Answers are grounded in this paper and cite their sources.
The success of the pharmacological treatment of epilepsy is not infrequently hampered by the fact that an initial early period of temporary efficacy (the “honeymoon” effect) is followed by a loss of effectiveness that may became complete within a period of 3-4 months; alternatively, or additionally, a reduction in the response to antiepileptic drugs (AEDs) may develop more gradually during prolonged treatment and lead to a condition of intractability epilepsy. One of the various factors that can account for such an unfavourable course is tolerance (i.e. a reduced response to a drug after repeated administrations), which may play a significant but still insufficiently acknowledged role. The review by Wolfang Löscher and Dieter Schmidt published in this issue of Epilepsia is therefore particularly welcome. The authors not only have great experience in the field, but their work has also significantly advanced our understanding of the “tolerance” phenomenon. In particular, this article provides readers with a comprehensive review that starts with definitions, analyzes experimental and clinical evidence, and discusses possible mechanisms and strategies aimed at preventing the development of tolerance to AEDs. What immediately catches the reader's attention is the imbalance between the wealth of experimental results and the rather limited amount of information that can be drawn from clinical studies. Since Millichap's work on acetazolamide (1), a number of studies have analysed the biological basis of tolerance to AEDs in various animal models. Most of these studies (including those by Löscher himself) have involved the kindling model, and have used experimental protocols capable of analyzing both acute tolerance (i.e. the tolerance that develops after a single drug exposure) and the chronic tolerance that develops gradually during repeated or continuous drug administrations. In principle, a decrease in the effect of a drug could be due to changes in its distribution or metabolism (pharmacokinetic tolerance) or adaptive changes in the AED target systems (pharmacodynamic tolerance). The occurrence of pharmacokinetic tolerance in patients treated with enzyme-inducing AEDs is usually easy to detect by means of plasma level monitoring, which provides a sound basis for preventing or resolving the problem. It is much more difficult to demonstrate the occurrence of pharmacodynamic tolerance in a patient who fails to respond to rational pharmacological therapy, because refractoriness to a drug may be due to the coexistence of factors related to the pathogenesis of the epilepsy. These factors can be important in a number of the epileptic disorders that are typically refractory to AEDs, such as infantile epileptic encephalopathies and temporal lobe epilepsy with hippocampal sclerosis (TLE HS). These disorders are thought to be the result of processes set in motion by “early” seizures that occur during an initial acute phase of epileptogenesis. In such cases, early seizures are thought to induce a cascade of events (2), and these seizure-dependent processes then lead to a secondary self-sustained epileptogenic mechanism that can be responsible for progression towards a condition of medical intractability. The concept of an epileptogenic process leading to increasingly severe epilepsies that eventually become refractory to AEDs is supported by the natural history of TLE HS and epileptic encephalopathies, and by experimental animal data drawn from the kindling, pilocarpine and kainic acid models. However, responsiveness to AEDs during the course of epilepsy may decrease not only because of the progression of the underlying epileptogenesis, but also because of the development of tolerance due to prolonged AED exposure. The limited information obtainable from the analysis of surgical specimens resected from drug-resistant patients undergoing epilepsy surgery does not allow us to choose decisively between these two possible interpretations. For example, the alterations in molecular structure and Na+ channels (3), or the altered effect of CBZ on Na+ currents observed in temporal lobe tissue taken from drug-resistant patients with TLE HS (4), could be induced by the epileptogenic process itself or by exposure to the drug (i.e. by the development of “tolerance”). More recently, Remy et al. (5) have used patch clamp recordings to re-investigate the drug sensitivity of voltage-dependent Na+ channels in hippocampal neurons from 13 patients with drug-resistant TLE HS, and shown that the use-dependent CBZ-induced blockade of these channels was completely lost in the CBZ-resistant patients, but completely preserved in three patients who had never been treated with CBZ or had responded well to CBZ received for only a short time before surgery. They also showed that the loss of use-dependent Na+ channel inhibition in the CBZ-resistant patients was due to the complete loss of CBZ effect on fast recovery from inactivation. These data demonstrated for the first time the usefulness of in vitro preparations of human tissue when investigating the mechanisms of drug resistance, as well as the existence of a relationship between long-term treatment with a given drug and the development of refractoriness to it. It is still too early to say if this is a cause-effect relationship (i.e., the loss of CBZ effect on Na+ channel recovery from inactivation develops through a tolerance mechanism during long-term exposure to the drug), but this reasonable hypothesis deserves further investigation. Indeed, if the contribution of tolerance to drug refractoriness is confirmed, appropriate preventive strategies can be developed. Interestingly, the tolerance induced by a specific drug also affects responsiveness to other drugs of the same category (cross-tolerance). The experimental evidence for cross-tolerance, extensively discussed by Löscher and Schmidt, supports the conclusion that the between-drug transfer of tolerance indicates common mechanisms of action (6). For example, CBZ and lamotrigine (LTG), both of which act by modulating voltage-dependent Na+ channels, exhibit bidirectional tolerance; in contrast, there is no cross-tolerance between the benzodiazepines (BZs) (which act through a different mechanism) and CBZ or LTG. However, this conclusion is not supported by examples of cross-tolerance between drugs with different mechanisms of action, such as levetiracetam (LEV) and CBZ, or the absence of cross-tolerance between drugs with a similar mechanism of action, as in the case of the two Na+ channel modulators phenytoin (PHT) and CBZ (7). Given the potential relevance of cross-tolerance to the clinical problem of multidrug resistance, it is clear that further studies are needed. In general, despite the experimental evidence, the real contribution of tolerance to the development of drug refractoriness in TLE HS and epileptic encephalopathies is still largely unknown. Specifically designed, controlled studies are warranted with the aim of disentangling the roles of epileptogenesis and long-term drug exposure. One intriguing tolerance–related phenomenon, first reported by Pinel et al. (8), is the influence of the temporal relationship between drug administration and seizures on the development of tolerance: i.e. tolerance contingent upon seizure occurrence. Contingent tolerance can therefore be defined as a decline in the therapeutic efficacy of a given drug that emerges only if the drug is present while the treated episode occurs (9). Candidate mechanisms that may contribute to this phenomenon include the adaptation of GABAA receptors, changes in the release of peptides (e.g., thyrotropin-releasing-hormone (TRH) and neuropeptide Y (NPY)), and alterations in adenosine receptors and voltage-gated ion channels (9, 10). The concept of contingency adds another dimension to interpreting the role of tolerance in the development of drug refractoriness, since we must evaluate the temporal relationship between seizure occurrence and drug exposure. In most instances, it is practically impossible to assess the influence of this variable in a clinical setting in which the therapy is administered at fixed intervals and the seizures occur randomly. Contingent tolerance, therefore, may have no practical clinical relevance - with the notable exception of status epilepticus (SE). During SE, the acutely administered drugs (usually different from those taken chronically by the patient) are delivered while the seizure activity is ongoing. It is therefore reasonable to wonder whether the acute development of contingent tolerance contributes to the loss of efficacy of the drug administered during SE, and whether there is any strategy that is capable of reducing such negative interference. Common clinical experience shows that the longer SE lasts, the more difficult it is to stop; i.e., SE becomes refractory to pharmacological treatment over time. Experimental results support the idea that this gradual loss of drug responsiveness is due to SE-dependent changes in pharmacological targets, such as receptors and voltage-gated channels, which become less sensitive to AEDs (11). The potential role of tolerance to AEDs acutely administered in a condition corresponding to the definition of contingent tolerance (i.e., due to the concomitant occurrence of seizures) is difficult to demonstrate. For example, it can be argued that, when SE persists for more than 60 minutes, it is because seizure activity failed to respond to early pharmacological treatments and was thus refractory from the start - and that the development of tolerance has nothing to do with its pathogenesis. However, it is not unusual to see patients with ongoing SE that has been repeatedly treated elsewhere with insufficient doses of first-line drugs (namely BZs), and where subsequent administration of an “adequate” i.v. bolus dose does not stop the SE; in such cases, the contribution of tolerance can be reasonably suspected. To the best of my knowledge, this hypothesis has never been systematically tested, but there is circumstantial evidence of the presence of acute drug tolerance during SE. For example, Osorio and Reed (12) observed some degree of pharmacodynamic tolerance (manifested as a loss of anesthetic effect and ability to induce an EEG pattern of burst suppression) in four out of 12 cases of refractory SE treated with pentobarbital infusion; after 4-5 days of infusion, these patients regained consciousness at pentobarbital concentrations that had earlier caused unresponsiveness and the loss of all brainstem reflexes. Although tolerance to anesthetic effects does not necessarily parallel tolerance to antiepileptic activity, this observation does demonstrate the occurrence of some tolerance to pentobarbital during SE. Controlled studies of the role (if any) of contingent tolerance in the development of SE refractoriness could help in developing sorely needed evidence-based guidelines for SE treatment (13, 14). Another interesting and clinically-relevant point, appropriately discussed by Löscher and Schmidt, is the fact that pharmacodynamic tolerance to side-effects usually develops more rapidly than tolerance to efficacy; moreover, tolerance to adverse effects seems to develop more easily with low than with high drug doses, whereas the opposite is true in the case of tolerance to therapeutic effects. These observations have major implications for clinical practice as they suggest that antiepileptic treatment should be started using slow titration schedules rather than loading doses (unless the clinical situation requires rapid seizure control as is often the case when dealing with infant and childhood epilepsies), because slow titration reduces the risk of loss of pharmacological effect and simultaneously minimizes adverse effects. Other questions of particular interest to clinicians are: how can pharmacodynamic or functional tolerance be detected in a clinical setting? Are there any differences in the propensity of different AEDs to induce tolerance? And are there any effective strategies other than slow titration to prevent the development of tolerance to efficacy or counteract it once it has developed? The available information on which to base the answers to these questions can be found throughout the paper by Löscher and Schmidt, and are summarised and discussed by these authors. Functional tolerance can be operationally defined as a shift from responder status: i.e., regressing from 50% or greater reduction in seizures (compared to initial seizure baseline) to a less than 50% reduction. Functional tolerance can be measured, in groups of patients, as the number of responders between two visits. While this operational definition makes it possible to design controlled studies, it may overlook significant changes occurring in individual patients. In general, clinical studies aimed at assessing tolerance should be designed in such a way as to rule out confounding factors, including lack of compliance, inclusion biases, disease progression, and changes in plasma concentrations (pharmacokinetic tolerance). According to the observations reviewed by Löscher and Schmidt, different degrees of evidence for tolerance have been found for both traditional and newer AEDs. The most substantial data concerning the propensity to tolerance points toward the BZs. Potentially confounding pharmacokinetic factors should be taken into account for all drugs, since without this information it is impossible to differentiate between the roles of pharmacokinetic and pharmacodynamic mechanisms in inducing a decline in drug responsiveness. Finally, the current strategies regarding prevention (or counteraction) of tolerance are rather disappointing. Proposed strategies for restoring drug efficacy once tolerance has developed (i.e., dose increases, intermittent treatment, drug holidays, and the use of receptor antagonists) are impracticable or of unproved efficacy. Overall, the Löscher and Schmidt review should be highly prized for providing us with a comprehensive and thoughtful account of a subject that has not received adequate attention from epileptologists. I am sure that the readers of Epilepsia will enjoy it as much as I did.
No takes yet. Share an insight, caveat, or question.
G. Avanzini (2006) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: