1 week did not significantly modify the pharmacokinetics of OXC or MHD (24). Erythromycin is a macrolide antibiotic frequently implicated in clinically relevant pharmacokinetic drug interactions (25). Results of an eight-volunteer study indicated that a week of erythromycin therapy (1,000 mg/day) had no major influence on the pharmacokinetic parameters of OXC and MHD (26).
TABLE 46.1. PHARMACOKINETIC INTERACTIONS OF OXCARBAZEPINE
Dextropropoxyphene can interfere with AEDs (1), but the clinical significance of individual interactions varies. The effects of dextropropoxyphene on the kinetics of OXC and its metabolites were reported in a study in eight patients receiving chronic OXC treatment (27). Plasma concentrations of MHD were not significantly affected by dextropropoxyphene (195 mg/day for 1 week), even if the DHD concentrations were reduced as a possible consequence of inhibited MHD oxidation.
Clinically significant interactions have been described between viloxazine and AEDs (1). In a study in six epileptic patients receiving chronic OXC monotherapy (1,500 ± 465 mg/day), a 10-day add-on treatment with viloxazine did not modify OXC plasma concentrations, whereas MHD concentrations were increased by 11% and DHD concentrations decreased by an average of 31% (28). No adverse effects were reported in any patients. These results suggest that viloxazine coadministration causes a very modest inhibition of MHD oxidation and that viloxazine can be used safely in depressed epileptic patients receiving OXC treatment.
Table 46.1 summarizes the published data on OXC interactions. In general, OXC shows a favorable profile, with a
limited capacity to affect disposition of other AEDs in a clinically relevant fashion. Only the observed reduction of LTG concentrations and, possibly, the increase in PHT concentrations may require a dose adjustment. The effects of OXC on the clinical efficacy of other drugs should be a minor problem except for the possibly reduced efficacy of oral contraceptives.
On the other hand, the substitution of an inducing agent with OXC should be carefully monitored because of potentially significant deinduction, and the oral doses of associated drugs may need to be adjusted.
Plasma concentrations of OXC are moderately affected by enzyme-inducing AEDS like PB, PHT, and CBZ, and no relevant interactions were observed with VPA and FBM. The modifications of OXC kinetics caused by other drugs seem of little, if any, clinical significance.