SEARCH

SEARCH BY CITATION

Keywords:

  • Synergism;
  • Combination;
  • Epilepsy;
  • Literature analysis;
  • Interaction mechanism

Abstract

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS
  5. DISCUSSION
  6. Appendix
  7. Acknowledgments
  8. REFERENCES

Summary: Purpose: Combination therapy is often used in the treatment of seizures refractory to monotherapy. At the same time, the pharmacodynamic mechanisms that determine the combined efficacy of antiepileptic drugs (AEDs) are unknown, and this prevents a rational use of these drug combinations. We critically evaluate the existing evidence for pharmacodynamic synergism between AEDs from preclinical studies in animal models of epilepsy to identify useful combinations of mechanisms and to determine whether study outcome depends on the various research methods that are in use.

Methods: Published articles were included if the studies were placebo-controlled, in vivo, or ex vivo animal studies investigating marketed or experimental AEDs. The animal models that were used in these studies, the primary molecular targets of the tested drugs, and the methods of interpretation were recorded. The potential association of these factors with the study outcome (synergism: yes or no) was assessed through logistic regression analysis.

Results: In total, 107 studies were identified, in which 536 interaction experiments were conducted. In 54% of these experiments, the possibility of a pharmacokinetic interaction was not investigated. The majority of studies were conducted in the maximal electroshock model, and other established models were the pentylenetetrazole model, amygdala kindling, and the DBA/2 model. By far the most widely used method for interpretation of the results was evaluation of the effect of a threshold dose of one agent on the median effective dose (ED50) of another agent. Experiments relying on this method found synergism significantly more often compared with experiments relying on other methods (p < 0.001). Furthermore, experiments including antagonists of the AMPA receptor were more likely to find synergism in comparison with all other experiments (p < 0.001).

Conclusions: Intensive preclinical research into the effects of AED combinations has not led to an understanding of the pharmacodynamic properties of AED combinations. Specifically, the majority of the preclinical studies are not adequately designed to distinguish between additive, synergistic, and antagonistic interactions. Quantitative pharmacokinetic–pharmacodynamic studies of selectively acting AEDs in a battery of animal models are necessary for the development of truly synergistic drug combinations.

At least 25% of patients with epilepsy require treatment with combinations of antiepileptic drugs (AEDs) to improve the efficacy and tolerability (Schmidt and Gram, 1995). The common practice of combining AEDs is based on the idea that two drugs activating different molecular targets may complement each other's action (Brodie, 1992). Only a 6–30% minority of patients refractory to monotherapy become seizure free with combination therapy (Walker and Koon, 1988; Tanganelli and Regesta, 1996; Kwan and Brodie, 2000), but even if complete control cannot be achieved in the remaining refractory patients, improvement may be possible through the appropriate choice of combination therapy. A combination of drugs may be more efficacious by a synergistic anticonvulsant action, an antagonistic action with respect to adverse effects, or both.

However, clinical evidence for the superiority of one AED combination over another is not available, and the evidence coming from studies in animal models for epilepsy is also not conclusive (Deckers et al., 2000). One reason for this may be that the effects of the majority of AEDs available on the market are rather nonspecific in the sense that multiple mechanisms contribute to their anticonvulsant action. This can obscure synergistic anticonvulsant mechanisms and prevents one drug combination from clearly standing out from the others.

A large body of literature exists of experimental studies on the comparative efficacy of AED combinations, including anticonvulsant agents that are not on the market (Czuczwar and Borowicz, 2002). For many of these agents, a well-defined and specific molecular target is thought to be responsible for their anticonvulsant action and this may provide the key to identifying the pharmacodynamic (PD) mechanisms that confer a synergistic interaction in vivo. We therefore performed a meta-analysis of preclinical studies on the combined efficacy of AED combinations, including several novel anticonvulsant compounds. The primary aim of this study is to prove or disprove the hypothesis that compounds with different molecular actions are more likely to interact synergistically than are compounds with similar molecular actions. To this purpose, the available evidence is systematically analyzed by using statistical techniques to determine in an objective manner the likelihood of observing synergy with various combinations of drug classes. A secondary aim of this analysis is to evaluate whether study outcome depends on the various research methods and thereby guide the design of future AED interaction studies.

METHODS

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS
  5. DISCUSSION
  6. Appendix
  7. Acknowledgments
  8. REFERENCES

Data collection

Information from preclinical studies on combination therapy with AEDs was obtained from the MEDLINE bibliographic database of the National Library of Medicine from 1966 through June 2004 and by examination of the references listed in these articles. From these sources, the studies complying with the aim of this review were selected by inspection of the title and the abstract. Studies were included if they were placebo-controlled in vivo or ex vivo animal studies with compounds that have documented anticonvulsant activity in at least one animal model of epilepsy. To enable the evaluation of study methods, only full-length articles in the English language were included; published abstracts were rejected. The characteristics of the retrieved publications were entered into a database that was organized as shown in Table 1. If multiple drug combinations or animal models were evaluated within a single publication, these were defined as individual interaction experiments and entered into separate records in the database.

Table 1. Outline of the protocol used in the classification and selection of data collected in the literature search
A. Animal model of epilepsy
   1. Maximal electroshock
   2. Pentylenetetrazole
   3. Amygdala kindling
   4. Sound in DBA/2 mice
   5. Other animal models
Evaluated were experiments based on models 1–4
B. Dosing of individual agents
   1. Single dose or two doses
   2. Three or more doses
Evaluated were experiments in which one or both agents were tested in three or more doses
C. Method of interaction analysis
   1. Adding responses
   2. ED50 or EC50
   3. Interaction between responses
   4, Isobologram
All methods were evaluated except method 1, when both agents were (potentially) capable of eliciting a response
D. Measurement of drug concentrations
   1. No concentrations measured
   2. Total plasma concentrations
   3. Free plasma concentrations
   4. Brain concentrations
Study of PK interaction was not a requirement for analysis
E. Outcome of experiment
   1. Synergy
   2. Additivity
   3. Antagonism

Data selection

To identify potentially synergistic drug combinations, the interaction experiments fulfilling the following three requirements were selected from this database. The first requirement was that an established animal model was used to enable comparison of the results obtained with each model. Based on their frequent application, the four animal models listed in Table 1 were selected; studies using modifications of standard methods were excluded from analysis. The second requirement was that at least one of the agents in an interaction experiment was administered in three doses or more to allow assessment of the dose–response relation for the individual drug. If both agents are administered in only one or two doses, it is not possible to assess the synergistic, additive, or antagonistic nature of the interaction. This point is discussed in detail in Appendix A, “Making the distinction between synergy, additivity, and antagonism.”

The third and last requirement was that a (semi-) quantitative method was applied to interpret the results. In practice, a variety of methods is used for interaction analysis, and their merits and limitations are discussed in Appendix A. In this Appendix, it is argued that determination of the effect of a single threshold dose on the median effective dose (ED50) of a second agent is at best a semiquantitative method, because the observed degree of potentiation depends on the threshold dose chosen. Fully quantitative methods include (a) isobolographic analysis, (b) the use of logistic linear models, and (c) two-way analysis of variance. Studies comparing the magnitude of individual and combined responses without taking into account the dose–response relation of at least one of the individual agents are methodologically flawed, and these studies were excluded from analysis. Methods of interaction analysis were evaluated for studies based on (semi-) quantitative methods to determine whether study outcome was influenced by the choice of a particular method.

Data analysis

The number of experiments reporting success (synergistic interaction) and failure (no synergy detected) was evaluated. Records were also kept of the number of experiments reporting additive and antagonistic interactions, but the results on these two types of interactions were combined in the statistical analysis because only few antagonistic interactions were reported. Heterogeneity between experiments with respect to the probability of success was tested by using the χ2 test at the 5% significance level. Different combinations of AED classes were compared according to the primary molecular targets involved. Drugs with anticonvulsant action attributed to one primary target were classified by their action on ion-channel subtypes and receptor subtypes, the principal actions being Ca2+-channel blockade, enhancement of postsynaptic γ-aminobutyric acid (GABA)A-receptor activity and blockade of N-methyl-d-aspartate (NMDA) receptors (Table 2). Compounds with documented action on multiple targets were classified separately; examples of such drugs include phenobarbital (PB) and valproate (VPA). The molecular targets of the combined agents were not a criterion for selection per se, but only the combinations of targets that were tested in ≥10 experiments were included in the statistical evaluation. Apart from the molecular targets studied, the use of a specific animal model or method of interaction analysis could also influence the probability of success. Possible confounding by these variables was taken into account by logistic regression for factors with a generalized linear model by using the S-Plus computer program (version 2000, MathSoft Inc., Seattle, WA, U.S.A.).

Table 2. Classification of the retrieved anticonvulsant agents according to their (probable) molecular targets data
Multiple targetsIon channelsPostsynaptic receptorsPresynaptic receptorsVarious targets
Na+Ca2+GABAANMDAAdenosineAMPA/kainateGlutamateGABA
  1. Currently marketed antiepileptic drugs are shown in italics. Abbreviations are listed in Table 7.

 L-typeAgonistsAntagonistA1 Uptake inhibitionβ-Blockers
FelbamateCarbamazepineAmlodipineTHIPα-AACHACFM-2ACPT-1NNC 05-2045Propranolol
GabapentinphenytoinDiltiazemMuscimolAPV8-CPXGYKI 52466GDEETiagabineAtenolol
Lamotrigine Isradipine CPPL-PIALY300164PPGSNAP-5114Metoprolol
PEMA NicardipineNeurosteroidsCPP-ene NBQXLY354740 Acebutolol
Phenobarbital NifedipineAllopregnanoloneD-CPP-eneA2 GABA metabolism 
Primidone NiguldipineAlfaxaloneCGP 39551NECA AOAANitric oxide
SIB 1893 NimodipineGanaxoloneCGP 37849 GHBAMolsidomine
Topiramate VerapamilPregnanoloneCGP 40116A3 7-Nitroindazole
Valproate  Co 2-1068CGP 43487APNEA GABA (aselective) 
T-typeBZD siteLY235959 ProgabideDopamine
EthosuximideClobazamLY233053Aselective Apomorphine
FlunarizineChlordiazepoxideL-701,3242-CA GABABFluphenazine
ClonazepamMemantine BaclofenL-DOPA
DiazepamDizocilpine CGP36742 
FlunitrazepamPhencyclidine Unknown
NicotinamideProcyclidine Oxiracetam
Nitrazepam Piracetam
OxazepamGlycine site Retigabine
TofizopamD-cycloserine 
Triazolam eliprodilD-serine Ca2+ release
 Dantrolene
Other siteGlycine 
LoreclezoleIFENPRODIL Gap-junction blocker
 Carbenoxolone
Table 7. Abbreviations of compound names
Abbrev.Compound name
2-CA2-chloroadenosine
3α,5α-PAllopregnanolone
3α,5ß-PPregnanolone
7-NI7-nitroindazole
8-CPX8-cyclopentyl-1,3-dimethylxanthine
α-AAα-aminoadipic acid
ACPT-11-aminocyclopentane-1,2,4-tricarboxylic acid
ALFAlfaxalone
AOAAAminooxoacetic acid
APNEAN6-2- (4-aminophenyl)ethyl-adenosine
APOApomorphine
APV2-aminophosphonovaleric acid
BACBaclofen
CBZCarbamazepine
CGP 37849(D,L- (E)-2-amino-4-methyl-5-phosphono-3-pentenoic acid
CGP 39551(D,L- (E)-2-amino-4-ethylester-5-phosphono-3-pentenoic acid
CGP 40116D- (E)-2-amino-4-methyl-5-phospono-3-pentenoic acid
CGP 43487D- (E)-2-amino-4-methylester-5-phospono-3-pentenoic acid
CPP-ene3- (2-carboxypiperazine-4-yl)-1-propenyl-1-phosphonic acid
CFM-21- (4'-aminophenyl)-3,5-dihydro-7,8-dimethoxy-4H-2,3-benzodiazepine-4- one HCl
CHAN6-cyclohexyladenosine
CLBClobazam
CDZChlordiazepoxide
CZPClonazepam
D-CPP-ene2- (2-carboxypiperazine-4-yl)-1-propenyl-1-phosphonic acid
DCSD-cycloserine
DZPDiazepam
ESMEthosuximide
FBMFelbamate
FLUFlunarizine
FNPFlunitrazepam
FPZFluphenazine
GBPGabapentine
GDEEGlutamate diethylester
GLYGlycine
GYKI 524661- (4-aminophenyl)-4-methyl-7,8-methylenedioxy-5H-2,3-benzodiazepine HCl
L-PIAL-phenylisopropyladenosine
LY233053[cis- (±)-4-[ (2H-tetrazol-5-yl)methyl]piperidine-2- carboxylic acid
LY235959[ (-)-3R,4aS,6R,8aR-6- (phosphonomethyl)-decahydroisoqu- inoline-3-carboxylic acid
LY3001647-acetyl-5- (4-aminophenyl)-8,9-dihydro-8-methyl-7H-1,3-dioxolo (4,5H)-2,3- benzodiazepine
MEMMemantine
MK-801Dizocilpine
NAMNicotinamide
NBQX2,3-dihydroxy-6-nitro-7-sulfamoylbenzo (F)quinoxaline
NECA5'-N -ethylcarboxamidoadenosine
NICNicardipine
NIFNifedipine
NIGNiguldipine
NIMNimodipine
NNC 05-20451- (3- (9h-carbazol-9-yl)-1-propyl)-4- (4-methoxyphenyl)-4-piperidinol
NZPNitrazepam
PBPhenobarbital
PEMAPhenylethylmalonamide
PHTPhenytoin
PPGPhosphonophenylglycine
SIB 1893(E)-2-methyl-6- (2-phenylethynyl)pyridine
SNAP-5114(S)- (-)-1-[2-[tris- (4-methoxyphenyl)methoxy]ethyl]-3-piperidinecarboxylic acid
STPStiripentol
TGBTiagabine
THIP(4,5,6,7)-tetrahydroisoxazolo[5,4c]pyridin-3-ol
TPMTopiramate
TZLTriazolam
VPAValproate

RESULTS

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS
  5. DISCUSSION
  6. Appendix
  7. Acknowledgments
  8. REFERENCES

Characteristics of interaction studies

In total, 107 publications addressed the topic of combining AEDs in accordance with our inclusion criteria. In these studies, 536 interaction experiments were published in the period from 1955 to June 2004. The number of retrieved publications was about two per year throughout the 1980s, four per year throughout the 1990s, and increased to about seven per year from 2000 onward. This indicates a long-standing and continuing interest in the combination of AEDs. The majority of experiments included a conventional AED; phenytoin (PHT) was studied most frequently (99 experiments), followed by PB (98), benzodiazepines [BZDs; primarily diazepam (DZP) and clonazepam (CZP)] (96), VPA (90), and carbamazepine (CBZ) (76).

The individual characteristics of all retrieved experiments are outlined in Tables 3–6. In 293 (54%) experiments, a synergistic PD interaction was reported, whereas additivity was found in 232 (43%) experiments. In 11 (2%) experiments, an antagonistic PD interaction was reported. Interestingly, both synergism and antagonism were reported for the combination of oxcarbazepine (OXC) and CZP, dependent on the dose ratio of the drugs (Luszczki et al., 2003b; Luszczki and Czuczwar, 2003). In 42% of the experiments, the characterization of drug interactions was based on a single dose–response relation or a single response level. Three or more doses or response levels were characterized in 19% of all interaction experiments, but only two studies characterized drug interactions fully by evaluating AED combinations over a range of doses of both drugs (Gennings et al., 1995; Jonker et al., 2004).

Table 3. Electroshock. All studies are conducted in mice using earclip electrodes, unless indicated otherwise.
ReferenceMethodCombination(s) testedPK interaction?Outcome
  1. *Ref.: +/−: indicates whether the experiment is included in the analysis (+) or not (−) according to the criteria listed in Table 1. Method: MES, maximal electroshock; in mice using earclips, unless otherwise indicated; ES, electroshock; electroshocks of various intensities are administered to determine the CSx (usually CS50), that is the current intensity producing tonic hindlimb extension in x% of the animals TS: threshold stimulation; current is adjusted to determine the median convulsive current (CC50) CSM: cortical stimulation model, Combinations tested:, range: range of effective doses tested (≥ 3doses) ED: single effective dose tested SE: single subeffective dose tested NE: drug is not effective at any dose that was tested Outcome: FEC, fractional effective concentration (see Appendix A); CS50, convulsive threshold: current needed for tonic hindlimb extension in 50% of animals, ED50, dose preventing tonic hindlimb extension at MES in 50% of animals; EC50, concentration preventing tonic hindlimb extension at MES in 50% of animals LLM: linear logistic model (see Appendix A); E/F, ratio of timing of stimulus-induced flexion and extension periods.

+Weaver et al., 1955Corneal MESPB (range)PHT (range)Not excludedFEC index = 0.6
+Leppik and Sherwin, 1977Corneal MESPB (range, fixed ratios)PHT (range)Brain concentrations used ([PB])Sum of concentrations accounts for combined response in 2 fixed ratios
+Kleinrok et al., 1980 − +Corneal MESAPO (NE range) L-DOPA/(ED) GHBA/BAC/AOAA (ED)PB/PHT (ED)Not excludedNo effect on [UPWARDS ARROW]CS50s Dose-orderly CS50 with PB and PHT (except of PB with AMA) Dose-orderly [UPWARDS ARROW]CS50 with PB, not PHT (except with AOAA)
+Masuda et al., 1981Corneal MES (mice/rabbit)PHT (range)PB (range)Total plasma concentrations used; ratio brain/plasma concentration in mice not changedAllnon-responders below line of additivity on isobole; FEC index=0.51 (mouse)/0.58 (rabbit)
+Bourgeois et al., 1983aMESNAM (NE)PB (range)Not excludedNo effect on ED50 of PB
+Bourgeois et al., 1983bMESPB (range)PRM (1:1 with PB)Brain and total plasma concentrations determined[PB]+[PRM] at E50 smaller than either EC50s
+Czuczwar et al., 1984b + +Corneal MESGDEE (NE range) APV (range) a-AA (NE range)PB/PHT (range)Not excludedNo effect on either ED50 Dose-orderly [DOWNWARDS ARROW]ED50s [DOWNWARDS ARROW]ED50 PB at middle dose;no effect on ED50 PHT
+Czuczwar et al., 1984aCorneal MESMuscimol/baclofen/GHBA (range)PB/PHT (range)Not excludedDose-orderly ED50 of PB with baclofen or GHBA; [DOWNWARDS ARROW]ED50 of PHT with musci mol; no effect of other combinations
Rastogi and Ticku, 1985MESPB/DZP (SE)/progabide (NE)PB/DZP (SE-EDrange)Not excludedIncreased response compared to control
+Chweh et al., 1985 +Corneal MESProgabide (range) THIP (range)CDZ/CLB/DZP/CZP/TZL (range) CDZ/TZL (range)Not excluded[DOWNWARDS ARROW]ED50 for all BZDs
+Bourgeois, 1986MESPB (range)PHT (range)Brain concentrations used ([PHT])Additive on isobole (FEC =0.92)
+Mondadori and Schmutz, 1986Corneal MES (rat)Oxiracetam/piracetam (NE range)CBZ/VPA/PHT/PB/CZP (range)Not excludedLLM: interaction between VPA and oxiracetam, and of CBZ or CZP in all combinations
−Saano, 1986MES (up- and-down)Tofizopam (SE)CZP/DZP/FNP/CBZ/PB/PHT/ VPA/oxazepam (ED)Not excludedNo change in seizure threshold (mA/kg), except increase with DZP
+Morris et al., 1987Corneal MESCBZ (range)PHT (range)Brain concentrations of both drugs usedFEC index=0.89 0.05 (Additive)
+Bourgeois and Wad, 1988MESPB (range)CBZ (range)Brain concentrations of both drugs used (no interaction reported)Additive on isobole FEC index = 0.86
+Peterson et al., 1990Corneal MES (rat)GLY (NE)PB/CBZ/MK-801/DZP/PHT/VPA (range)Brain concentrations of GLY and all compounds but DZP determined ([GLY])ED50 of E/F ratio significantly decreased with PB, CBZ, MK-801 and DZP, not with PHT and VPA
+Czuczwar et al., 1990bCorneal MESL-PIA (range)DZP/VPA/PB/CBZ/PHT (range)Not excluded[DOWNWARDS ARROW]ED50 of DZP and VPA; no effect on ED50 of PB, CBZ and PHT
+Czuczwar et al., 1990bCorneal MESNECA (range)VPA/PB/CBZ (range)Not excluded[DOWNWARDS ARROW]ED50 of VPA, ED50 of PB, no effect on ED50 CBZ
Peterson, 1991aCorneal MES (rat)D-serine (NE/SE)PB/CBZ/PHT (ED)Not excludedDecrease in occurrence of THE
+Gordon et al., 1991Corneal MESFBM (range)DZP (range)Not excluded[DOWNWARDS ARROW]ED50 of DZP
+Urbanska et al., 1991 +MESMK-801 (SE range)VPA/PB (range) PHT/CBZ (range)No change in total plasma levels of VPA or PB Not excludedDose-orderly ED50 of VPA and PB No effect on ED50 of PHT and CBZ
+Urbanska et al., 1992MESMEM (SE range)/THP (NE)VPA (range)No change in total plasma levels of VPADose-orderly [DOWNWARDS ARROW]ED50 VPA
+Czechowska et al., 1993MESCGP37849/CGP39551 (range)VPA (range)No change in total plasma levels of VPA No change in totalDose-orderly [DOWNWARDS ARROW]ED50 VPA
+Pietrasiewicz et al., 1993MESCGP37849 (SE range)CBZ/PHT/PB (range)plasma levels of AEDs except lowering of PB with CGP39551Dose-orderly [DOWNWARDS ARROW]ED50 CBZ, PHT and PB
+ CGP39551 (SE range) Dose-orderly [DOWNWARDS ARROW]ED50 PHT and PB, but not ED50 CBZ
+Gordon et al., 1993Corneal MESPHT/CBZ/VPA/PB (SE)FBM (range)Rise of total plasma level from 37% to ± 55% of EC50 of FBM with VPA and CBZ; otherwise no changesDose-orderly [DOWNWARDS ARROW]ED50 of FBM in all combinations
+Zarnowski et al., 1993MESNBQX (range)CBZ/DZP/PHT/PB/VPA (range)No change in total plasma levels of AEDsDose-orderly [DOWNWARDS ARROW]ED50 of all AEDs tested
+Shank et al., 1994Corneal MESTPM (range, fixed ratio)PB/PHT/CBZ (range)Not excludedTPM/CBZ and TPM/PB ‘slightly synergistic’ on isobole, TPM/PHT additive
+Chez et al., 1994MESPHT (range)VPA (range)Brain concentrations of both drugs used (no interaction reported)[DOWNWARDS ARROW]EC50 PHT and VPA; FEC index = 0.67
Wlaz et al., 1994Corneal TSDCS (range)MK-801 (range)Not excludedNo additional [UPWARDS ARROW]CC50 in combinations
+Zarnowski et al., 1994b +MESProcyclidine (NE range) Ifenprodil (SE)CBZ/PHT/PB/VPA/DZP (range)No change in total plasma levels of AEDs Not excludedDose-orderly; [DOWNWARDS ARROW]ED50 of all AEDs No effect on any ED50
+Zarnowski et al., 1994aMESD-CCP-ene (SE range)CBZ/DZP/PHT/PB/VPA (range)No change in total plasma levels of AEDsDose-orderly; [DOWNWARDS ARROW]ED50 of all AEDs tested
+Borowicz et al., 1995 +MESGYKI 52466 (SE range)VPA/CBZ/PHT (range) PB (range)No change in total or free plasma levels of AEDs Not excludedDose-orderly; [DOWNWARDS ARROW]ED50 of all AEDs tested; No effect on ED50 of PB
Czuczwar et al., 1995ESGYKI 52466/NBQX (range)MK-801 (ED) D-Not excluded; considered unlikely since several fold changes in response were observedDose-orderly [UPWARDS ARROW]CS50 of both GYKI and NBQX
+MES CCP-ene (range) [DOWNWARDS ARROW]ED50 of D-CCP-ene
+Gennings et al., 1995Corneal MESCBZ and FBM and PHT (range) Not excludedLLM: only interaction occurs between CBZ and PHT at time where both ineffective when given separate; no interaction at later time
+Deutsch et al., 1995ESMK801 (SE/NE range)FLU (range)Not excluded; considered unlikelyEffect of MK801 on seizure threshold curve of FLU (2 way ANOVA)
+Deutsch et al., 1996ES3a,5ß-P(ED)/alfaxalone (range)FLU (range)Not excludedEffect of 3a,5ß-P and ALF on seizure threshold curve of FLU (2 way ANOVA)
+Gasior et al., 1996aESNIC/NIF/FLU (SE range)CGP 40116/CGP 43487 (ED)Not excluded; considered unlikely based on studies with conventional AEDsDose-orderly [UPWARDS ARROW]CS50 compared to saline + Ca2+-channel blocker
Wlaz et al., 1996Corneal ESDCS (SE range)PHT/CBZ (ED)Tendency to increase in free plasma levels of PHT in combination; no effect on free CBZ concentrationsDose-orderly [UPWARDS ARROW]CC50 with CBZ, biphasic response with PHT
+Corneal MESDCS (SE)PHT/CBZ (range) [DOWNWARDS ARROW]ED50 of PHT; no effect on ED50 of CBZ
+Borowicz et al., 1996MESLY235959/LY233053 (SE range)CBZ/PHT/VPA/PB (range)No change in total plasma levels of CBZ, VPA, PHT or PBDose-orderly [DOWNWARDS ARROW]ED50 in all drug combinations
+Borowicz et al., 1997aMESNIG (range)PHT/VPA (range)Not excludedNo effect on ED50 of PHT and VPA
+ + NIG (range) Isradipine/dantrolene (NE range)CBZ/PB (range) CBZ/PB/PHT/VPA (range)No change in free plasma levels of CBZ and PB Not excludedDose-orderly [UPWARDS ARROW]ED50 of CBZ and PB no effect on ED50
+Borowicz et al., 1997bMESAPNEA (SE range)CBZ/PHT/VPA/PB (range)No change in total or free plasma levels of any AED; PB: total levels onlyDose-orderly [DOWNWARDS ARROW]ED50 of CBZ, PHT, VPA and PB
+Gasior et al., 1998MESNIC (NE)/NIF (SE range)/FLU (SE range)FBM (range)Not excludedNo effect on ED50 of FBM
+Czuczwar et al., 1998MESLY 300164 (range)CBZ/VPA/PHT/PB (range)No effect of LY 300164 on free plasma level of AEDsDose-orderly [DOWNWARDS ARROW]ED50
Balakrishnan et al., 1998Corneal MES (rat)NIM (range)PHT/VPA (range)Not excludedED25+ ED25 gives more than 50% protection
Joseph et al., 1998CSM (rat)FLU (SE)VPA (ED)/VPA (SE)Not excluded(Higher) increase in TLS and TGS over time with combination
Della Paschoa et al., 1998CSM (rat)VPA (SE)PHT (range)Total and free serum levels determined in vitroEffect of combination on both outcomes > drug effect alone in the conc. tested
+Dalby, 2000MESTGB (NE range)NNC 05-2045 (ED)/SNAP-5114 (NE)Not excludedIncreased protection at intermediate dose with NNC 05-2045, not with SNAP-5114
+Swiader et al., 2000MESTPM (SE range)VPA/CBZ/PHT/PB (range)No effect of TPM on free plasma level of other AEDs, except increase of CBZDose-orderly [DOWNWARDS ARROW]ED50
+Borowicz et al., 2000bMESLY300164 (SE)DZP (range)No effect of LY 300164 on free plasma level of DZPDose-orderly [DOWNWARDS ARROW]ED50
Assi, 2001ESCHA/CPP-ene (SE)VPA/DZP/PHT/PB/CBZ (ED)No effects on total plasma levels of AEDs[UPWARDS ARROW]CS50 with all AEDs studied
+MESCHA/CPP-ene (SE)VPA/DZP/PHT/PB/CBZ (range) [DOWNWARDS ARROW]ED50 of all AEDs studied
+Lu and Yu, 2001Corneal MES3a,5a-P (range)PB (range)No exclusion[DOWNWARDS ARROW]ED50 of combination in fixed ratio
+Borowicz et al., 2002bMES2-CA (range)VPA/CZP/PHT/PB/CBZ (range)No effects on free plasma level of CBZ[DOWNWARDS ARROW]ED50 of CBZ but no change in other ED50s
+Borowicz et al., 2002dMESALF (ED)VPA/CZP/PHT/PB/CBZ (range)[UPWARDS ARROW]Free plasma concentration VPA; all other levels not studied[UPWARDS ARROW]ED50 of VPA but no change in other ED50s
+Borowicz et al., 2002cMESGBP (range)VPA/CZP/PHT/PB/LTG/LY-300164 (range)Free plasma level PB and LTG, no change in other concentrationsCBZ, VPA, PHT and LY-300164 synergistic on isobole
+Luchowska et al., 2002MESPropranolol/atenolol/metoprolol /acebutolol (range)VPA/DZP/CBZ/PHT (range)No effects on free plasma or brain levels of VPA or DZP; other AED levels not studied[DOWNWARDS ARROW]ED50 of VPA and DZP with propranol/metoprolol & of VPA with acebutolol
+Luszczki et al., 2002MESSIB 1893 (range)VPA (range)No effects on free plasma levelsSynergistic on isobole
+Sun et al., 2002MESVPA (range)CBZ (range)Not excludedAdditive at all dose ratios (isobole at 50% level)
+Borowicz et al., 2003MESSIB 1893 (NE)CBZ/PHT/PB (range)Not excludedAdditive on isobole
+Luszczki and Czuczwar, 2003MESOXC (range)PHT/PB/VPA/CBZ/CZP (range)Not excludedPHT antagonistic, CZP synergistic and antagonistic, others additive on isobole
+Luszczki et al., 2003cMESLTG (range)PHT/PB/VPA/CBZ/TPM (range)Free plasma levels of PB, no effect on other AED levelsCBZ antagonistic, PHT additive and others synergistic on isobole
+Luszczki et al., 2003bMESPHT (range)OXC/CBZ (range)CBZ not excluded, no effect on free plasma levels of OXCOXC antagonistic, CBZ additive on isobole
+ CZP (range)OXC/CBZ/PHT (range)Not excluded for CBZ/PHT, no effect on free plasma levels of OXCCBZ/PHT synergistic, OXC synergistic and antagonistic on isobole
+Luszczki et al., 2003aMESTGB (range)PHT/CBZ/VPA/PB/LTG /TPM/FBM (range)TGB not tested, brain levels of PHT, VPA, no change in other drug levelsAll combinations additive on isobole
+Luszczki et al., 2003dMESTGB (range)GBP (range)No change in brain concentrations of GBPSynergistic on isobole
Jonker et al., 2004CSM (rat)TGB (range)LTG (range)No change in total plasma levels of both drugsSynergistic on isobole for eye closure and head jerk, but not for forelimb clonus and forelimb tonus
+Luszczki and Czuczwar, 2004aMESTGB (range)GBP (range)Not excludedSynergistic on isobole
Mazarati et al., 2004Status Epilepticus (rat)LEV (range)DZP (ED)Not excluded[DOWNWARDS ARROW]Seizure time and number of spikes (one way ANOVA)
Sills et al., 2004MESTPM (range)PB/PHT/PRM/CBZ/VPA /CLB/LTG/FBM/TGB (range)Not excludedResponse not enhanced
+ ESM/VGB/GBP/LEV (NE range)Not excludedResponse not enhanced
Table 4. Chemically induced seizures
ReferenceDoseCombination (s) testedPK interaction?Outcome
Stach and Kacz, 1977Cortical ouabain (rabbit)APO (SE)GHBA/AOAA (SE)Not excluded[DOWNWARDS ARROW]Behavioral seizure activity with both combinations (no statistical analysis)
+Czuczwar et al., 1981PTZ 110 mg/kg BIC isoniazidPHT (NE)DZP (range)Not excluded[DOWNWARDS ARROW]ED50 of DZP No change in ED50 of DZP
+Czuczwar et al., 1982PTZ 110 mg/kg BIC 3.5 mg/kgPHT (NE)CZP, NZP, CDZ (range)Not excluded[DOWNWARDS ARROW]ED50 of CZP and NZP, not of CDZ [DOWNWARDS ARROW]ED50 of CZP, not of NZP and CDZ
+Bourgeois et al., 1983aPTZ 85 mg/kg BIC 2.7 mg/kgNAM (range)PB (range)Brain concentrations of PB used; no interaction reported[DOWNWARDS ARROW]EC50 of PB [DOWNWARDS ARROW]EC50 of PB
+Bourgeois et al., 1983bPTZ ? mg/kgPB (range)PRM (1:1 with PB)Brain concentrations used; no interaction reportedNo change in EC50 of PB
+ PEMA (1:1 with PB) [DOWNWARDS ARROW]EC50 of PB
+Chweh et al., 1985 +PTZ 85 mg/kgProgabide (range) THIP (range)CDZ/CLB/ DZP/CZP/TZL (range) CDZ/TZL (range)Not excludedNo effect on ED50 of any BZD
Hawkins and Mellanby, 1986TT 0.6 or 1 μ lCBZ (ED)piracetam (SE)Not excluded[DOWNWARDS ARROW]No. of fits with combination compared to CBZ alone
Saano, 1986BIC up to 0.09 mgTofizopam (SE)CZP/DZP/FNP /oxazepam (ED) CBZ/PB/PHT/VPA (NE)Not excludedIncrease of seizure threshold (dose BIC), except with oxazepam No increase of seizure threshold (dose BIC)
+Bourgeois, 1988PTZ 85 mg/kgVPA (range)ESM (range)Brain concentrations of both drugs used [[UPWARDS ARROW]ESM]Additive on isobole FEC index = 1.02
+Czuczwar et al., 1990aPTZ 115 mg/kgVPA, PB, ESM,DZP (SE) NIF (range)diltiazem/verapamil (NE range)No change in total plasma levels for effective combinations (ineffective combinations ND) [UPWARDS ARROW]Total plasma levels of VPA and PBDose-orderly [DOWNWARDS ARROW]% seizing mice after diltiazem + ESM but not in other combinations Dose-orderly [DOWNWARDS ARROW]% seizing mice except with DZP
+Peterson, 1991bPTZ 75 mg/kg in ratsGLY (SE)DZP/VPA/ESM/PB (range)Not excluded[DOWNWARDS ARROW]ED50 of DZP and VPA, not of ESM and PB
+Gordon et al., 1991PTZ 100 mg/kg/BIC/isoniazidFBM (range)DZP (range)Not excluded[DOWNWARDS ARROW]ED50 of DZP
+Musolino et al., 1991PTZ 75 mg/kg in ratsVPA (range)ESM (range)No change in kinetic parameters based on total plasma concentrations in separate experiment[DOWNWARDS ARROW]ED50 for tonic–clonic seizures, not for myoclonias
Gasior et al., 1996bPTZ 81 mg/kgESM/VPA/CZP (SE)NIC (SE range)No change in total plasma concentrations of ESM and in free concentration of VPA; CZP not determinedDose-orderly [DOWNWARDS ARROW]% mice with clonus (not tonus) in combinations with ESM or VPA but not CZP
+ NIM/FLU (NE range) NIM same effect, FLU ineffective
+Gasior et al., 1997PTZ 70 mg/kg3a,5a-P/3a,5ß-P/ganaxolone/Co 2-1068/PB (SE range)DZP (range)Not excluded[DOWNWARDS ARROW]ED50 of DZP for all combinations, with slope of D-E curve
+ DZP (SE)3a,5a-P, 3a,5ß-P, ganaxolone, Co 2-1068, PB (range)Not excluded[DOWNWARDS ARROW]ED50 of all steroids and PB except 3a,5a-P
Stephen et al., 1998PTZ 85 mg/kgLTG (SE)TPR (SE)Not excludedWith combination no seizure in 15 min.
Shantilal et al., 1999PTZ 20 mg/kg in ratFLU (SE)VPA (ED)Not excludedWith combination stronger decrease in number and duration of SWD
+Klodzinska et al., 2000PTZ 80 mg/kgLY354740 (range)DZP/ESM/VPA (range)Free plasma level of DZP, other AEDs not studied[DOWNWARDS ARROW]ED50 of DZP but no change in other ED50s
+Borowicz et al., 2000bPTZ 85 mg/kgLY300164 (SE)DZP (range)No effect of LY 300164 on free plasma level of DZPDose-orderly [DOWNWARDS ARROW]ED50
+Assi, 2001PTZ ? mg/kgCHA/CPP-ene (SE)VPA/DZP/PHT/PB/CBZ (range)No effects on total plasma levels of AEDs[DOWNWARDS ARROW]ED50 of all AEDs studied
PTZ i.v. thres-holdCHA/CPP-ene (SE)VPA/DZP/PHT/PB/CBZ (ED) [UPWARDS ARROW]Seizure threshold with all AEDs studied
+Kaminski et al., 2001aPTZ 95 mg/kgAmlodipine (SE)VPA/ESM/PB (range)No effect of amlodipine on free plasma levels of AEDs[DOWNWARDS ARROW]ED50 of all AEDs studied
+Borowicz et al., 2002bPTZ 95 mg/kg2-CA (SE)VPA/PHT/PB/CZP (range)Not excluded[DOWNWARDS ARROW]ED50 of CZP but no change in other ED50s
+Borowicz et al., 2002dPTZ 85 mg/kgALF (NE)PB/CZP/ESM (range)Not excludedNo change in ED50s
+Borowicz et al., 2002dPTZ 85 mg/kgALF (NE)VPA (range)[UPWARDS ARROW]Free plasma level of VPA, ALF not tested[UPWARDS ARROW]ED50 of VPA
+Cuadrado et al., 2002PTZ 110 mg/kgLTG (SE range)VPA (SE range)[UPWARDS ARROW]LTG brain levels, no influence on VPA levels[DOWNWARDS ARROW]EC50s of both agents
+Tutka et al., 2002PTZ 118 mg/kgMolsidomine (SE range)VPA/CZP/ESM/PB[UPWARDS ARROW]Free plasma level of VPA, other AEDs not tested[DOWNWARDS ARROW]ED50 of VPA attributed to PK interaction; no change in other ED50s
+Borowicz et al., 2003PTZ 103 mg/kgSIB 1893 (NE)VPA/ESM/PB/CLZ (range)Not excludedNo change in ED50s
Swiader et al., 2003aAminophylline 273 mg/kgLY 300164 (SE/ED)DZP (range)Not excluded[DOWNWARDS ARROW]ED50 at ED of LY 300164, and at SE of LY 300164 with DZP alone
+Swiader et al., 2003bPTZ 99 mg/kgVGB (SE)ESM/VPA/CZP (range)No change in total plasma levels of VPA or CZP, but [UPWARDS ARROW]ESM levels[DOWNWARDS ARROW]ED50 of ESM due to PK interaction, no change in other ED50s
+Luszczki and Czuczwar, 2004aPTZ 100 mg/kgTGB (range)GBP (range)Not excludedAdditive on isobole
+Sills et al., 2004PTZ 85 mg/kgTPM (NE range)PB/PRM/ESM/VPA /FBM/TGB (range) Response not enhanced
+ + PHT/CBZ/VGB/GBP (NE range) CLB/LTG/LEV (NE range)Not excludedResponse not enhanced Increased response
Table 5. Amygdala kindling model
ReferenceMethodCombination (s) testedPK interaction?Outcome
  1. †Outcome

  2. ADT, afterdischarge threshold; GST, generalized seizure threshold; SS, seizure severity SD, seizure duration; ADD, afterdischarge duration.

Peterson, 1986Male ratsGLY (NE)DZP/PB (SE range)Not excluded[DOWNWARDS ARROW]ADD (both) and SS (only DZP)
Dziki et al., 1992Female ratsMK-801 (SE)VPA (range)No change in total plasma concentration of VPA[UPWARDS ARROW]ADT and [DOWNWARDS ARROW]SS at one dose, no change in SD
Löscher et al., 1993Female ratsMK-801/CGP 39551 (SE range)NBQX (range)Not excluded[UPWARDS ARROW]ADT and [DOWNWARDS ARROW]SS and SD
Löscher and Hönack, 1993Female ratsCGP 37849 (range)VPA (2 ED)No change in total plasma concentration of VPA[UPWARDS ARROW]ADT at one dose; no change in SS or SD
Löscher and Honack, 1994Female ratsMEM (SE range)NBQX (range)Not excluded[UPWARDS ARROW]ADT and [DOWNWARDS ARROW]SS and SD
+Ebert et al., 1997Female ratsL-701,324 (range)Ifenprodil (2 NE)Not excluded[UPWARDS ARROW]ADT and [DOWNWARDS ARROW]SS, no change in SD
+Wlaz et al., 1999 + +Female rats, to ADT Female rats, to GSTCGP40116 (NE range) L-701,324 (SE) L-701,324 (NE)Eliprodil (range) Eliprodil (range) Eliprodil (NE)Not excluded Not excludedNo change in any parameter Reversed dose-orderly [UPWARDS ARROW]ADT and SS and SD [UPWARDS ARROW]GST and [DOWNWARDS ARROW]SS and SD, no change in ADD
+Borowicz et al., 2000c + +Male ratsAPNEA (range)CBZ/CZP (range) VPA/PB (SE) PHT (SE)No effect of APNEA on free plasma levels of CBZ, PB or VPA; CZP levels ND Not excludedDose-orderly [DOWNWARDS ARROW]SS, SD and ADD Dose-orderly [DOWNWARDS ARROW]SD and ADD No effect on any parameter
Borowicz et al., 2000bMale ratsLY300164 (SE on SD and ADD, NE on SS)DZP (range)No effect of LY 300164 on free plasma level of DZPDose-orderly [DOWNWARDS ARROW]SS, SD and ADD
Borowicz et al., 2000aMale rats7-NI (SE)VPA/PHT/CBZ/PB/CZP (SE)No effect of 7-NI on free plasma levels of CBZ or PB; other levels not measured[DOWNWARDS ARROW]SS and SD with CBZ or PB, not with other AEDs
Borowicz et al., 2001bMale ratsLY300164 (range) LY235959 (NE)VPA/PHT/CBZ/PB (SE)Free plasma levels of PHT and VPA did not differ from control[DOWNWARDS ARROW]SD and ADD with VPA only [DOWNWARDS ARROW]SS, SD and ADD with PHT only
Borowicz et al., 2001aMale ratsGYKI 52466 (SE)CZP/VPA/CBZ/PB/PHT (SE)No change in free plasma level of VPA, other levels not determined[DOWNWARDS ARROW]SD and ADD of CZP and VPA; not with other AEDs
Borowicz et al., 2002aMale ratsNIG (SE)VPA/PHT/PB/CZP/CBZ (SE)No effect of NIG on free plasma levels of CBZ or PBDose-orderly [UPWARDS ARROW]SD and ADD with CBZ and PB
+Borowicz et al., 2002dMale ratsALF (NE)VPA/PHT/PB/CZP/CBZ (SE)Free plasma level of VPA, all other levels not testedNo change in any parameter
Zhang et al., 2003Male ratsLTG (ED)GBP (ED)/MK-801 (NE)Not excluded[DOWNWARDS ARROW]Tolerance development to LTG for SS, SD and ADD (2-way ANOVA)
Borowicz et al., 2004Male ratsLoreclezole (SE)PB/PHT/VPA/CBZ/CLZ (SE)Free plasma levels of CBZ, other levels not tested[DOWNWARDS ARROW]ADD with VPA, [DOWNWARDS ARROW]ADD and SD with CLZ, [DOWNWARDS ARROW]ADD, SD and SS with CBZ, no effects in other combinations
Table 6. Audiogenic seizures (DBA/2 mice).
ReferenceCombination (s) testedPK interaction?Outcome
  1. In the presence of FBM, the ED50 of metoprolol decreased significantly with respect to the tonic phase alone of the seizures.

  2. ND, Not Determined

+De Sarro et al., 1996LTG (range)PB/PHT/VPA/DZP/CBZ (range)No change in total plasma levels of AEDs (LTG ND)Dose-orderly [DOWNWARDS ARROW]ED50 in all combinations
+De Sarro et al., 1998GBP (range)CBZ/DZP/FBM/LTG/PHT /PB/VPA (range)No change in total and free plasma levels of AEDs (GBP ND)[DOWNWARDS ARROW]ED50 for all combinations, except GBP + LTG
+De Sarro et al., 2000a7-NI (range)CBZ/DZP/LTG/PHT/PB /VPA (range)No change in total and free plasma levels of AEDs (7-NI ND)[DOWNWARDS ARROW]ED50 consistent with DZP/PB/VPA, N.S. with CBZ/PHT/LTG
+De Sarro et al., 2000bTPM (range)CBZ/DZP/FBM/LTG/PHT /PB/VPA (range)No change in total and free plasma levels of AEDs (TPM ND)[DOWNWARDS ARROW]ED50 in all combinations
+De Sarro et al., 2000cDCS (range)CBZ/DZP/FBM/LTG/PHT /PB/VPA (range)No change in total and free plasma levels of AEDs (DCS ND)Consistent [DOWNWARDS ARROW]ED50 with DZP/PB/PHT/VPA, N.S. with CBZ/FBM/LTG
+De Sarro et al., 2001Retigabine (range)CBZ/DZP/FBM/LTG/PHT /PB/VPA (range)No change in total and free plasma levels of AEDs (retigabine ND)[DOWNWARDS ARROW]ED50 with DZP/PB/PHT/VPA/CBZ, inconsistent with FBM/LTG
+De Sarro et al., 2002bPropranolol (range)CBZ/DZP/FBM/LTG/PHT /PB/VPA (range)No change in total and free plasma levels of AEDs (propranolol ND)[DOWNWARDS ARROW]ED50 consistent with DZP/PB/VPA/LTG, N.S. with CBZ/FBM/PHT
+Metoprolol (range) Not excluded[DOWNWARDS ARROW]ED50 consistent with DZP/PB/VPA/LTG, N.S. with CBZ/FBM/PHT
+Atenolol (NE range) Not excludedNo change in ED50
+De Sarro et al., 2002aPPG (range) / ACPT-1 (range)CFM-2 / CPPene (range)Not excluded[DOWNWARDS ARROW]ED50 in all combinations
+Gareri et al., 2004Carbenoxolone (SE)CBZ (range)No change in free plasma levels of AEDs[DOWNWARDS ARROW]ED50 in all combinations

Table 2 lists the AEDs grouped by molecular targets. In almost every interaction experiment (91%), at least one of the agents in the combination has a specific mechanism of action in the sense that it is thought to act through a single primary molecular target. Specificity of action is essential for an evaluation that aims to identify the mechanisms of action. Among the agents acting through a single specific target, especially (a) NMDA-receptor antagonists, (b) modulators of GABAA-receptor function, (c) agents blocking Na+ or Ca2+ ion channels, and (d) agonists of adenosine-receptor subtypes were evaluated (Table 2). This table also lists additional mechanisms that were evaluated, but not studied in a frequency sufficient for quantitative evaluation.

Figure 1 shows that potential pharmacokinetic (PK) interactions were not studied in 292 (54%) of all retrieved interaction experiments. In 85% of the experiments studying a PK interaction, the concentrations of only one of the two administered agents were determined. The line graph in Fig. 1 shows that a PK interaction was more often detected when comparing brain with total and unbound plasma. In the experiments taking brain concentrations into account, a PK interaction was reported in 30% of the experiments, and this is significantly more often than in experiments based on total plasma concentrations (p < 0.003; Fisher's exact test). This is a strong indication that, in the evaluation of PD drug–drug interactions, PK interactions at the level of the brain distribution must be taken into consideration.

image

Figure 1. The number of interaction experiments investigating a potential pharmacokinetic (PK) interaction for all retrieved experiments. The line graph shows the percentage of experiments detecting a PK interaction.

Download figure to PowerPoint

Selection of studies

A selection was made of all retrieved interaction experiments according to the protocol shown in Table 1. Study designs were classified by the animal model that was used in the experiment, the dosing strategy, and the method of interaction analysis. Based on this classification, flawed and nonstandard study designs were identified that could bias a group-wise comparison of study results, leading to the exclusion from analysis of 157 experiments of the total of 536. This leaves 379 (70%) experiments that were used in the evaluation of the interaction mechanisms contributing to synergy. The selection of these experiments is discussed later.

Animal models

From the retrieved studies, four groups of established animal models emerged based on the method of seizure induction (Fig. 2). By far the most frequently used animal model was the maximal electroshock (MES) model in mice. In these experiments, an electric current of fixed intensity and duration is applied via ear-clip or corneal electrodes, resulting in tonic hindlimb extension and flexion that is followed by clonus. In 53 of 280 electroshock experiments, rats were used, or threshold stimulation applied, and these experiments were excluded from analysis because of the relatively small numbers involved. In 130 experiments, seizures were induced by systemic administration of a convulsant compound. Most commonly (in 100 experiments), pentylenetetrazole was given subcutaneously to mice in a fixed dose that induces, under control conditions, generalized convulsions in 97% of the animals. Other agents, primarily bicuculline, were used less frequently, and the results obtained with this convulsant are therefore not included in the present analysis.

image

Figure 2. Primary animal models used in the study of anticonvulsant drug combinations. The bars show the number of interaction experiments that were performed in four different established models (light) and in variations thereon (dark). Two experiments that were performed in a genetic epilepsy model (Kaminski et al., 2001b; Bouwman et al., 2004) are not included here.

Download figure to PowerPoint

A third established model is the amygdala kindling model in the rat (51 experiments); in this model, the animals are repeatedly stimulated before the drug experiment. This results in enhanced electrical excitability due to long-term plastic changes that are the hallmark of kindling models. With 72 experiments, the fourth established model is a genetic model that uses the DBA/2 strain of mice, in which seizures can be induced by a loud sound. All retrieved experiments in this model were included in the analysis because they were performed within the same research group and according to the same methods (Table 6).

Dosing of individual agents

For the interpretation of interaction studies, it is necessary to measure not only the combined drug response, but also the response to the individual drugs. In the absence of individual response measurements at multiple doses or concentrations, the observed combined response cannot be evaluated in a quantitative manner. Figure 5 in Appendix A outlines how the design of experiments determines which methods of interaction analysis are applicable. Experiments determining the individual drug effects at only one or two doses of both agents were excluded from analysis; this was the case for 66 (12%) interaction experiments. In the remaining experiments, the individual drug effects were determined at multiple doses of both agents (61%) or at multiple doses of one of the agents (27%).

image

Figure 5. Decision scheme for interpreting the outcome of interaction experiments. The interpretation of an experiment depends on the number of doses tested and on whether a response was observed to none, one, or both of the single drugs.

Download figure to PowerPoint

PD interaction analysis

The most commonly used interpretation of experimental results relied on the change in ED50 (i.e., the dose corresponding to the half-maximal response). In 307 (57%) interaction experiments, the degree of potentiation by a threshold dose of the other agent was determined. This method is at best semiquantitative because the observation of a change in ED50 depends on the threshold dose that is chosen (Fig. 7).

image

Figure 7. The relative potency of the additive response calculated by using the method of concentration–addition shown in Fig. 6. The horizontal axis represents the normalized response to the drug administered in a fixed concentration. Along the vertical axis, the EC50 of the drug combination is shown relative to the EC50 of the drug alone; therefore values <1 indicate potentiation of the response. Calculations were based on the sigmoidal function shown in the inset, with different values for EB and curve steepness (nH).

Download figure to PowerPoint

A second, cruder interpretation of drug interactions is based on changes in the magnitude of response to a fixed dose of one drug under the influence of a second drug. This approach was used in 166 (31%) experiments, mostly determining the effect of a threshold dose of one drug on the response to a single dose of second drug. An important question for the interpretation of these experiments is whether at least one of the agents was tested individually at multiple doses. In 49 (9%) experiments, one of the drugs was found to be ineffective at all doses tested, and in these experiments, any increase in response to the drug combination indicates synergy (Fig. 5). A clear example, published after completion of our data analysis, is the enhancement of the anticonvulsant effect of topiramate (TPM), CBZ, and OXC by levetiracetam (LEV) in the MES test, in which LEV itself is inactive (Luszczki et al., 2006). Conversely, if both agents are individually effective or if this was not determined, interpretation of the results is not straightforward. This becomes clear if we imagine combining the ED25 of an agent with the same dose of this agent itself; in the case of a sigmoid-shaped dose–response relation, the combined response will exceed the half-maximal response. Altogether, 109 (20%) experiments did not take this into account and were therefore excluded from the analysis. Another 65 (12%) interaction experiments used various methods to take into account the dose-dependency of responses to the individual agents. These methods included (a) isobolographic analysis, (b) fractional effective concentrations, (c) two-way analysis of variance, and (d) linear logistic models. All of these methods are discussed in more detail in Appendix A.

Study outcomes in experiments fulfilling the selection criteria

Overall results

The influence of study methods on the outcome of the experiment was evaluated by logistic regression analysis. Figure 3 shows the number of experiments finding synergy for the method of interaction analysis, the animal model, and the method for evaluating a potential PK interaction. Experiments in which the degree of synergy was determined semiquantitatively by assessment of the ED50 value arrived significantly more often at a positive result (synergy) compared with quantitative evaluation methods (p < 0.001, ANOVA). No significant differences were found between animal models in the propensity for the finding of synergy. Conversely, the method used for taking into account the possibility of a PK interaction significantly influenced the outcome of interaction experiments (p < 0.001, ANOVA). Of the 65 experiments in which total plasma concentrations were measured, 63 experiments arrived at a conclusion of synergy. This result may, however, be confounded by the method of interaction analysis, because 63 of the 65 experiments were performed by assessment of the ED50. In the 27 experiments measuring drug concentrations in the brain, synergy was significantly less often found, and this may be due to the relatively more frequent detection of a PK interaction in these experiments.

image

Figure 3. Outcomes of selected experiments (n = 379). Dark bars represent the number of experiments that found synergism; light bars represent the number of experiments reporting additivity or antagonism. The corresponding regression estimates are shown to the right side, with values less than zero indicating a negative effect on the probability of synergism. Error bars represent 95% confidence intervals.

Download figure to PowerPoint

Specific molecular targets

Drug combinations also were evaluated according to their postulated molecular targets when ≥10 experiments were available for a combination of molecular targets. According to this criterion, 11 combinations of mechanisms of action were evaluated (Fig. 4), amounting to 244 experiments (45% of all retrieved experiments). This subset of the data was evaluated with a logistic regression analysis to identify synergistic molecular targets. The regression model included three factor variables: the method of interaction analysis, the method for evaluating a potential PK interaction, and the molecular targets. Figure 4 shows that the first two of these factors influenced the study outcome significantly, and this is in agreement with the regression analysis that was performed on all selected data. Conversely, the influence of the molecular targets that were combined on study outcome only approached statistical significance (p = 0.07, ANOVA). The present analysis does not provide conclusive evidence that AED combinations including a drug with a specific molecular target are likely to be more effective than combinations of drugs that each have multiple molecular targets.

image

Figure 4. Outcomes of experiments with combinations including specific molecular targets (n = 244). The dark bars represent the number of experiments that found synergism; light bars represent the number of experiments reporting additivity or antagonism. To the right side, values of the regression estimates less than zero indicate a negative effect on the probability of synergism. Error bars represent 95% confidence intervals.

Download figure to PowerPoint

Novel molecular targets

The AMPA receptor and the presynaptic glutamate receptors represent relatively novel drug targets for seizure suppression and have not been studied as extensively as other molecular targets. Antagonists of the AMPA receptors have not been studied with any particular target in ≥10 experiments, and these experiments were therefore not included in the statistical data analysis. Nevertheless, the results obtained with antagonists of the AMPA receptor are noteworthy, because in a total of 20 experiments, 19 experiments reported synergism and only one experiment reported additivity. AMPA-receptor antagonists were combined most frequently with agents acting through multiple targets (seven experiments) and with Na+-channel blockers (six experiments). The success rate in the 20 experiments cited was above average, even when taking into consideration that all but one of these experiments were conducted by determination of a change in ED50. The percentage of ED50 experiments with AMPA-receptor antagonists finding synergism was 95%, and this value was 57% in ED50 experiments not including these agents (p = 0.003 for the difference, χ2 test).

Selective agonists of presynaptic glutamate-receptor subtypes represent another novel target for AED action. It is not yet possible to assess the potential of these agents for synergism because only seven experiments were retrieved, five reporting synergism, and the remaining two, additivity. In four experiments in a recent study, two group III glutamate-receptor agonists potentiated the anticonvulsant effect of both AMPA- and NMDA-receptor block in DBA/2 mice (De Sarro et al., 2002a). These results are encouraging but remain to be confirmed in other seizure models, preferably by using a quantitative method of interaction analysis.

DISCUSSION

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS
  5. DISCUSSION
  6. Appendix
  7. Acknowledgments
  8. REFERENCES

The combination of AEDs is believed to be more likely to be beneficial when drugs are used that differ in their molecular action. In the present literature analysis, 536 animal experiments were evaluated to determine whether preclinical studies support the hypothesis that AEDs may be combined rationally on a mechanistic basis. These experiments form a heterogeneous group in which the study design was found to influence the study outcome. Synergism was significantly more often found in threshold experiments compared with strictly quantitative methods. Although accounting for this difference, no single combination of mechanisms was significantly more, or less, frequently concluded to be synergistic compared with other combinations. However, when considering all tested drug combinations including an AMPA-receptor antagonist, an increased probability for synergism was found.

These findings unfortunately do not constitute a broad mechanistic basis for the selection of drug combinations that are synergistic with respect to their anticonvulsant action. It is useful to consider the limitations that are encountered in the interpretation of preclinical studies with drug combinations. In the present analysis, only the anticonvulsant effects of drug combinations were evaluated and not their toxicity. It would be of interest to take into account toxicity measures, such as the rotorod test, to derive a therapeutic index from the ratio TD50/ED50, where TD50 represents the dose resulting in half-maximal toxicity. No attempt, however, has been made to demonstrate the correspondence of this index to the tolerability of AEDs in humans, and toxicity was evaluated in only a part of all retrieved experiments. The complexity of this topic deserves to be covered in a separate review, and the reader interested in the application of a therapeutic index for interpreting drug interactions is referred to Luszczki et al. (2003c).

The retrieved studies were, quite understandably, conducted exclusively with combinations of drugs that possess differing molecular targets. It is therefore not possible to determine whether such combinations offer an additional advantage over combinations of drugs with an identical action. It was possible, however, to compare the probability of synergism between those drug combinations that include agents with multiple molecular targets and those that include only more specifically acting drugs; no difference was observed. It must be emphasized that the comparison of drug combinations on a mechanistic basis is limited to the current knowledge of the molecular targets involved. These insights will certainly change in the future with the discovery of new mechanisms of drug action and with a better understanding of the functioning of known molecular targets. The interpretation of drug-interaction studies is further complicated by the widely differing study protocols used. Most important, the animal model used, the dose range that is studied, the PK analysis, and the method of interpretation all differ between studies. All of these factors were taken into account in the present analysis, and their impact on the outcome of interaction experiments is discussed later.

Animal models

Activity in the common screening models for AEDs is a good predictor for efficacy in the treatment of seizure disorders in humans (White, 1999). Because AEDs with varying molecular targets are active in the established animal models for epilepsy, it can be argued that these “mechanism-neutral” models are likely to reflect various mechanisms underlying combined drug action. This may well explain why, in the present analysis, no significant differences were found between animal models with respect to their general tendency for predicting synergism. On the level of individual drug combinations, however, the comparison of activity in different animal models may still be useful. For example, in the MES model and in the threshold pentylenetetrazole (PTZ) test, nonepileptic rodents are used, whereas amygdala kindling and the induction of seizures in DBA/2 mice are tests in rodents with increased seizure susceptibility or genuine epilepsy (Meldrum, 2002). The DBA/2 model is notable in that it reflects the activity of all AEDs tested so far, suggesting that this model may be sensitive to a wide range of antiepileptic mechanisms. MES and PTZ tests seem to reflect primarily an action on voltage-sensitive Na+ channels and potentiation of GABAA-mediated inhibition, respectively, although other mechanisms can also produce activity in the PTZ test (Meldrum, 2002). However, with none of these animal models was a dependence of the study outcome on the mechanisms of action of the evaluated drugs apparent. Other experimental epilepsy models that more closely reflect epileptogenesis and pharmacoresistance development in refractory types of epilepsy in humans are needed to demonstrate such actions (Stables et al., 2002).

Dose range

With the exception of two experiments (Gennings et al., 1995; Jonker et al., 2004), all retrieved drug interactions were characterized at only a single or a few doses or concentrations. In experiments detecting no synergy, the possibility therefore exists that synergy is present, but at doses that were not studied. From the experiments in which synergy is found, it is not clear if synergy is also present at other doses and whether the dose may be optimized to obtain a higher degree of synergy. In a study using four different subthreshold dose levels of the AMPA/kainate-receptor antagonist LY300164 (Czuczwar et al., 1998), a dose-dependent potentiation of the anticonvulsant effect of conventional AEDs was observed in the MES test in mice, with the lowest doses of LY300164 being without effect on the ED50 values. In the majority of the retrieved interaction experiments, drug interactions were characterized by intraperitoneal administration of drugs, followed by a single response measurement. It should be noted that with this study design, an exceedingly large number of animals would be required to characterize a drug interaction over the full effective dose range of both drugs.

However, advances in the design of interaction studies and improved statistical models have made the assessment feasible of complex in vivo PD drug interactions (Jonker et al., 2005). A recent study from our laboratory illustrates that the resources required for a full interaction analysis can remain relatively limited through the application of PK/PD modeling (Jonker et al., 2005). In this study, the interaction between tiagabine (TGB) and lamotrigine (LTG) was fully characterized in 40 rats, by using counts of ictal signs as a measure of drug response. With intravenous drug infusion, a considerable variability in drug concentrations was observed. Through repeated measurement of both concentration and effect in individual animals, the drug interaction could be evaluated with much higher precision than would have been possible with measurements of only a single response per animal and without characterizing the time course of drug concentrations in that animal. This study also demonstrated that variability in drug concentrations can actually be exploited to facilitate the full characterization of a drug–drug interaction instead of hindering its interpretation.

Pharmacokinetic interaction

The PK/PD approach sketched earlier has the additional advantage that a potential PK interaction is taken into account. PD interactions are inseparable from the PK processes that govern the time course of drug action (el-Mashri et al., 1997). A change in the plasma concentration profile or a change in the permeability of the blood–brain barrier due to a PK interaction influences the magnitude of the response to a drug combination. Thus an apparent synergistic effect may in reality be caused by an increased concentration (e.g., due to liver enzyme inhibition). Ideally, the combined response is related to the concentrations at the effect site (the brain). Measurement of unbound plasma concentrations may have very little added value because changes in protein binding have relevance in only a very limited number of cases (Benet and Hoener, 2002). In the present analysis, PK interactions were significantly more often observed when studied at the level of brain concentrations compared with the level of total or unbound plasma concentrations. However, effect-site concentrations are in practice not directly accessible, and the measurement of whole brain concentrations does not reveal concentration differences between brain regions. By simultaneous measurement of drug concentrations and response, it is possible to characterize the time delay between the two entities and thereby obtain an estimate of effect-site concentrations. The most efficient way to achieve this is in animal models that allow repeated measurement of the response in conjunction with repeated blood sampling (e.g., the cortical stimulation model in rats) (Voskuyl et al., 1992).

In most experiments included in the present analysis, PK interactions were evaluated not simultaneously but in a separate experiment, and only for drug combinations that appear to interact synergistically, based on response data. However, a drug interaction may appear additive because of a decrease in drug concentrations, whereas it is really a synergistic effect. This necessitates the routine evaluation of PK interactions.

Methods of interaction analysis

Although the isobolographic analysis is considered the gold standard for the evaluation of drug combinations (Tallarida, 2001), the majority of drug interactions were evaluated by determining the influence of a fixed dose of one compound on the potency (ED50 and sometimes EC50) of the other compound. Synergy was significantly more often found in these experiments compared with experiments using other methods, including isobolographic analysis. This finding is in line with a study of the interaction between the glutamatergic neurotransmission inhibitor riluzole and four AEDs (CBZ, PB, VPA, and diphenylhydantoin) using the MES test in mice (Borowicz et al., 2004). In this study, riluzole decreased the ED50 value of each of the four AEDs when administered at a threshold dose in a dose-dependent fashion, whereas an isobolographic analysis indicated an additive interaction for each of the four drug combinations. A statistically significant decrease in ED50 value has commonly been interpreted as indicative of a favorable interaction with disregard for the magnitude of the decrease. In the study on riluzole (Borowicz et al., 2004), the smallest statistically significant response was a decrease in the ED50 from a value of 274 mg/kg to a value of 245 mg/kg. The mere demonstration of statistical significance at the 5% level for a decrease in ED50 value in an animal study is insufficient for arguing the potential clinical benefit of a drug combination.

In addition, experiments interpreted with quantitative methods like isobolographic analysis take into account the dose–response relations of both individual agents. The criterion for synergy is less strict in experiments using determination of ED50 values, because the dose–response relation of only one of the individual agents is taken into account. Although the present evaluation was not designed to assess the validity of interpreting changes in the ED50 value, the finding that the outcomes of drug-interaction studies relying on this method are considerably more often positive than are studies using other methods, highlights the problem of relying on statistical significance alone for interpreting study results.

To conclude, a mechanistic basis?

From the present analysis, only the AMPA-receptor antagonists emerged as a drug class that is likely to confer synergism. Because these compounds have in general high toxicities, their most appropriate use may be in combination therapy exclusively (Rogawski and Donevan, 1999). Current research methods have their limitations, and this may have hampered the search for effective combinations. In view of the fact that multiple target sites are attributed to the action of most marketed AEDs, a search for compounds that act with high specificity at a single target site may not be relevant for the development of clinically useful new and more efficacious AEDs (Rogawski, 1998). At present, a greater need exists for studies on interactions of the newer AEDs (Luszczki and Czuczwar, 2004b), both because such combinations will inevitably appear in clinical practice and because favorable interactions may emerge as their mechanisms of actions may differ from the older AEDs. Also, the combination of different mechanisms into one molecule is to be preferred for greater ease of use in practice (Schmidt et al., 1997; White, 1997).

However, for the identification of synergistically interacting target sites, highly selective acting agents are indispensable. Only with this knowledge may it become possible to develop AEDs that activate the optimal combination of target sites with the optimal intensity. The use of good and efficient study designs, analysis of potential PK interactions, and application of good quantitative methods for data analysis is indispensable as well. All these ingredients are available but should be further extended and optimized.

Appendix

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS
  5. DISCUSSION
  6. Appendix
  7. Acknowledgments
  8. REFERENCES

APPENDIX A

Making the distinction between synergy, additivity, and antagonism

A variety of methods for interpretation of the responses to drug combinations was encountered in the present literature study. For an exhaustive review of evaluation methods, the interested reader is referred to (Berenbaum, 1989). More recently, the quantitative assessment of the effects of combined drugs was discussed in (Tallarida, 2001). In the following, the most frequently encountered methods are only briefly discussed, starting from the information that will be available from a given study design. Based on whether the dose–response relations of none, one, or both of the individual agents are characterized, methods for interaction analysis can be classified according to the scheme shown in Fig. 5. When one of the agents is not effective at any dose, an increase in the response to the second agent indicates a favorable interaction. Conversely, when both agents are individually effective at some concentrations, the question is raised as to how to define the additive response to the combination.

For assessing the effect of one agent on the concentration–response relation of a second agent, the method of concentration–addition is a useful approach (Trendelenburg, 1962). The application of this method is demonstrated in Fig. 6. Drug B is present at a fixed concentration that results in the response EB, and this response is associated with a unique concentration C'A. This concentration is added to the concentration–response relation of A to construct the additive response to A in the presence of B. In the case that EB equals zero, however, a unique solution for C'A does not exist. This hampers the interpretation of experiments that assess the effect of a threshold dose of agent B on the EC50 of agent A. It has been argued on basis of the method of concentration–addition that the EC50 value decreases in the presence of a threshold dose when a drug interaction is additive (Deckers et al., 2000). However, when the influence of EB and the Hill slope is taken into consideration, it becomes clear that both decreases and increases of the EC50 value are possible (Fig. 7). Importantly, the EC50 value increases with EB when the Hill slope is unity, and therefore the observation of a decrease in EC50 indicates synergism under this condition. Conversely, when the concentration–response relation is steep, the EC50 value decreases substantially at very low values of EB. Under this condition, the observation of a decrease in EC50 should be interpreted with caution, because it does not necessarily indicate synergism. Figure 7 is useful as a reference for the interpretation of experimental observations but could not be used in the present analysis because the Hill slope was, in most studies, not reported.

image

Figure 6. Method of concentration–addition for calculating the response to drug A in the presence of drug B (EA,B). C'A is the concentration of drug A corresponding to the response achieved with CB when administered alone. EA,B is obtained by adding C'A to CA. Note that the apparent change in the horizontal distance between the curves is due to the use of a logarithmic scale. From Jonker et al., 2005.

Download figure to PowerPoint

A limitation to the EC50 method is that it characterizes the effect of one drug on the other, but not the reverse. It is possible that the result obtained with the EC50 method depends on which drug is fixed to a threshold concentration and which drug concentration is varied, especially when the concentration–response relations of the individual drugs differ in Hill factor or maximal response. In contrast, isobolographic analysis is bidirectional and based on the definition of Loewe additivity (Loewe, 1953), shown in Fig. 8. In this figure, linearity of the iso-effect curve indicates Loewe additivity. Linearity is directly assessed by plotting concentrations for the individual drugs and the combination that elicit a selected level of response, but in practice, such observations may be difficult to obtain. More practical is to determine the concentration–response relation for a drug combination that is administered in a fixed proportion of concentrations; such data can also be interpreted on the basis of Loewe additivity (Tallarida, 2001). The older literature that uses similar experimental designs often refers to the fractional effective concentration (FEC) (Elion et al., 1954). The FEC of a drug is the concentration ratio given in the equation for Loewe additivity that is shown in Fig. 8. The FEC index of a drug combination is the sum of the FEC values of the individual drugs at a selected response level, with values close to 1 signifying additivity. An index of 0.7–1.3 is considered to indicate an additive interaction; values <0.7 indicate potentiation, and values >1.3 indicate infraadditive interaction (Bourgeois and Wad, 1988). For the assessment of synergy, additivity, and antagonism at multiple response levels, a statistical procedure is available that is based on linear regression (Tallarida et al., 1997).

image

Figure 8. Outcomes of an isobolographic analysis. The shape of the isobole is determined by the concentration pairs that correspond to a given magnitude of the response. C'A and C'B indicate the concentrations of A and B that elicit this response when these drugs are present alone. From Jonker et al., 2005.

Download figure to PowerPoint

A limitation to the isobolographic approach is that it cannot be fully used when two drugs differ in the maximal response. In that case, Loewe additivity is defined only within the range of possible effects of the least efficacious drug. We recently proposed an alternative approach that has general application and relies on determination of the maximum additive response (MAR) (Jonker et al., 2005). The MAR is governed by the most efficacious drug in a combination and is derived by concentration–addition of drug A to drug B and the reverse. This approach was recently applied to interpret the combined anticonvulsant effect of tiagabine and lamotrigine (Jonker et al., 2004).

The methods that were discussed so far have in common that they were developed from the viewpoint of interpreting combined drug responses. However, to this specific end, general-purpose statistical methods can also be used. Two such methods were encountered in the present literature analysis, logistic regression and two-way ANOVA. These methods estimate the responses to the individual drugs (the “main effects”) and the responses to the combination (the “interaction effect”). A significant main effect for a drug indicates that a relation between dose and response exists for that drug. A significant interaction effect indicates that the drug response to the doses of drug A in the combination depends on the doses of drug B and vice versa. The general form of the linear logistic regression models is: f (x) =β01× x12× x212× x12. Although these methods are powerful tools to classify the magnitude of drug interaction, the outcome of these analyses is not readily translated into a pharmacologic measure of response such as the change in EC50. In this respect, a fully parametric approach that was recently developed is of interest (Minto et al., 2000). In the case of a sigmoid Emax model for drug effect, this parameterization involves estimation of Emax, EC50, and nH as a function of the dose ratio of the drugs in the combination. An additional advantage of this method is that it is not constrained to a specific dose–response function and can also be applied if one of the drugs in the combination is inefficacious (Jonker et al.).

Acknowledgments

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS
  5. DISCUSSION
  6. Appendix
  7. Acknowledgments
  8. REFERENCES

Acknowledgment:  We appreciate the expert advice of Dr. P.H.C. Eilers regarding statistical evaluations.

REFERENCES

  1. Top of page
  2. Abstract
  3. METHODS
  4. RESULTS
  5. DISCUSSION
  6. Appendix
  7. Acknowledgments
  8. REFERENCES
  • Assi AA. (2001) N6-cyclohexyladenosine and 3- (2-carboxypiperazine-4-yl)-1-propenyl-1-phosphonic acid enhance the effect of antiepileptic drugs against induced seizures in mice. Journal of Pharmacy and Pharmaceutical Science 4:4251.
  • Balakrishnan S, Pandhi P, Bhargava VK. (1998) Effect of nimodipine on the efficacy of commonly used antiepileptic drugs in rats. Indian Journal of Experimental Biology 36:5154.
  • Benet LZ, Hoener B. (2002) Changes in plasma protein binding have little clinical relevance. Clinical Pharmacological Therapeutics 71:115121.
  • Berenbaum MC. (1989) What is synergy? Pharmacological Reviews 41:93141.
  • Borowicz KK, Gasior M, Kleinrok Z, Czuczwar SJ. (1995) The non-competitive AMPA/kainate receptor antagonist, GYKI 52466, potentiates the anticonvulsant activity of conventional antiepileptics. European Journal of Pharmacology 281:319326.
  • Borowicz KK, Gasior M, Kleinrok Z, Czuczwar SJ. (1996) Competitive NMDA-receptor antagonists, LY 235959 and LY 233053, enhance the protective efficacy of various antiepileptic drugs against maximal electroshock-induced seizures in mice. Epilepsia 37:618624.
  • Borowicz KK, Gasior M, Kleinrok Z, Czuczwar SJ. (1997a) Influence of isradipine, niguldipine and dantrolene on the anticonvulsive action of conventional antiepileptics in mice. European Journal of Pharmacology 323:4551.
  • Borowicz KK, Kleinrok Z, Czuczwar SJ. (1997b) N6-2- (4-aminophenyl)ethyl-adenosine enhances the anticonvulsive activity of antiepileptic drugs. European Journal of Pharmacology 327:125133.
  • Borowicz KK, Kleinrok Z, Czuczwar SJ. (2000a) 7-Nitroindazole differentially affects the anticonvulsant activity of antiepileptic drugs against amygdala-kindled seizures in rats. Epilepsia 41:11121118.
  • Borowicz KK, Kleinrok Z, Czuczwar SJ. (2000b) The AMPA/kainate receptor antagonist, LY 300164, increases the anticonvulsant effects of diazepam. Naunyn-Schmiedeberg's Archives of Pharmacology 361:629635.
  • Borowicz KK, Kleinrok Z, Czuczwar SJ. (2000c) N6-2- (4-aminophenyl)ethyl-adenosine enhances the anticonvulsive action of convential antiepileptic drugs in the kindling model of epilepsy in rats. European Neuropsychopharmacology 10:237243.
  • Borowicz KK, Duda AM, Kleinrok Z, Czuczwar SJ. (2001a) Interaction of GYKI 52466, a selective non-competitive antagonist of AMPA/kainate receptors, with conventional antiepileptic drugs in amygdala-kindled seizures in rats. Polish Journal of Pharmacology 53:101108.
  • Borowicz KK, Kleinrok Z, Czuczwar SJ. (2001b) Glutamate receptor antagonists differentially affect the protective activity of conventional antiepileptics against amygdala-kindled seizures in rats. European Neuropsychopharmacology 11:6168.
  • Borowicz KK, Kleinrok Z, Czuczwar SJ. (2002a) Niguldipine impairs the protective activity of carbamazepine and phenobarbital in amygdala-kindled seizures in rats. European Neuropsychopharmacol 12:225233.
  • Borowicz KK, Luszczki J, Czuczwar SJ. (2002b) 2-Chloroadenosine, a preferential agonist of adenosine A1 receptors, enhances the anticonvulsant activity of carbamazepine and clonazepam in mice. European Neuropsychopharmacol 12:173179.
  • Borowicz KK, Swiader M, Luszczki J, Czuczwar SJ. (2002c) Effect of gabapentin on the anticonvulsant activity of antiepileptic drugs against electroconvulsions in mice: an isobolographic analysis. Epilepsia 43:956963.
  • Borowicz KK, Zadrozniak M, Swiader M, Kowalska A, Kleinrok Z, Czuczwar SJ. (2002d) Interaction of the neurosteroid alphaxalone with conventional antiepileptic drugs in different types of experimental seizures. European Journal of Pharmacology 449:8590.
  • Borowicz KK, Piskorska B, Luszczki J, Czuczwar SJ. (2003) Influence of SIB 1893, a selective mGluR5 receptor antagonist, on the anticonvulsant activity of conventional antiepileptic drugs in two models of experimental epilepsy. Polish Journal of Pharmacology 55:735740.
  • Borowicz KK, Sawiniec A, Czuczwar SJ. (2004) Interaction of loreclezole with conventional antiepileptic drugs in amygdala-kindled rats. European Neuropsychopharmacol 14:251257.
  • Bourgeois BF. (1986) Antiepileptic drug combinations and experimental background: the case of phenobarbital and phenytoin. Naunyn Schmiedebergs Archives of Pharmacology 333:406411.
  • Bourgeois BF. (1988) Combination of valproate and ethosuximide: antiepileptic and neurotoxic interaction. Journal of Pharmacological Experimental Therapeutics 247:11281132.
  • Bourgeois BF, Wad N. (1988) Combined administration of carbamazepine and phenobarbital: effect on anticonvulsant activity and neurotoxicity. Epilepsia 29:482487.
  • Bourgeois BF, Dodson WE, Ferrendelli JA. (1983a) Potentiation of the antiepileptic activity of phenobarbital by nicotinamide. Epilepsia 24:238244.
  • Bourgeois BF, Dodson WE, Ferrendelli JA. (1983b) Primidone, phenobarbital, and PEMA: II. Seizure protection, neurotoxicity, and therapeutic index of varying combinations in mice. Neurology 33:291295.
  • Bouwman BM, Heesen E, Van Rijn CM. (2004) The interaction between vigabatrin and diazepam on the electroencephalogram during active behaviour in rats: an isobolic analysis. European Journal of Pharmacology 495:119128.
  • Brodie MJ. (1992) Drug interactions in epilepsy. Epilepsia 33 (Suppl 1):S1322.
  • Chez MG, Bourgeois BF, Pippenger CE, Knowles WD. (1994) Pharmacodynamic interactions between phenytoin and valproate: individual and combined antiepileptic and neurotoxic actions in mice. Clinical Neuropharmacology 17:3237.
  • Chweh AY, Swinyard EA, Wolf HH, Kupferberg HJ. (1985) Effect of GABA agonists on the neurotoxicity and anticonvulsant activity of benzodiazepines. Life Science 36:737744.
  • Cuadrado A, De Las Cuevas I, Valdizán EM, Armijo JA. (2002) Synergistic interaction between valproate and lamotrigine against seizures induced by 4-aminopyridine and pentylenetetrazole in mice. European Journal of Pharmacology 453:4352.
  • Czechowska G, Dziki M, Pietrasiewicz T, Kleinrok Z, Turski WA, Czuczwar SJ. (1993) Competitive antagonists of NMDA receptors, CGP 37849 and CGP 39551, enhance the anticonvulsant activity of valproate against electroconvulsions in mice. European Journal of Pharmacology 232:5964.
  • Czuczwar SJ, Borowicz KK. (2002) Polytherapy in epilepsy: the experimental evidence. Epilepsy Research 52:1523.
  • Czuczwar SJ, Turski L, Kleinrok Z. (1981) Diphenylhydantoin potentiates the protective effect of diazepam against pentylenetetrazol but not against bicuculline and isoniazid-induced seizures in mice. Neuropharmacology 20:675679.
  • Czuczwar SJ, Turski L, Kleinrok Z. (1982) Effects of combined treatment with diphenylhydantoin and different benzodiazepines on pentylenetetrazol- and bicuculline-induced seizures in mice. Neuropharmacology 21:563567.
  • Czuczwar SJ, Chmielewska B, Turski WA, Kleinrok Z. (1984a) Differential effects of baclofen, gamma-hydroxybutyric acid and muscimol on the protective action of phenobarbital and diphenylhydantoin against maximal electroshock-induced seizures in mice. Neuropharmacology 23:159163.
  • Czuczwar SJ, Turski L, Schwarz M, Turski WA, Kleinrok Z. (1984b) Effects of excitatory amino-acid antagonists on the anticonvulsant action of phenobarbital or diphenylhydantoin in mice. European Journal of Pharmacology 100:357362.
  • Czuczwar SJ, Malek U, Kleinrok Z. (1990a) Influence of calcium channel inhibitors upon the anticonvulsant efficacy of common antiepileptics against pentylenetetrazol-induced convulsions in mice. Neuropharmacology 29:943948.
  • Czuczwar SJ, Szczepanik B, Wamil A, Janusz W, Kleinrok Z. (1990b) Differential effects of agents enhancing purinergic transmission upon the antielectroshock efficacy of carbamazepine, diphenylhydantoin, diazepam, phenobarbital, and valproate in mice. Journal of Neural Transmission. General Section 81:153166.
  • Czuczwar SJ, Borowicz KK, Kleinrok Z, Tutka P, Zarnowski T, Turski WA. (1995) Influence of combined treatment with NMDA and non-NMDA receptor antagonists on electroconvulsions in mice. European Journal of Pharmacology 281:327333.
  • Czuczwar SJ, Swiader M, Kuzniar H, Gasior M, Kleinrok Z. (1998) LY 300164, a novel antagonist of AMPA/kainate receptors, potentiates the anticonvulsive activity of antiepileptic drugs. European Journal of Pharmacology 359:103109.
  • Dalby NO. (2000) GABA-level increasing and anticonvulsant effects of three different GABA uptake inhibitors. Neuropharmacology 39:23992407.
  • De Sarro G, Nava F, Aguglia U, De Sarro A. (1996) Lamotrigine potentiates the antiseizure activity of some anticonvulsants in DBA/2 mice. Neuropharmacology 35:153158.
  • De Sarro G, Spagnolo C, Gareri P, Gallelli L, De Sarro A. (1998) Gabapentin potentiates the antiseizure activity of certain anticonvulsants in DBA/2 mice. European Journal of Pharmacology 349:179185.
  • De Sarro G, Gareri P, Falconi U, De Sarro A. (2000a) 7-Nitroindazole potentiates the antiseizure activity of some anticonvulsants in DBA/2 mice. European Journal of Pharmacology 394:275288.
  • De Sarro G, Gratteri S, Bonacci F, Musumeci SA, Elia M, De Sarro A. (2000b) Topiramate potentiates the antiseizure activity of some anticonvulsants in DBA/2 mice. European Journal of Pharmacology 388:163170.
  • De Sarro G, Gratteri S, Naccari F, Pasculli MP, De Sarro A. (2000c) Influence of D-cycloserine on the anticonvulsant activity of some antiepileptic drugs against audiogenic seizures in DBA/2 mice. Epilepsy Research 40:109121.
  • De Sarro G, Di Paola ED, Conte G, Pasculli MP, De Sarro A. (2001) Influence of retigabine on the anticonvulsant activity of some antiepileptic drugs against audiogenic seizures in DBA/2 mice. Naunyn Schmiedeberg's Archives of Pharmacology 363:330336.
  • De Sarro G, Chimirri A, Meldrum BS. (2002a) Group III mGlu receptor agonists potentiate the anticonvulsant effect of AMPA and NMDA receptor block. European Journal of Pharmacology 451:5561.
  • De Sarro G, Di Paola ED, Ferreri G, De Sarro A, Fischer W. (2002b) Influence of some ß-adrenoceptor antagonists on the anticonvulsant potency of antiepileptic drugs against audiogenic seizures in DBA/2 mice. European Journal of Pharmacology 442:205213.
  • Deckers CLP, Czuczwar SJ, Hekster YA, Keyser A, Kubova H, Meinardi H, Patsalos PN, Renier WO, Van Rijn CM. (2000) Selection of antiepileptic drug polytherapy based on mechanisms of action: the evidence reviewed. Epilepsia 41:13641374.
  • Della Paschoa OE, Kruk MR, Hamstra R, Voskuyl RA, Danhof M. (1998) Pharmacodynamic interaction between phenytoin and sodium valproate changes seizure thresholds and pattern. British Journal of Pharmacology 125:9971004.
  • Deutsch SI, Park CH, Lukacs LG, Morn C, Koetzner L, Mastropaolo J. (1995) MK-801 alters the GABAA receptor complex and potentiates flurazepam's antiseizure efficacy. Pharmacology, Biochemistry and Behavior 51:909915.
  • Deutsch SI, Rosse RB, Steinberg K, Morn C, Koetzner L, Riggs R, Mastropaolo J. (1996) Evaluation of in vivo interactions in mice between flurazepam and two neuroactive steroids. Pharmacology, Biochemistry and Behavior 55:323326.
  • Dziki M, Honack D, Loscher W. (1992) Kindled rats are more sensitive than non-kindled rats to the behavioural effects of combined treatment with MK-801 and valproate. European Journal of Pharmacology 222:273278.
  • Ebert U, Wlaz P, Löscher W. (1997) Anticonvulsant effects by combined treatment with a glycine (B) receptor antagonist and a polyamine site antagonist in amygdala-kindled rats. European Journal of Pharmacology 322:179184.
  • Elion GB, Singer S, Hitchings GH. (1954) Antagonists of nucleic acid derivatives: Part VIII. Synergism in combinations of biochemically related antimetabolites. Journao of Biological Chemistry 208:477488.
  • El-Mashri HA, Reardon KF, Yang RS. (1997) Integrated approaches for the analysis of toxicologic interactions of chemical mixtures. Critical Review of Toxicology 27:175197.
  • Gareri P, Condorelli D, Belluardo N, Gratteri S, Ferreri G, Donato Di Paola E, De Sarro A, De Sarro G. (2004) Influence of carbenoxolone on the anticonvulsant efficacy of conventional antiepileptic drugs against audiogenic seizures in DBA/2 mice. European Journal of Pharmacology 484:4956.
  • Gasior M, Borowicz K, Starownik R, Kleinrok Z, Czuczwar SJ. (1996a) Ca2+ channel blockade and the antielectroshock activity of NMDA receptor antagonists, CGP 40116 and CGP 43487, in mice. European Journal of Pharmacology 312:2733.
  • Gasior M, Kaminski R, Brudniak T, Kleinrok Z, Czuczwar SJ. (1996b) Influence of nicardipine, nimodipine and flunarizine on the anticonvulsant efficacy of antiepileptics against pentylenetetrazol in mice. Journal of Neural Transmission, General Section 103:819831.
  • Gasior M, Carter RB, Goldberg SR, Witkin JM. (1997) Anticonvulsant and behavioral effects of neuroactive steroids alone and in conjunction with diazepam. Journal of Pharmacology and Experimental Therapeutics 282:543553.
  • Gasior M, Swiader M, Przybylko M, Borowicz K, Turski WA, Kleinrok Z, Czuczwar SJ. (1998) Felbamate demonstrates low propensity for interaction with methylxanthines and Ca2+ channel modulators against experimental seizures in mice. European Journal of Pharmacology 352:207214.
  • Gennings C, Sofia RD, Carchman RA, Carter WH Jr, Swinyard EA. (1995) Analysis of anticonvulsant and neurotoxic responses to combination therapy with carbamazepine, felbamate and phenytoin by response-surface modeling. Arzneimittelforschung 45:739748.
  • Gordon R, Gels M, Diamantis W, Sofia RD. (1991) Interaction of felbamate and diazepam against maximal electroshock seizures and chemoconvulsants in mice. Pharmacology, Biochemistry and Behavior 40:109113.
  • Gordon R, Gels M, Wichmann J, Diamantis W, Sofia RD. (1993) Interaction of felbamate with several other antiepileptic drugs against seizures induced by maximal electroshock in mice. Epilepsia 34:367371.
  • Hawkins CA, Mellanby JH. (1986) Piracetam potentiates the antiepileptic action of carbamazepine in chronic experimental limbic epilepsy. Acta Neurologica Scandinavica suppl 109:117121.
  • Jonker DM, Voskuyl RA, Danhof M. (2004) Pharmacodynamic analysis of the anticonvulsant effects of tiagabine and lamotrigine in combination in the rat. Epilepsia 45:424435.
  • Jonker DM, Visser SAG, Van der Graaf PH, Voskuyl RA, Danhof M. (2005) Towards a mechanism-based analysis of pharmacodynamic drug-drug interactions in vivo. Pharmacology and Therapeutics 106:118.
  • Joseph S, David J, Joseph T. (1998) Synergistic effect of flunarizine and sodium valproate on seizure thresholds elicited by cortical stimulation in conscious rats. Indian Journal of Physiology and Pharmacology 42:3949.
  • Kaminski RM, Mazurek M, Turski WA, Kleinrok Z, Czuczwar SJ. (2001a) Amlodipine enhances the activity of antiepileptic drugs against pentylenetetrazole-induced seizures. Pharmacology, Biochemistry and Behavior 68:661668.
  • Kaminski RM, Van Rijn CM, Turski WA, Czuczwar SJ, Van Luijtelaar G. (2001b) AMPA and GABA-B receptor antagonists and their interaction in rats with a genetic form of absence epilepsy. European Journal of Pharmacology 430:251259.
  • Kleinrok Z, Czuczwar SJ, Kozicka M. (1980) Effect of dopaminergic and GABA-ergic drugs given alone or in combination on the anticonvulsant action of phenobarbital and diphenylhydantoin in the electroshock test in mice. Epilepsia 21:519529.
  • Klodzinska A, Bijak M, Chojnacka-Wojcik E, Kroczka B, Swiader M, Czuczwar SJ, Pilc A. (2000) Roles of group II metabotropic glutamate receptors in modulation of seizure activity. Naunyn Schmiedeberg's Archives of Pharmacology 361:283288.
  • Kwan P, Brodie MJ. (2000) Early identification of refractory epilepsy. New England Journal of Medicine 342:314319.
  • Leppik IE, Sherwin AL. (1977) Anticonvulsant activity of phenobarbital and phenytoin in combination. Journal of Pharmacology and Experimental Therapeutics 200:570575.
  • Loewe S. (1953) The problem of synergism and antagonism of combined drugs. Arzneimittelforschung 3:285290.
  • Löscher W, Hönack D. (1993) Effects of the competitive NMDA receptor antagonist, CGP 37849, on anticonvulsant activity and adverse effects of valproate in amygdala-kindled rats. European Journal of Pharmacology 234:237245.
  • Löscher W, Honack D. (1994) Over-additive anticonvulsant effect of memantine and NBQX in kindled rats. European Journal of Pharmacology 259:R3-R5.
  • Löscher W, Rundfeldt C, Honack D. (1993) Low doses of NMDA receptor antagonists synergistically increase the anticonvulsant effect of the AMPA receptor antagonist NBQX in the kindling model of epilepsy. European Journal of Neuroscience 5:15451550.
  • Lu YQ, Yu R. (2001) Synergistic protection of allopregnanolone and phenobarbital against maximal electroshock seizures in mice. Acta Pharmacologica Sinite 22:361364.
  • Luchowska E, Luchowski P, Wielosz M, Kleinrok Z, Czuczwar SJ, Urbanska EM. (2002) Propranolol and metoprolol enhance the anticonvulsant action of valproate and diazepam against maximal electroshock. Pharmacology, Biochemistry and Behavior 71:223231.
  • Luszczki JJ, Czuczwar SJ. (2003) Isobolographic and subthreshold methods in the detection of interactions between oxcarbazepine and conventional antiepileptics-a comparative study. Epilepsy Research 56:2742.
  • Luszczki JJ, Czuczwar SJ. (2004a) Isobolographic profile of interactions between tiagabine and gabapentin: a preclinical study. Naunyn Schmiedebergs Archives of Pharmacology 369:434446.
  • Luszczki JJ, Czuczwar SJ. (2004b) Preclinical profile of combinations of some second-generation antiepileptic drugs: an isobolographic analysis. Epilepsia 45:895907.
  • Luszczki J, Borowicz KK, Czuczwar SJ. (2002) Sib 1893 possesses pro- and anticonvulsant activity in the electroshock seizure threshold test in mice. Polish Journal of Pharmacology 54:517520.
  • Luszczki J, Swiader M, Czuczwar M, Kis J, Czuczwar SJ. (2003a) Interactions of tiagabine with some antiepileptics in the maximal electroshock in mice. Pharmacology, Biochemistry and Behavior 75:319327.
  • Luszczki JJ, Borowicz KK, Swiader M, Czuczwar SJ. (2003b) Interactions between oxcarbazepine and conventional antiepileptic drugs in the maximal electroshock test in mice: an isobolographic analysis. Epilepsia 44:489499.
  • Luszczki JJ, Czuczwar M, Kis J, Krysa J, Pasztelan I, Swiader M, Czuczwar SJ. (2003c) Interactions of lamotrigine with topiramate and first-generation antiepileptic drugs in the maximal electroshock test in mice: an isobolographic analysis. Epilepsia 44:10031013.
  • Luszczki JJ, Swiader M, Parada-Turska J, Czuczwar SJ. (2003d) Tiagabine synergistically interacts with gabapentin in the electroconvulsive threshold test in mice. Neuropsychopharmacology 28:18171830.
  • Luszczki JJ, Andres MM, Czuczwar P, Cioczek-Czuczwar A, Ratnaraj N, Patsalos PN, Czuczwar SJ. (2006) Pharmacodynamic and pharmacokinetic characterization of interactions between levetiracetam and numerous antiepileptic drugs in the mouse maximal electroshock seizure model: an isobolographic analysis. Epilepsia 47:1020.
  • Masuda Y, Utsui Y, Shiraishi Y, Karasawa T, Yoshida K, Shimizu M. (1981) Evidence for a synergistic interaction between phenytoin and phenobarbital in experimental animals. Journal of Pharmacology and Expeimental Therapeutics 217:805811.
  • Mazarati AM, Baldwin R, Klitgaard H, Matagne A, Wasterlain CG. (2004) Anticonvulsant effects of levetiracetam and levetiracetam-diazepam combinations in experimental status epilepticus. Epilepsy Research 58:167174.
  • Meldrum B. (2002) Do preclinical seizure models preselect certain adverse effects of antiepileptic drugs. Epilepsy Research 50:3340.
  • Minto CF, Schnider TW, Short TG, Gregg KM, Gentilini A, Shafer SL. (2000) Response surface model for anesthetic drug interactions. Anesthesiology 92:16031616.
  • Mondadori C, Schmutz M. (1986) Synergistic effects of oxiracetam and piracetam in combination with antiepileptic drugs. Acta Neurologica Scandinavica suppl 109:113116.
  • Morris JC, Dodson WE, Hatlelid JM, Ferrendelli JA. (1987) Phenytoin and carbamazepine, alone and in combination: anticonvulsant and neurotoxic effects. Neurology 37:11111118.
  • Musolino R, Gallitto G, De Domenico P, Bonazinga MM, Sturniolo R, Labate C, Di Perri R. (1991) Synergistic anticonvulsant effect of valproic acid and ethosuximide on pentylenetetrazole-induced epileptic phenomena in rats. Journal of International Medical Research 19:5562.
  • Peterson SL. (1986) Glycine potentiates the anticonvulsant action of diazepam and phenobarbital in kindled amygdaloid seizures of rats. Neuropharmacology 25:13591363.
  • Peterson SL. (1991a) Anticonvulsant drug potentiation by glycine in maximal electroshock seizures is mimicked by D-serine and antagonized by 7-chlorokynurenic acid. European Journal of Pharmacology 199:341348.
  • Peterson SL. (1991b) Glycine potentiation of anticonvulsant drugs in pentylenetetrazol seizures in rats. Brain Research Bulletin 26:4347.
  • Peterson SL, Trzeciakowski JP, Frye GD, Adams HR. (1990) Potentiation by glycine of anticonvulsant drugs in maximal electroshock seizures in rats. Neuropharmacology 29:399409.
  • Pietrasiewicz T, Czechowska G, Dziki M, Turski WA, Kleinrok Z, Czuczwar SJ. (1993) Competitive NMDA receptor antagonists enhance the antielectroshock activity of various antiepileptics. European Journal of Pharmacolofy 250:17.
  • Rastogi SK, Ticku MK. (1985) Involvement of a GABAergic mechanism in the anticonvulsant effect of pentobarbital against maximal electroshock-induced seizures in rats. Pharmacology, Biochemistry and Behavior 22:141146.
  • Rogawski MA. (1998) Mechanism-specific pathways for new antiepileptic drug discovery. In FrenchJA, LeppikI, DichterMA (eds), Antiepileptic drug development, Vol 76. Philadelphia : Lippincott-Raven Publishers, pp. 1127.
  • Rogawski MA, Donevan SD. (1999) AMPA receptors in epilepsy and as targets for antiepileptic drugs. In Delgado-EscuetaAV, WilsonWA, OlsenRW, PorterRJ (eds), Advances in Neurology, Third ed, Vol 79. Philadelphia : Lippincott Williams & Wilkins, pp. 947963.
  • Saano V. (1986) Tofizopam selectively increases the action of anticonvulsants. Medical Biology 64:201206.
  • Schmidt D, Gram L. (1995) Monotherapy versus polytherapy in epilepsy. A reappraisal. CNS Drugs 3:194208.
  • Schmidt D, Straub HB, Meencke HJ. (1997) Putative dual (or multiple) cellular mechanism of action of anti-myoclonic epilepsy medication. Epilepsia 38:44.
  • Shank RP, Gardocki JF, Vaught JL, Davis CB, Schupsky JJ, Raffa RB, Dodgson SJ, Nortey SO, Maryanoff BE. (1994) Topiramate: preclinical evaluation of structurally novel anticonvulsant. Epilepsia 35:450460.
  • Shantilal P, David J, Joseph T. (1999) Effect of sodium valproate and flunarizine administered alone and in combination on pentylenetetrazole model of absence seizures in rat. Indian Journal of Experimental Biology 37:228233.
  • Sills GJ, Butler E, Thompson GG, Brodie MJ. (2004) Pharmacodynamic interaction studies with topiramate in the pentylenetetrazol and maximal electroshock seizure models. Seizure 13:287295.
  • Stables JP, Bertram EH, White HS, Coulter DA, Dichter MA, Jacobs MP, Loscher W, Lowenstein DH, Moshe SL, Noebels JL, Davis M. (2002) Models for epilepsy and epileptogenesis: report from the NIH workshop, Bethesda, Maryland. Epilepsia 43:14101420.
  • Stach R, Kacz D. (1977) Effect of combined dopaminergic and GABA-ergic stimulation on ouabain-induced epileptiform activity. Epilepsia 18:417423.
  • Stephen LJ, Sills GJ, Brodie MJ. (1998) Lamotrigine and topiramate may be a useful combination [letter]. Lancet 351:958959.
  • Sun M, Van Rijn CM, Liu Y, Wang M. (2002) Combination of carbamazepine and valproate in different dose proportions in maximal electroshock seizure model in mice. Epilepsy Research 51:511.
  • Swiader M, Kotowski J, Gasior M, Kleinrok Z, Czuczwar SJ. (2000) Interaction of topiramate with conventional antiepileptic drugs in mice. European Journal of Pharmacology 399:3541.
  • Swiader M, Kuzniar H, Kleinrok Z, Czuczwar SJ. (2003a) Influence of LY 300164, an AMPA/kainate receptor antagonist upon the anticonvulsant action of antiepileptic drugs against aminophylline-induced seizures in mice. Polish Journal of Pharmacology 55:103107.
  • Swiader M, Luszczki J, Wielosz M, Czuczwar SJ. (2003b) Influence of vigabatrin, a novel antiepileptic drug, on the anticonvulsant activity of conventional antiepileptics in pentetrazole-induced seizures in mice. Polish Journal of Pharmacology 55:363370.
  • Tallarida RJ. (2001) Drug synergism: its detection and applications. Journal of Pharmacology and Experimental Therapeutics 298:865872.
  • Tallarida RJ, Stone DJ Jr., Raffa RB. (1997) Efficient designs for studying synergistic drug combinations. Life Science 61:L 417–425.
  • Tanganelli P, Regesta G. (1996) Vigabatrin vs. carbamazepine monotherapy in newly diagnosed focal epilepsy: a randomized response conditional cross-over study. Epilepsy Research 25:257262.
  • Trendelenburg U. (1962) The action of acetylcholine on the nictitating membrane of the spinal cat. Journal of Pharmacology and Experimental Therapeutics 135:3944.
  • Tutka P, Luszczki J, Kleinrok Z, Arent K, Wielosz M. (2002) Molsidomine enhances the protective activity of valproate against pentylenetetrazole-induced seizures in mice. Journal of Neural Transmission 109:455466.
  • Urbanska E, Dziki M, Kleinrok Z, Czuczwar SJ, Turski WA. (1991) Influence of MK-801 on the anticonvulsant activity of antiepileptics. European Journal of Pharmacology 200:277282.
  • Urbanska E, Dziki M, Czuczwar SJ, Kleinrok Z, Turski WA. (1992) Antiparkinsonian drugs memantine and trihexyphenidyl potentiate the anticonvulsant activity of valproate against maximal electroshock-induced seizures. Neuropharmacology 31:10211026.
  • Voskuyl RA, Hoogerkamp A, Danhof M. (1992) Properties of the convulsive threshold determined by direct cortical stimulation in rats. Epilepsy Research 12:111120.
  • Walker JE, Koon R. (1988) Carbamazepine versus valproate versus combined therapy for refractory partial complex seizures with secondary generalization. Epilepsia 29:693.
  • Weaver LC, Swinyard EA, Woodbury LA, Goodman LS. (1955) Studies on anticonvulsant drug combinations: phenobarbital and diphenylhydantoin. Journal of Pharmacology and Experimental Therapeutics 113:359370.
  • White HS. (1997) Clinical significance of animal seizure models and mechanism of action studies of potential antiepileptic drugs. Epilepsia 38 (Suppl 1):S917.
  • White HS. (1999) Comparative anticonvulsant and mechanistic profile of the established and newer antiepileptic drugs. Epilepsia 40 (Suppl 5):S210.
  • Wlaz P, Baran H, Loscher W. (1994) Effect of the glycine/NMDA receptor partial agonist, D-cycloserine, on seizure threshold and some pharmacodynamic effects of MK-801 in mice. European Journal of Pharmacology 257:217225.
  • Wlaz P, Rolinski Z, Czuczwar SJ. (1996) Influence of D-cycloserine on the anticonvulsant activity of phenytoin and carbamazepine against electroconvulsions in mice. Epilepsia 37:610617.
  • Wlaz P, Ebert U, Loscher W. (1999) Anticonvulsant effects of eliprodil alone or combined with the glycineB receptor antagonist L-701,324 or the competitive NMDA antagonist CGP 40116 in the amygdala kindling model in rats. Neuropharmacology 38:243251.
  • Zarnowski T, Kleinrok Z, Turski WA, Czuczwar SJ. (1993) 2,3-dihydroxy-6-nitro-7-sulfamoylbenzo (F)quinoxaline enhances the protective activity of common antiepileptic drugs against maximal electroshock-induced seizures in mice. Neuropharmacology 32:895900.
  • Zarnowski T, Kleinrok Z, Turski WA, Czuczwar SJ. (1994a) The competitive NMDA antagonist, D-CPP-ene, potentiates the anticonvulsant activity of conventional antiepileptics against maximal electroshock-induced seizures in mice. Neuropharmacology 33:619624.
  • Zarnowski T, Kleinrok Z, Turski WA, Czuczwar SJ. (1994b) The NMDA antagonist procyclidine, but not ifenprodil, enhances the protective efficacy of common antiepileptics against maximal electroshock-induced seizures in mice. Journal of Neural Transmission. General Section 97:112.
  • Zhang ZJ, Russell S, Obeng K, Postma T, Obrocea G, Weiss SR, Post RM. (2003) Coadministration of gabapentin or MK-801 with lamotrigine slows tolerance to its anticonvulsant effects on kindled seizures. Pharmacology Biochemistry an Behavior 74:565571.