SEARCH

SEARCH BY CITATION

Keywords:

  • Levetiracetam, Antiepileptic drugs, Drug interactions, Epilepsy, Animal models.

Summary

  1. Top of page
  2. Summary
  3. Interactions of Levetiracetam with Anticonvulsants in the Audiogenic Seizure Model
  4. Interactions of Levetiracetam with Anticonvulsants in the MES and PTZ Seizure Models
  5. Interactions of Levetiracetam with Anticonvulsants in the Rat Amygdala Kindling Model
  6. Interactions of Levetiracetam with Anticonvulsants in Other Seizure Models
  7. Discussion
  8. Acknowledgments
  9. References

Levetiracetam (Keppra) is an antiepileptic drug (AED) characterized by a novel mechanism of action, unique profile of activity in seizure models, and broad-spectrum clinical efficacy. The present report critically reviews several preclinical studies focused on combination therapy with levetiracetam and other anticonvulsants in various seizure and epilepsy models. Administration of levetiracetam together with many different clinically used AEDs or other anticonvulsants generally enhances their protective activity and, among existing AEDs, this was particularly prevalent with valproate. The protective activity of other AEDs was also enhanced by levetiracetam, which seems to be a universal finding that is independent of seizure model or drug combination studied. However, particularly strong enhancement was observed when levetiracetam was combined with agents either enhancing GABAergic or reducing glutamatergic neurotransmission. Importantly, these combinations were not associated with exacerbation of side effects or pharmacokinetic interactions. Based on the available preclinical data, it appears that combination treatment with levetiracetam and other anticonvulsants provides additional therapeutic benefit that may be attributed to its novel and distinct mechanism of action. Moreover, combinations of levetiracetam with clinically used AEDs that enhance GABAergic inhibition may be considered for rational polytherapy, which is often necessary in drug-resistant patients.

The issue of clinical utility of monotherapy versus polytherapy for epilepsy has been raised over 25 years ago (Reynolds & Shorvon, 1981). At that time, however, very few antiepileptic drugs (AEDs) were available. During the last two decades, 13 new AEDs have been introduced, and several other compounds are in late stages of clinical development (Rogawski, 2006; Perucca et al., 2007). As a consequence, neurologists are now equipped with an unprecedented number of AEDs, and the issue has become far more complex (Kwan & Brodie, 2006).

Monotherapy remains the treatment of choice for newly diagnosed epilepsy, and polytherapy with more than one AED is considered only after at least one failure of a monotherapy regimen. This therapeutic strategy is driven by concerns of excessive drug load and increased toxicity. Thus, combinations of AEDs should be carefully selected based on potential for synergy that is not associated with unfavorable pharmacokinetic interactions and toxicity. Such a methodical approach to selection of AED combinations can be called rational polytherapy (Kwan & Brodie, 2006). Unfortunately clinical evidence on how and when to combine AEDs is still very limited, thus preclinical studies could offer some experimental guidelines as to which AED combinations appear to be therapeutically advantageous. In fact, a large number of studies have concentrated on AED interactions in preclinical models of epilepsy (for recent reviews see Deckers et al., 2000 and Jonker et al., 2007). Despite this extensive effort, it is still very difficult to predict which AED combinations could be most efficient in clinics. This is partly due to the fact that a majority of the currently available AEDs affect multiple neurotransmitter systems or ion channels, and it is virtually impossible to delineate which cellular mechanism could be responsible for the synergistic effect. Nevertheless, preclinical studies may still provide valuable guidance for rational polytherapy in patients with epilepsy (Löscher & Wauquier, 1996; Czuczwar & Borowicz, 2002). Studies involving pharmacological probes with well-defined selectivity towards specific neuronal targets may potentially offer better understanding of synergistic mechanisms translating into better seizure protection. Results of such studies can even guide the development of therapeutics with dual mechanism combined in one molecule (Löscher & Wauquier, 1996; Löscher, 1998).

Levetiracetam (Keppra) is an AED with a unique profile of activity in preclinical models of epilepsy. It is ineffective in standard models that have been traditionally used for AED discovery, namely the maximal electroshock (MES) and pentylenetetrazol (PTZ) tests (Klitgaard et al., 1998; Klitgaard, 2001). However, it displays protective actions against audiogenic and kindled seizures (Klitgaard et al., 1998; Löscher et al., 1998), 6-Hz electrical seizures (Barton et al., 2001), and spike-wave discharges in genetic models of absence epilepsy (Gower et al., 1995; Bouwman & van Rijn, 2004). This preclinical profile correlates well with broad-spectrum clinical efficacy of levetiracetam (Ben Menachem & Falter, 2000; Shorvon et al., 2000; Berkovic et al., 2007; Brodie et al., 2007; Noachtar et al., 2008). Another unique feature of levetiracetam is its distinctive binding site, which recently has been characterized as the synaptic vesicle protein 2A (SV2A) (Lynch et al., 2004). Molecular mechanisms by which this binding protects from seizures are not well understood at present, but the existence of a strong functional correlation between SV2A binding affinity and anticonvulsant potency is a clear indicator of a mechanistic link between the two (Lynch et al., 2004). However, several other mechanisms that may contribute to the anticonvulsant effects of levetiracetam have also been identified. For example, levetiracetam opposes inhibitory effects of Zn2+ and β-carbolines on GABA- and glycine-gated currents (Rigo et al., 2002). Furthermore, significant inhibitory effects of levetiracetam on voltage-gated Ca2+ and K+ channels were reported (Niespodziany et al., 2001; Madeja et al., 2003). Finally, inhibitory effects of levetiracetam on intracellular Ca2+ release have also been reported (Angehagen et al., 2003; Nagarkatti et al., 2008).

Levetiracetam has recently been approved in the European Union (EU) for monotherapy of partial onset seizures, with or without secondary generalization, in patients from 16 years or older with newly diagnosed epilepsy; however it is frequently used in combination with other AEDs, which is often necessary in drug-resistant patients (De Smedt et al., 2007). Therefore, it is important to understand what the effects of levetiracetam in combination with these drugs are. In fact, several studies addressed this issue in preclinical models of epilepsy (Mazarati et al., 2004; Sills et al., 2004; Luszczki et al., 2005; Luszczki et al., 2006; Donato Di Paola et al., 2007; Luszczki et al., 2007). The results of these studies and data previously published only in abstract form (Matagne et al., 2001; Klitgaard & Matagne, 2002) together with some unpublished results from our group will be discussed in the present review.

Interactions of Levetiracetam with Anticonvulsants in the Audiogenic Seizure Model

  1. Top of page
  2. Summary
  3. Interactions of Levetiracetam with Anticonvulsants in the Audiogenic Seizure Model
  4. Interactions of Levetiracetam with Anticonvulsants in the MES and PTZ Seizure Models
  5. Interactions of Levetiracetam with Anticonvulsants in the Rat Amygdala Kindling Model
  6. Interactions of Levetiracetam with Anticonvulsants in Other Seizure Models
  7. Discussion
  8. Acknowledgments
  9. References

Levetiracetam has been reported to potentiate the protective effects of several clinically used AEDs and other anticonvulsant compounds against audiogenic seizures in mice (Matagne et al., 2001; Donato Di Paola et al., 2007). Levetiracetam displays very good efficacy in this model with an effective dose of 50% (ED50) value against clonic convulsions equal to 9.7 mg/kg following intraperitoneal (i.p.) administration (Gower et al., 1992). However, when a low dose of levetiracetam (5.5 mg/kg), which is equivalent to the effective dose of 25% (ED25) value against clonic convulsions, is combined with other anticonvulsants, their potency against audiogenic seizures is substantially increased (Table 1). In particular, the anticonvulsant potency of drugs enhancing GABAergic inhibition (i.e., benzodiazepines, phenobarbital, chlordiazepoxide, and valproate) was most effectively enhanced by levetiracetam. The potency of these drugs was increased by more than 16-fold (Table 1). The potency of other drugs augmenting GABAergic neurotransmission [i.e., bretazenil (partial agonist of the benzodiazepine receptor), allopregnanolone (positive allosteric modulator of GABAA receptors), NO-711 (GABA transporter inhibitor), or vigabatrin (GABA transaminase inhibitor)] was also considerably increased (up to 11-fold increase) (Table 1). Likewise, the protective effects of glutamate receptor antagonists [i.e., MK-801 (N-methyl-D-aspartate [NMDA] receptor antagonist) and NBQX (AMPA/kainate receptor antagonist)] were also markedly enhanced in combination with levetiracetam; up to 19-fold increase in potency (Table 1). In contrast, the potency of several drugs inhibiting sodium channels (i.e., carbamazepine or phenytoin) was enhanced to much lesser degree (Table 1). The smallest degree of potentiation was observed with the calcium channel inhibitor (flunarizine) and β-adrenergic receptor blocker (propranolol) (Table 1).

Table 1.   Effects of anticonvulsants administered alone and in combination with LEV in the mouse audiogenic seizure model
Name of thePretreatmentED50A (mg/kg)bED50B (mg/kg)cChange in potency
compound timea (min)VEH plus CompoundLEVd plus CompoundED50A/ED50B
  1. Audiogenic seizures were induced in genetically sound susceptible mice (Animal Husbandry Unit, UCB, Belgium) with 90-dB, 10- to 20-kHz acoustic stimulus applied for 30 s. Each experimental group consisted of 10 mice that responded positively in the preselection testing performed 24 h before the experiment.

  2. aAll compounds were administered i.p.

  3. bED50A, dose of an anticonvulsant that was required to protect 50% animals against clonic seizures induced by audiogenic stimulation; 95% confidence intervals in parenthesis.

  4. cED50B, dose of an anticonvulsant in combination with levetiracetam that was required to protect 50% animals against clonic seizures induced by audiogenic stimulation; 95% confidence intervals in parenthesis.

  5. dLevetiracetam (LEV) was administered at the dose of 5.5 mg/kg i.p. 60 min prior to testing.

  6. eReported only in the abstract form (Matagne et  al., 2001).

  7. fPreviously unpublished.

Valproate30121 (110–144)4.3 (1.8–9.7)28e
Clonazepam300.036 (0.033–0.039)0.0016 (0.0007–0.0031)23e
Diazepam300.33 (0.31–0.35)0.017 (0.0004–0.8)19f
NBQX1527.9 (18.6–41.7)1.5 (0.68–3.31)19f
MK-801300.17 (0.15–0.2)0.01 (0.0004–0.28)17f
Phenobarbital309.6 (6.8–12.1)0.6 (0.2–1.3)16e
Chlordiazepoxide302.9 (2.2–3.8)0.18 (0.11–0.31)16f
Bretazenil300.19 (0.17–0.21)0.017 (0.008–0.012)11f
NO-711302.5 (2.1–3.1)0.5 (0.22–1.22)5f
Lamotrigine3016.8 (14.3–19.7)4.1 (2.0–8.7)4.1f
Allopregnanolone106.3 (5.8–6.9)1.7 (0.9–5.5)3.7f
Carbamazepine3021.2 (13.3–28.4)5.9 (3.9–8.1)3.6e
Vigabatrin240 1367 (1331–1404)490 (409–587)2.8f
Phenytoin3025.7 (19.6–32.8)13.2 (9.3–16.5)1.9e
Propranolol3019.9 (18.5–21.5)11.6 (9.8–13.6)1.7f
Flunarizine60132 (118–147) 77.5 (48.1–124.7)1.7f

Donato Di Paola et al. (2007) used a very similar experimental design in the audiogenic seizure model and reported synergistic interactions of levetiracetam with felbamate, gabapentin, and topiramate (see Table 2). Interestingly, the enhancing effect of levetiracetam was also less evident in combination with sodium channel inhibitors, i.e., phenytoin (Donato Di Paola et al., 2007).

Table 5.   Effects of AEDs administered alone and in combination with LEV in the rat amygdala kindling model
 ED50A (mg/kg)aED50B (mg/kg)bChange in
 VEH plusLEVc pluspotency
AEDscompoundcompoundED50A/ED50B
  1. aED50A, dose of an anticonvulsant that was required to protect 50% animals against generalized seizures; 95% confidence intervals in parenthesis.

  2. bED50B, dose of an anticonvulsant in combination with LEV that was required to protect 50% animals against generalized seizures; 95% confidence intervals in parenthesis.

  3. cLevetiracetam (LEV) was administered at the dose of 17 mg/kg i.p.

  4. dAll data have been previously reported only in the abstract form (Klitgaard & Matagne, 2002).

Clonazepam0.17 (0.14–0.19)0.04 (0.03–0.06)4.2d
Valproate191 (163–225)64.2 (62.8–65.8)3.0d
Carbamazepine68.6 (64.7–72.6)27.2 (14.6–50.6)2.5d
Phenobarbital33.9 (30.4–37.8)16.1 (12.8–20.1)2.1d

It is important to underline that the potentiation of the anticonvulsant response of AEDs by levetiracetam was not associated with increased adverse effects (i.e., motor impairment) as assessed by the rotarod test (Matagne et al., 2001; Donato Di Paola et al., 2007). To the contrary, the therapeutic index, defined as the ratio between toxic dose of 50% (TD50) and ED50, was dramatically increased in combinations of levetiracetam with valproate, clonazepam, and phenobarbital (Fig. 1) (Matagne et al., 2001). The therapeutic index was also considerably higher in drug combinations for which potentiation of the anticonvulsant effects by levetiracetam has been observed (Donato Di Paola et al., 2007).

image

Figure 1.   Effects of combination therapy with levetiracetam and AEDs on the therapeutic index assessed in the rotarod test in mice. Note: Therapeutic index is the ratio between TD50 value (a dose of an anticonvulsant that produces motor impairment in 50% of animals) and ED50 value (a dose of an anticonvulsant that was required to protect 50% animals against audiogenic seizures). All data have been previously reported only in the abstract form (Matagne et  al., 2001).

Download figure to PowerPoint

Furthermore, plasma and brain concentrations of AEDs were not affected by concomitant administration of levetiracetam (Matagne et al., 2001; Donato Di Paola et al., 2007). The only exception was valproate, but its plasma and brain concentrations were actually reduced in combined treatment with levetiracetam (Matagne et al., 2001) (Table 3). Finally, levetiracetam's concentration in the brain and plasma remained unchanged when it was administered together with other AEDs (Table 4) (Matagne et al., 2001). This appears to exclude the existence of pharmacokinetic interaction responsible for the enhanced anticonvulsant effects in the combination experiments.

Table 2.   Interactions of LEV with 11 clinically used AEDs in five seizures models
 Seizure model  
Mechanism ofAudiogenic  Amygdala AdverseIncrease of AED plasma
action/AEDsmicePTZMESkindlingSSSEeffectsor brain concentration
  1. +, potentiation by levetiracetam (subthreshold method); (+), additive effect of levetiracetam (isobolographic method); [+], synergy with levetiracetam (isobolographic method); –, lack of interaction with levetiracetam; n.t., not tested.

  2. aPreviously reported only in the abstract form (Matagne et  al., 2001).

  3. bPresent report.

  4. hPreviously reported only in the abstract form (Klitgaard & Matagne, 2002).

GABA-mediated
 Clonazepam +an.t.n.t.+hn.t.aa
 Diazepam +b,cn.t.n.t.n.t.+icc
 Phenobarbital+acn.t.(+)e+hn.t.a,c,ea,c,e
Sodium channel
 Phenytoin +acn.t.(+)en.t.n.t.a,c,ea,c,e
 Carbamazepine +a,cn.t.[+]e+hn.t.c,e+ga,c,e
 Oxcarbazepinen.t.n.t.[+]en.t.n.tee
 Lamotrigine+bcn.t.(+)en.t.n.t.c,ec,e
Mixed mechanisms
 Valproate +a,cn.t.(+)e+hn.t.a,c,ea,c,e
 Felbamate+cn.t.[+]fn.t.n.t.c,fc,f
 Topiramate+c+d[+]en.t.n.t.ce+gc,g
 Gabapentin+cn.t.n.t.n.t.n.t.cc
Table 3.   Influence of LEV on plasma and brain concentrations of AEDs
 AED concentrations 
TreatmentPlasma (μg/ml)Brain (μg/g)Brain/plasma ratio
  1. Values given are means ± SD (n = 7–10 per group). Levetiracetam (LEV; 5.4 mg/kg) and AEDs were administered orally 60 min prior to testing. n.d., level below detection limit; *p < 0.05. All data have been previously reported only in the abstract form (Matagne et  al., 2001).

Valproate (166 mg/kg) plus vehicle222.5 ± 44.729.1 ± 8.20.13 ± 0.02
Valproate plus LEV149.3 ± 32.4*21.1 ± 6.6*0.14 ± 0.03
Clonazepam (0.018 mg/kg) plus vehiclen.d.n.d.
Clonazepam plus LEVn.d.n.d.
Phenobarbital (2.54 mg/kg) plus vehicle1.8 ± 0.311.6 ± 1.06.76 ± 1.35
Phenobarbital plus LEV1.8 ± 0.312.6 ± 1.56.93 ± 0.79
Phenytoin (2.52 mg/kg) plus vehicle1.7 ± 0.81.2 ± 0.30.82 ± 0.33
Phenytoin plus LEV1.7 ± 0.71.3 ± 0.30.88 ± 0.36
Carbamazepine (5.9 mg/kg) plus vehicle1.3 ± 0.20.6 ± 0.20.44 ± 0.15
Carbamazepine plus LEV1.3 ± 0.40.7 ± 0.20.58 ± 0.17

Interactions of Levetiracetam with Anticonvulsants in the MES and PTZ Seizure Models

  1. Top of page
  2. Summary
  3. Interactions of Levetiracetam with Anticonvulsants in the Audiogenic Seizure Model
  4. Interactions of Levetiracetam with Anticonvulsants in the MES and PTZ Seizure Models
  5. Interactions of Levetiracetam with Anticonvulsants in the Rat Amygdala Kindling Model
  6. Interactions of Levetiracetam with Anticonvulsants in Other Seizure Models
  7. Discussion
  8. Acknowledgments
  9. References

Levetiracetam is not effective against MES (Klitgaard et al., 1998) and only raises the seizure threshold in this model (Löscher & Honack, 1993). Nevertheless, interactions between levetiracetam and several other AEDs have been studied in the MES model with the use of isobolography (Luszczki et al., 2006, 2007). This method of studying drug interactions, which are administrated at several fixed dose ratios, allows determination of synergistic, antagonistic, or additive effects of drug combinations (Tallarida, 2001). In fact, isobolographic methods are increasingly used for studying interactions of AEDs in preclinical models of epilepsy, because they appear to have more predictive validity than traditional subthreshold methods involving dose-response analysis (Luszczki & Czuczwar, 2003; Jonker et al., 2007). Levetiracetam administered at several fixed dose ratios together with topiramate (1:2, 1:1, 2:1, and 4:1), carbamazepine (16:1), or oxcarbazepine (8:1 and 16:1) displayed synergistic (supraadditive) effects, while combinations with phenytoin, lamotrigine, phenobarbital, or valproate produced additive effects in the MES

model (Luszczki et al., 2006). It should also be noted that levetiracetam in these drug combinations neither altered the motor and memory performance of mice nor altered the concentrations of AEDs in the brain. Finally, brain concentrations of levetiracetam remained unchanged when it was administered with the above listed AEDs (Luszczki et al., 2006). In contrast, supraadditive combinations with felbamate (1:2, 1:1, 2:1, and 4:1) in the MES model were associated with increased brain concentrations of levetiracetam, but again no potentiation of motor impairment was observed (Luszczki et al., 2007). However, significant potentiation of the motor impairment induced by topiramate and carbamazepine was observed when levetiracetam (150 mg/kg; a dose increasing the threshold for electrically induced seizures by 50%) was used in combination with high doses of these AEDs (Luszczki et al., 2005). It is noteworthy that the motor coordination impairment produced by the other AEDs (i.e., phenytoin, phenobarbital, valproate, lamotrigine, oxcarbazepine, and felbamate) was not affected by addition of levetiracetam (Luszczki et al., 2005).

Similarly to the MES model, levetiracetam is also not effective against seizures induced by supramaximal doses of PTZ (Klitgaard et al., 1998), but raises the threshold for this chemoconvulsant (Löscher & Honack, 1993). Probably for that reason, comprehensive studies of AED combinations with levetiracetam have not been performed in this model. However, Sills et al. (2004) observed an interesting result in a study that evaluated pharmacodynamic interactions of topiramate with several other AEDs (including levetiracetam) in the PTZ model. Topiramate, like levetiracetam, did not protect mice against PTZ-induced seizures, but when these two drugs were administered in combination, significant protection was achieved (Sills et al., 2004).

Interactions of Levetiracetam with Anticonvulsants in the Rat Amygdala Kindling Model

  1. Top of page
  2. Summary
  3. Interactions of Levetiracetam with Anticonvulsants in the Audiogenic Seizure Model
  4. Interactions of Levetiracetam with Anticonvulsants in the MES and PTZ Seizure Models
  5. Interactions of Levetiracetam with Anticonvulsants in the Rat Amygdala Kindling Model
  6. Interactions of Levetiracetam with Anticonvulsants in Other Seizure Models
  7. Discussion
  8. Acknowledgments
  9. References

Levetiracetam shows very good efficacy in kindling models, particularly against amygdala kindling in rats (Klitgaard et al., 1998; Löscher et al., 1998). Consequently, the interactions between levetiracetam and other AEDs have been studied in this model (Klitgaard & Matagne, 2002). Valproate, clonazepam, carbamazepine, and phenobarbital displayed dose-dependent protection against secondarily generalized motor seizures (Fig. 2A–2D).

image

Figure 2.   Effects of AEDs administered alone and in combination with levetiracetam (LEV) in the rat amygdala kindling model. Note: All data have been previously reported only in the abstract form (Klitgaard et  al., 2002).

Download figure to PowerPoint

Coadministration of an ineffective dose of levetiracetam (17 mg/kg) afforded up to four-fold increase in their  potency (Table 5) and leftward shifts of the dose-response curves (Fig. 2). Higher dose of levetiracetam (108 mg/kg) enhanced the potency of the above mentioned AEDs even further (i.e., up to nine-fold potency increase in the case of valproate) (Klitgaard & Matagne, 2002).

Table 4.   Influence of AEDs on plasma and brain concentrations of LEV
 LEV concentrations 
TreatmentPlasma (μg/ml)Brain (μg/g)Brain/plasma ratio
  1. Values given are means ± SD (n = 7–10 per group). Levetiracetam (LEV; 5.4 mg/kg) and AEDs were administered orally 60 min prior to testing. All data have been previously reported only in the abstract form (Matagne et  al., 2001).

LEV plus vehicle4.9 ± 0.52.1 ± 0.30.43 ± 0.03
LEV plus valproate (166 mg/kg)4.6 ± 0.31.9 ± 0.20.43 ± 0.03
LEV plus vehicle4.2 ± 0.62.4 ± 0.70.57 ± 0.13
LEV plus clonazepam (0.018 mg/kg)4.3 ± 0.42.4 ± 0.50.54 ± 0.08
LEV plus vehicle4.8 ± 0.52.3 ± 0.20.49 ± 0.05
LEV plus phenobarbital (2.54 mg/kg)5.0 ± 0.42.5 ± 0.40.52 ± 0.05
LEV plus vehicle4.0 ± 0.72.4 ± 0.40.59 ± 0.07
LEV plus phenytoin (2.52 mg/kg)4.2 ± 0.62.4 ± 0.30.58 ± 0.09
LEV plus vehicle4.2 ± 0.42.3 ± 0.30.54 ± 0.07
LEV plus carbamazepine (5.9 mg/kg)4.2 ± 0.42.3 ± 0.30.54 ± 0.05

Interactions of Levetiracetam with Anticonvulsants in Other Seizure Models

  1. Top of page
  2. Summary
  3. Interactions of Levetiracetam with Anticonvulsants in the Audiogenic Seizure Model
  4. Interactions of Levetiracetam with Anticonvulsants in the MES and PTZ Seizure Models
  5. Interactions of Levetiracetam with Anticonvulsants in the Rat Amygdala Kindling Model
  6. Interactions of Levetiracetam with Anticonvulsants in Other Seizure Models
  7. Discussion
  8. Acknowledgments
  9. References

Levetiracetam is highly active in an experimental model of self-sustained status epilepticus (SSSE) (Mazarati et al., 2004). It is important to note that it was more efficacious than diazepam, which becomes ineffective after prolonged seizures in this model. Unlike diazepam, levetiracetam treatment caused permanent inhibition of SSSE without late recurrence of seizure activity (Mazarati et al., 2004). Interestingly, very strong potentiation of diazepam's protective activity against SSSE was observed in combination with levetiracetam (Mazarati et al., 2004). For example, when levetiracetam (50 and 200 mg/kg) was administered together with diazepam at 1 mg/kg (a dose that is 10 times lower than that required to attenuate seizures), nearly complete protection against SSSE was achieved (Mazarati et al., 2004).

Levetiracetam has shown protective activity against seizures induced by 6 Hz electrical stimulation (Barton et al., 2001). This finding prompted many researchers to introduce the model to their standard experimental repertoire, and now it is commonly used to characterize putative anticonvulsant compounds. For example lacosamide, similarly to levetiracetam, is very active in this model (Stöhr et al., 2007). Interestingly, combination of these two AEDs produces very robust synergistic effects in the 6-Hz model, which is not associated with augmentation of adverse effects (Beyreuther et al., 2007). Combinations of other AEDs with levetiracetam in the 6-Hz model have not been reported so far.

Discussion

  1. Top of page
  2. Summary
  3. Interactions of Levetiracetam with Anticonvulsants in the Audiogenic Seizure Model
  4. Interactions of Levetiracetam with Anticonvulsants in the MES and PTZ Seizure Models
  5. Interactions of Levetiracetam with Anticonvulsants in the Rat Amygdala Kindling Model
  6. Interactions of Levetiracetam with Anticonvulsants in Other Seizure Models
  7. Discussion
  8. Acknowledgments
  9. References

Levetiracetam possesses a completely novel mechanism of action that has not been described for any other AED. It is postulated that binding of the drug to SV2A protein represents its primary mode of action leading to a unique profile in preclinical models (Klitgaard et al., 1998; Lynch et al., 2004). This emphasizes the interest in understanding how that novel mechanism of action influences the  anticonvulsant activity of other AEDs acting on more traditional targets.

There has been a debate whether combining AEDs with different mechanisms of action may be more beneficial than targeting the same cellular mechanisms by two different drugs (Deckers et al., 2000; Jonker et al., 2007). Although one can find examples supporting both scenarios, it appears that, overall, combining two drugs with similar mechanisms of action can provide greater therapeutic benefit than that of two drugs acting on different mechanisms (Klitgaard et al., 1993; Löscher et al., 1993; Löscher & Honack, 1994; Gasior et al., 1997). Interestingly, combinations of drugs acting on the GABAergic system appear to be particularly beneficial in this regard. For example, combination of a GABA uptake inhibitor together with diazepam (allosteric modulator of the GABAA receptor) provided marked synergy in the audiogenic seizure model, while combinations of the same GABAergic drugs with glutamate receptor antagonists remained without any additional therapeutic effect (Klitgaard et al., 1993). Even more pronounced potentiation of the anticonvulsant effects of diazepam was observed when it was administered together with other allosteric modulators of GABAA receptors in the PTZ model (Gasior et al., 1997). In fact, several neurosteroids and phenobarbital increased 25- to 50-fold the potency of diazepam in this model (Gasior et al., 1997). The degree of potentiation observed in that particular study was truly remarkable and amazingly within the same range as observed in some combinations of levetiracetam with GABAergic drugs reviewed in the present report; for example clonazepam or diazepam (Table 1).

It has been well established that SV2A protein is levetiracetam's specific binding site and that the drug does not have any affinity for GABAA receptors (Noyer et al., 1995). However, the functional consequences of levetiracetam binding to SV2A, or the role of SV2A itself, are still not fully understood. It has been shown that SV2A protein modulates exocytosis of transmitter-containing vesicles (Crowder et al., 1999; Xu & Bajjalieh, 2001). Moreover, mice lacking SV2A are characterized by a decrease in the calcium-dependent exocytotic burst, which is a measure of the availability of neurotransmitter vesicles ready to release their content (Xu & Bajjalieh, 2001). Finally, the lack of SV2A results in a decreased action potential-dependent neurotransmission, while action potential-independent neurotransmission remains normal (Crowder et al., 1999; Janz et al., 1999). Thus, it is conceivable that levetiracetam, under certain conditions, may actually facilitate GABAergic neurotransmission. This putative effect has not been demonstrated experimentally, but several studies suggested possible links between levetiracetam and the GABAergic system (Löscher et al., 1996; Poulain & Margineanu, 2002; Rigo et al., 2002; Palma et al., 2007).

Even though the anticonvulsant activity of all compounds promoting GABAergic neurotransmission has been enhanced by levetiracetam, the degree of potentiation was not equal for all of them (Table 1). This is perhaps best exemplified by the difference in potency shift between benzodiazepines (clonazepam and diazepam; approximately 20-fold shift), barbiturates (phenobarbital; approximately 16-fold shift) and neurosteroids (allopregnanolone; approximately 4-fold shift) (Table 1). This difference is even more intriguing since all these compounds are considered positive allosteric modulators of GABAA receptors (Sieghart, 2006). Although they have different binding sites on the GABAA receptor, their allosteric effects, in general, are somewhat similar (Sieghart, 2006). There is, however, at least one major difference between these compounds. Neurosteroids, in contrast to barbiturates or benzodiazepines, seem to preferentially act on extrasynaptic GABAA receptors containing the delta subunit (Mihalek et al., 1999; Stell et al., 2003). If we consider that levetiracetam most likely has a presynaptic action on neurotransmitter release, due to presynaptic location of SV2A protein, then its effects should be more significant for synaptic rather than extrasynaptic receptors. On this basis, the anticonvulsant effects of allosteric GABAA modulators could be differentially enhanced by levetiracetam. Another argument supporting this hypothesis could be the relatively weaker potentiation of anticonvulsant effects of compounds increasing ambient GABA concentration in the brain, i.e., NO-711 (GABA transporter inhibitor) or vigabatrin (GABA transaminase inhibitor) (Table 1). These compounds also seem to exert their anticonvulsant action mainly due to enhancement of so-called “tonic inhibition” that is mediated via extrasynaptic GABAA receptors (Wu et al., 2001; Nusser & Mody, 2002). Thus, it might be suggested that levetiracetam seems to preferentially enhance only synaptic responses of GABAergic compounds, leaving extrasynaptic neurotransmission intact.

The anticonvulsant activity of excitatory amino acid receptors antagonists (i.e., MK-801 and NBQX) in the audiogenic seizure model is also strongly enhanced by levetiracetam. This observation may be related to the role of SV2A and its effects on excitatory neurotransmission (Janz et al., 1999). In fact, significant reduction of stimulation-evoked excitatory postsynaptic currents (EPSCs) in cultured hippocampal neurons from SV2A knockout mice has been reported (Custer et al., 2006). Interestingly, the neurotransmitter release machinery, in which SV2A is involved, may operate differentially in GABAergic and glutamatergic synapses. This concept is supported by a series of recent studies indicating major differences in neurotransmitter release properties of GABAergic and glutamatergic presynaptic terminals (Moulder & Mennerick, 2005, 2006; Moulder et al., 2007). Consequently, levetiracetam may modulate presynaptic glutamate release, which perhaps underlies the strong potentiation of the anticonvulsant effects of glutamate receptor antagonists (Table 1). It would also be consistent with the fact that combination therapy with different classes of glutamate receptor antagonists robustly potentiates their anticonvulsant effects (Löscher, 1998). These results are also in line with the observation that glutamate receptor antagonists are generally very likely to yield synergistic effects in combination with a whole range of AEDs (Löscher, 1998). For example, an antagonist of the NMDA receptor, MK-801, significantly enhanced the anticonvulsant effects of several AEDs in the MES model (Urbanska et al., 1991). Furthermore, the antagonist of non-NMDA receptors, NBQX, also produced similar effects (Zarnowski et al., 1993).

Although levetiracetam is characterized by a binding site distinct from other AEDs (i.e., SV2A protein), it is not fully ascertained what is the contribution of other mechanisms of action to its anticonvulsant activity and thereby potential for pharmacodynamic interactions with AEDs (see above). In fact, levetiracetam's effects on voltage activated Ca2+ and K+ channels or intracellular Ca2+ release may be at least partially responsible for the observed interaction with several AEDs, which are also likely to have more than one mode of action including effects on various ion channels. Precise mechanisms underlying interactions of levetiracetam with these AEDs cannot be fully understood simply on the basis of the data from in vivo seizure models. Thus, such results need to be interpreted with caution and need further experimental scrutiny.

The effects of combined treatments with levetiracetam and other anticonvulsants characterized by well-defined mechanisms of action form an interesting pattern (Table 1). It is clear that levetiracetam strongly enhances the anticonvulsant effects of compounds affecting either GABAergic or glutamatergic neurotransmission, while the effects of agents acting on ion channels were the least enhanced in the audiogenic seizure model (Table 1). This pattern may be yet another indication of levetiracetam's unique mechanism of action on synaptic neurotransmission associated with its binding to SV2A protein. Audiogenic seizures may be particularly sensitive to SV2A ligands, since very strong correlation between binding affinity and anticonvulsant potency has been described in this model (Lynch et al., 2004). However, when interactions of levetiracetam with AEDs were tested in other seizure models (i.e., PTZ and MES), the pattern is not that apparent. Because SV2A ligands are generally inactive in these models, their affinity-potency correlation cannot be determined. Limited protective activity of levetiracetam, affecting only threshold for these seizures, may be a reflection of other modes of action (Niespodziany et al., 2001; Rigo et al., 2002; Madeja et al., 2003). Recently, we have reported that the affinity-potency correlation of SV2A, originally reported for audiogenic seizures, can be also observed in corneally  kindled mice and rats with genetic absence epilepsy (Kaminski et al., 2008). Thus, it would be interesting to establish whether a similar pattern to that observed in the audiogenic seizure model (i.e., strong potentiation of GABAergic and antiglutamatergic compounds) could also be observed in these models.

Preferential effects of AEDs on discrete brain regions could also play important role in mediation of their antiseizure activity. Indeed, substantia nigra pars reticulata (SNR) appears to be a key structure for propagation of seizures in the brain (Iadarola & Gale, 1982). It is important to bear in mind that AEDs enhancing GABAergic neurotransmission have been shown to inhibit neuronal firing in the SNR, while sodium channel inhibitors (i.e., phenytoin and carbamazepine) were ineffective (Löscher & Wauquier, 1996). Levetiracetam resembles GABAergic drugs by inhibiting neuronal firing in the SNR (Löscher et al., 1996). Furthermore, local injections of levetiracetam into the SNR, as well as compounds either enhancing GABAergic or inhibiting glutamatergic neurotransmission, produced significant protection against pilocarpine-induced seizures (Turski et al., 1986; Klitgaard et al., 2003). This has been contrasted by the lack of anticonvulsant effects of phenytoin in this model following local injection to the SNR (Klitgaard et al., 2003). In that context it is interesting to note that a combination of levetiracetam with sodium channel inhibitors consistently produced much less potentiation of their anticonvulsant effects than combination therapy with drugs enhancing GABAergic transmission or inhibitors of glutamate receptors (Tables 1 and 5).

It has been suggested that advantageous drug combinations observed in preclinical studies, although with obvious limitations, may be predictive of clinical efficacy (Czuczwar & Borowicz, 2002). After reviewing all the experimental data regarding combination therapy with levetiracetam and clinically used AEDs, it appears that such treatment generally leads to enhancement of their anticonvulsant effect (Table 2). Indeed, almost all of the 26 experiments in which levetiracetam was combined with 11 different AEDs in various experimental models revealed synergistic or additive effects (Table 2). It should be noted though that most of these experiments were performed with the use of the so-called subthreshold method, which is based on coapplication of compounds at a fixed dose that is without significant effect on its own together with another compound at protective doses. The coapplied compound may potentiate, attenuate, or remain without effect upon the protective efficacy of a given AED. Such study design is not suitable for the full assessment of the interaction between two compounds (Luszczki & Czuczwar, 2003). These issues could be resolved by isobolographic methods, in which fixed combinations of drugs at different ratios are tested allowing clear determination of synergistic, antagonistic, or additive effects (Tallarida, 2001; Luszczki & Czuczwar, 2003; Jonker et al., 2007). Although it is noteworthy that therapeutically beneficial combinations of levetiracetam with AEDs were observed irrespectively of seizure model and study design (Table 2), it remains to be determined whether such combinations also translate into a particular clinical benefit.

Strong enhancement of anticonvulsant effects of valproate by levetiracetam clearly stood out among all other combinations with clinically used AEDs (Tables 1 and 5). Valproate is characterized by diverse mechanisms of action and broad spectrum of clinical efficacy in both partial and generalized epilepsies (Löscher, 1999). Thus it is perhaps not surprising that combinations of valproate with other AEDs or different anticonvulsant compounds generally lead to synergistic/additive effects regardless of preclinical model utilized (Deckers et al., 2000; Jonker et al., 2007). In fact, as recently reviewed by Jonker et al. (2007), combinations of valproate with other AEDs or experimental anticonvulsants in nearly all cases produced at least additive effects in the MES, PTZ, amygdala kindling, and audiogenic seizure models. The present review, which includes a substantial portion of unpublished data, confirms these observations since levetiracetam markedly enhanced the anticonvulsant activity of valproate in all three models (i.e., audiogenic seizures, MES, and amygdala kindling) (Table 2). Moreover, the degree of potentiation was truly remarkable reaching an almost 30-fold increase in potency against audiogenic seizures (Table 1). This level of potentiation was only on par with clonazepam (Tables 1 and 5), which is a strong and selective positive modulator of GABAA receptors. Interestingly, one of the main mechanisms responsible for anticonvulsant effects of valproate is also the enhancement GABAergic neurotransmission (Löscher, 1999).

It is very important to recognize that combination therapy should provide better seizure protection, which does not come at the expense of added toxicity and adverse effects. In fact, adverse effects can be expected to be reduced in combination treatments, because lower doses of AEDs may be used (Kwan & Brodie, 2006). In this respect, it is essential to emphasize that additive/synergistic effects of levetiracetam, administered at subthreshold doses with most AEDs, were generally not associated with worsening of their side effects (Table 2). In fact, the separation between doses of valproate, clonazepam, or phenobarbital providing seizure protection and doses inducing motor impairment has been dramatically increased in combined treatments with levetiracetam (Fig. 1). When levetiracetam was administered at a relatively high dose (150 mg/kg) required to elevate the electroconvulsive threshold by 50%, motor impairment produced by several AEDs, with the exception of carbamazepine and topiramate, was not affected (Luszczki et al., 2005).

Pharmacokinetic interactions may play an important role in combination therapy of epilepsy (Patsalos et al., 2002). Levetiracetam has a very low potential for pharmacokinetic interactions with other AEDs, because it does not undergo hepatic metabolism and has low protein binding (Patsalos, 2000). Indeed, combination therapy with AEDs was generally not associated with pharmacokinetic interactions (Tables 2, 3, and 4). However, in the case of valproate and felbamate in combination with levetiracetam, possible pharmacokinetic interactions have been observed. Brain concentrations of levetiracetam have increased when administered in combination with felbamate (Luszczki et al., 2007), while in the case of valproate and levetiracetam in combination, valproate plasma and brain concentrations decreased, but the ratio between these two concentrations remained unchanged (Table 3). Thus, this pharmacokinetic interaction cannot be considered as being responsible for the observed enhancement of valproate's anticonvulsant activity. Furthermore, such interaction has not been observed in clinical trials involving both drugs (Patsalos, 2000; Coupez et al., 2003; Patsalos, 2004).

Interaction studies discussed in the present review were performed after single administration of AEDs, which are used chronically in the management of patients with epilepsy. Therefore, different results could potentially be observed after chronic treatment in animal models, but such data are not available yet.

On the basis of the numerous experiments reviewed, it is concluded that combinations of levetiracetam with other AEDs, particularly those enhancing GABAergic inhibition, lead to additive/synergistic effects on seizure protection, which are not associated with more pronounced side effects and pharmacokinetic interactions. These preclinical data warrant further investigation as to what extent such drug combinations provide a therapeutic benefit in patients with epilepsy.

Acknowledgments

  1. Top of page
  2. Summary
  3. Interactions of Levetiracetam with Anticonvulsants in the Audiogenic Seizure Model
  4. Interactions of Levetiracetam with Anticonvulsants in the MES and PTZ Seizure Models
  5. Interactions of Levetiracetam with Anticonvulsants in the Rat Amygdala Kindling Model
  6. Interactions of Levetiracetam with Anticonvulsants in Other Seizure Models
  7. Discussion
  8. Acknowledgments
  9. References

Conflict of interest: We confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this report is consistent with those guidelines. R.K., A.M., and H.K. are full-time employees of UCB. P.P. received speaker's or consultancy fees and/or research grants from UCB.

References

  1. Top of page
  2. Summary
  3. Interactions of Levetiracetam with Anticonvulsants in the Audiogenic Seizure Model
  4. Interactions of Levetiracetam with Anticonvulsants in the MES and PTZ Seizure Models
  5. Interactions of Levetiracetam with Anticonvulsants in the Rat Amygdala Kindling Model
  6. Interactions of Levetiracetam with Anticonvulsants in Other Seizure Models
  7. Discussion
  8. Acknowledgments
  9. References