It has been known for a long time that almost 30% of patients treated for epilepsy belong to the group of patients resistant to standard pharmacological treatment. We are not able to predict, when starting the treatment, how the patient will respond to it; if it will have a satisfactory effect on the reduction of frequency of seizures; and finally in which group the patient will be – in the one responding well to treatment or in the drug-resistant one (1). The situation is complicated further by the fact that still there is not a uniform, commonly accepted definition of drug-resistant epilepsy, and taking into account what a heterogeneous disease epilepsy is, how multifactorial its neurobiological origin is, one should not expect such a definition in the near future.

Among numerous definitions proposed by various authors the most lucid and clear in terms of clinical practice seems to be the definition used in clinical studies of new drugs. In its light, epilepsy may be recognized as resistant to drugs if at least one tonic-clonic seizure or two partial (focal) seizures occur within a month and repeat in the following three months under the conditions of appropriate selection of the kind of drug for the type of seizure, their application in a dose adequate for the age and weight of the patient, and the lack of satisfactory control of seizures after applying the second-choice and third-choice drugs. A similar view is also expressed by Majkowski, by stating that we are dealing with drug-resistant epilepsy when the patient has six generalized seizures or twelve partial seizures within a year under the condition of their appropriate treatment (2). On the other hand, Genton believes the term drug-resistant epilepsy should be used when the seizures remain after two years of treatment with at least three basic antiepileptic drugs in monotherapy or their combination after assessment of the case by at least two neurologists (3). Undoubtedly we can observe several pathomechanisms conditioning resistance in a given patient with seizures resistant to treatment, which makes it even more difficult to provide a rational explanation. Although numerous studies on epilepsy resistance to pharmacological treatment have been carried out in many leading research centres worldwide, we are still far from knowing all of its mechanisms. Among those taken into account and partly understood one should list: differences in pharmacogenetic profile of individual patients and, connected with it, differences in metabolism of drugs and occurrence of some, particularly serious effects of undesirable drugs; the hypothesis of decreased sensitivity of brain receptors and ion channels to antiepileptic drugs; the phenomenon of increased expression of proteins participating in drug transport; neurodegeneration in targeted places of action of drugs; and reorganisation of synaptic contacts triggered by seizure activity in epileptic foci in the brain.

Moreover, in explaining the causes of resistance to treatment, many clinical aspects are considered, such as: early onset of disease, high frequency of seizures, in particular in the early period of treatment, organic brain damage, that is symptomatic aetiology of seizures, co-existing mental disability, family history of epilepsy, febrile convulsions in childhood, and interactions with other drugs or substances simultaneously used, e.g. caffeine or theophylline (4).

An example of studies on the role of changes in receptors in the brain generating epileptogenic activity is a study on the protective role of a lack of alpha-1b adrenergic receptors, reported this year by an Italian group of researchers. Perhaps finding a molecule blocking this type of receptor at the level of the brain would be significant in preventing the occurrence of seizures as well as their spread in the brain and modulating the seizure threshold. It was found that laboratory animals deprived of the mentioned receptor were resistant to convulsions triggered by giving them pilocarpine or kainates, and also that severity of the seizure itself and escalation of post-seizure symptoms were definitely less intense than in the group having this receptor. Furthermore, post-seizure neurodegeneration in the programmed cell death mechanism was not observed in them (5).

On the other hand, genetic polymorphism and its influence on the occurrence of resistance to antiepileptic drugs are illustrated well by the study of the relationship of polymorphism of the MCR2 gene to the response to the adrenocorticotropic hormones used in West syndrome (serious encephalopathy with infantile spasms, salaam spasms). It has been known for a long time that not all children with this syndrome respond well to the use of renal cortex hormones. A study carried out by Chinese authors and reported last year shows that polymorphism of the MC2R (melanocortin 2 receptor) gene promoter – homozygous or heterozygous carriers of the TCCT haplotype compared with non-carriers – is associated with a good response to ACTH or a lack of response (6).

Another interesting example of the influence of genetic polymorphism on the response to the applied treatment may be the already well studied polymorphism of microsomal enzymes of cytochrome P450: four main isoenzymes are distinguished, namely CYP 3A4, 2D6, 2C9 and 1A2. Metabolism of approximately 95% of all known drugs, including antiepileptic drugs, takes place with their participation. The CYP 2C9*6 variety is connected for instance with the occurrence of toxic responses to phenytoin, which was found in a group of Afro-American females from Florida (7). The CYP 2C9*2 and 3 varieties dominate among Caucasians and their presence is connected with decreased metabolism of phenytoin in comparison with other varieties of CYP 2C9; other varieties are dominant among Mongoloids (8, 9). Another known problem among patients treated with carbamazepine (the most frequently used antiepileptic drug worldwide) is its adverse interaction with erythromycin, leading sometimes to the toxic increase of carbamazepine concentration in the blood serum. This results from suppression of CYP 3A4 isoform by erythromycin, taking part in metabolism of carbamazepine (10).

Serious allergic responses to some antiepileptic drugs are also connected with polymorphism of protein coding genes involved in metabolism of drugs. A tendency to allergic responses after taking carbamazepine is shown by individuals having a group of three heat shock proteins (HSP) coded on the short arm of chromosome 6 in position 6p21.3. These proteins also take part in development of agranulocytosis after taking clozapine and abacavir allergy (11, 12). These observations confirm Garrod´s thesis, put forward at the beginning of the 20^th century, stating that the adverse effects of drugs were genetically conditioned. Later observations led to the realisation that apart from genetic predispositions such factors as age, environmental impacts, type of diet, substances and other drugs may have an influence on metabolism of a given drug, and in the process its effectiveness, interactions and adverse effects.

In some countries it is already possible to determine the pharmacogenetic profile of the whole or selected isoenzymes of cytochrome P450 in vivo and in vitro. The cost of such an analysis in the USA is approximately 300 dollars and it uses various techniques, including PCR, protein chromatography and mass spectrometry. Therefore the identification of CYP 2C9 polymorphism allows one to determine in some cases the basis of drug resistance in epilepsy. Similar to epilepsy, identification of isoforms of cytochrome P450 will be most likely used soon in determining sensitivity to drugs applied in the treatment of depression, schizophrenia, Parkinson´s disease, asthma, breast cancer and many other diseases. These studies will also permit one to predict interaction of the prescribed drug with other simultaneously used drugs and potential adverse effects connected with its use.

An interesting example of genetically conditioned drug-resistant epilepsy with a bad prognosis is the ring chromosome 20 syndrome. Epileptic seizures and frequent non-convulsive status epilepticus are accompanied here with mental retardation, body dysmorphic disorder and various kinds of conduct disorders. Epilepsy in this syndrome, and in particular a tendency to non-convulsive status epileptici, results most likely from deficiency of dopaminergic transmission, which presents a certain new aspect in the discussion about the role of neurotransmitters in generating epileptic seizures – so far GABA and glutamates have been recognized as the main ones (13).

Excessive expression of carrier proteins of drugs may also be responsible for resistance to drugs in epilepsy. A group of over 50 ATP binding proteins (ABC proteins, i.e. ATP-binding cassette) take part in removing metabolites from the nerve cell. Representatives are P-glycoprotein (Pgp) and MRP1, 2 and 5 (multidrug resistance-associated protein). They are found in the endothelium of cerebral small vessels, and they operate like the pumps removing lipophilic foreign substances (xenobiotics) from the brain through the blood-brain barrier (14). Both above-mentioned proteins – Pgp and MRP – transport many antiepileptic drugs, such as phenytoin, carbamazepine, valproic acid, phenobarbital, lamotrigine, topiramate and gabapentin. Therefore excessive activity of these proteins may hinder access of the listed drugs to the brain, and in particular the epileptogenic area. Such phenomena have been observed in laboratory animals in which convulsions were pharmacologically induced, and Pgp expression was even two times higher than in the control group (15).

On the other hand, the selective P-glycoprotein inhibitor tariquidar restored the anticonvulsive effect of phenobarbital without influencing either its pharmacokinetic parameters or its neurotoxicity. Hughes claims that in as many as 40% of patients with epilepsy resistant to drugs one should expect a beneficial response to its application (16). In some patients also a beneficial influence of verapamil bringing a definite improvement of the effectiveness of antiepileptic drugs used earlier with poor effect was observed. Probenecid in patients resistant to drugs treated with phenytoin is also supposed to act similarly (17).

It seems, although it is not unequivocally proven in clinical research, that excessive expression of Pgp is a result of polymorphism of the MDR1 gene, but this requires further studies on larger groups of patients. A further impediment results from the fact that for drugs such as vigabatrin, pregabalin, zonisamide and tiagabine specific carrier proteins have not been identified so far. Loescher et al. found that in epileptogenic areas of the brain removed during surgery for epilepsy resistant to drugs, expression of the gene for Pgp increases as much as by 130%, and of MRP protein-coding genes even by 225% (18).

Changes in sensitivity of the potential-dependent sodium channel to preparations (carbamazepine and phenytoin) blocking its functions may also partially explain the phenomenon of resistance to drugs developing with the duration of treatment. An example of such a well described syndrome with loss of sensitivity to the above-mentioned drugs is inherited autosomal dominant nocturnal frontal lobe epilepsy (ADNFLE), in which the mutation concerns the gene located on chromosome 20q for the acetylcholine receptor á4 subunit, and a large gene polymorphism causes patients to be insensitive to carbamazepine or, conversely, hypersensitive to it (19).

The theory of insufficient sensitivity of receptors and ion channels to antiepileptic drugs focuses in turn on receptor-channel mechanisms of stimulation and suppression of a nerve cell. The main role falls here to the GABAergic membrane receptors whose three types are in the principal synapses. They are marked as GABA A, B and C receptors. Types A and C are included in the chloride channel, type B constitutes presynaptic metabotropic receptors suppressing the calcium channel, whereas postsynaptic ones stimulate the potassium channel, whose result is hyperpolarization of the nerve cell. But the greatest significance in quick suppressive modulation of nerve cell stimulation can be ascribed to the GABA A receptor. Its quantitative decline in areas of the brain strategic for generating epileptic seizures, e.g. in the CA1 or CA3 zone of the hippocampus, may be a basis for the resistance to treatment of seizures. Declines of this kind are found in animals in experimental models of epilepsy (chemogenic convulsions) as well as in people with structural damage of the hippocampal complex in mesial temporal lobe epilepsy (20). Thus there are numerous antiepileptic drugs having an effect through modulation of synthesis (e.g. vigabatrin), synaptic secretion (gabapentin) or GABA reuptake (tiagabine), but also opening chloride channels (barbiturates) or causing allosteric modification of the GABA A receptor (benzodiazepines). On the other hand, Fisher et al. showed how the mutation of only one amino acid may change GABA A receptor affinity in comparison with its agonists. Replacement of alanine with asparagine in the á1 subunit evidently decreases its sensitivity to stimulating factors, whereas replacement of lysine with methionine in the γ2 subunit results in a significant decrease of inhibitory postsynaptic potentials (21).

The results of experimental studies of the epileptic state in animals show quickly occurring changes in the structure of GABA A receptors consisting of five subunits. This explains the higher effectiveness of benzodiazepines in the initial stage of the epileptic state and barbiturates in its later stages. Meanwhile Gambardella et al. showed an association of polymorphism of a fragment of metabotropic GABA B receptor coding gene with the increase of risk of development of drug-resistant temporal epilepsy (22).

An interesting mechanism of drug resistance is also suggested by the discoverers of antibodies directed against GABA decarboxylase – the main enzyme responsible for decomposition of this neurotransmitter in synapses. A similar phenomenon was also observed in relation to subunits of GluR3 receptor, which results in the disturbance of balance of synaptic stimulation and suppression processes (23).

Equally interesting are recent reports by Chinese authors who studied glutathione S-transferase activity (GST) in surgically removed fragments of brain nervous tissue. This enzyme plays a key role in hepatic metabolism of numerous antiepileptic drugs. Alpha, mu and pi isoforms of this enzyme are known. Comparison of 32 patients operated on due to drug-resistant epilepsy with 8 persons operated on for other reasons revealed that the pi isoform shows clearly higher activity in the endothelium and astrocytes of patients with drug-resistant epilepsy than in the control group, and higher in comparison with other GST isoforms (24).

As one may readily surmise on the basis of this cursory review of new directions in studies on the neurobiological basis of drug-resistant epilepsy, the attention of researchers is focused on the genetic-molecular and metabolic processes causing ineffectiveness of the applied pharmacological treatment. Undoubtedly, precisely here lies the cause of this serious problem of contemporary epileptology. But despite advanced studies we are still far from an unequivocal answer to why almost 1/3 of patients do not respond satisfactorily to applied treatment, often already at the time of its initiation; what is the cause of development of secondary resistance to drugs after an initially satisfactory period of treatment; and why in some patients the opposite situation is observed, i.e. passage from a period of drug resistance to a period of good control of seizures (25). Having an explanation of the basis of drug resistance as well as the possibility of earlier, if only approximate, determination of the response to the drugs concerned, it would be possible to expect a significant reduction of the percentage of patients for whom at the moment we cannot offer much among available pharmacotherapeutic methods.