SEARCH

SEARCH BY CITATION

Keywords:

  • Antiepileptic drug;
  • Potassium channel;
  • KCNQ;
  • Kv7;
  • Seizure;
  • Epilepsy;
  • Preclinical models;
  • Refractory epilepsy

Summary

  1. Top of page
  2. Summary
  3. Human Serum and Brain Concentrations
  4. In Vitro Studies
  5. Models of Epileptiform Activity
  6. Human Epileptic Brain Tissue
  7. In Vivo Models of Partial Epilepsy (Partial and Secondarily Generalized Seizures)
  8. Temporal Lobe Epilepsy and Epileptogenesis
  9. Generalized Tonic–Clonic Seizures
  10. Generalized Epilepsy (Absence Seizures)
  11. Genetic Models
  12. Status Epilepticus
  13. Neuroprotection
  14. CNS Tolerability Profile of AEDs
  15. Comparison with Other AEDs
  16. Combination Treatments
  17. Conclusions
  18. Acknowledgments
  19. Disclosure
  20. References
  21. Supporting Information

Retigabine [RTG (international nonproprietary name); ezogabine (EZG; U.S. adopted name)] is a first-in-class antiepileptic drug (AED) that reduces neuronal excitability by enhancing the activity of KCNQ (Kv7) potassium (K+) channels. RTG/EZG has recently been approved by the European Medicines Agency and the U.S. Food and Drug Administration as adjunctive therapy in adults with partial-onset seizures. In this review we discuss the activity that RTG/EZG has demonstrated across a broad spectrum of in vitro/in vivo animal models of seizures, including generalized tonic–clonic, primary generalized (absence), and partial seizures, in addition to the compound’s ability to resist and block the occurrence of seizures induced by a range of stimuli across different regions of the brain. The potency of RTG/EZG in models refractory to several conventional AEDs and the work done to assess antiepileptogenesis and neuroprotection are discussed. Studies that have evaluated the central nervous system side effects of RTG/EZG in animals are reviewed in order to compare these effects with adverse events observed in patients with epilepsy. Based on its demonstrated effect in a number of animal epilepsy models, the synergistic and additive activity of RTG/EZG with other AEDs supports its potential use in therapeutic combinations for different seizure types. The distinct mechanism of action of RTG/EZG from those of currently available AEDs, along with its broad preclinical activity, underscores the key role of KCNQ (Kv7) K+ channels in neuronal excitability, and further supports the potential efficacy of this unique molecule in the treatment of epilepsy.

Retigabine [RTG (international nonproprietary name); ezogabine (EZG; U.S. adopted name)] is a first-in-class antiepileptic drug (AED) that is well known to neuropharmacologists interested in the physiologic and pathophysiologic role of KCNQ (Kv7) channels in the central nervous system (CNS). Its unique mechanism of action (MoA), and the historical breadth and depth of data from in vitro and in vivo animal seizure and epilepsy models predict a potentially effective treatment for a range of seizure disorders (Czuczwar et al., 2010; Owen 2010). This potential is supported by three double-blind, placebo-controlled trials of adjunct therapy in patients with partial seizures. In the first of these studies, a dose-ranging phase II trial, RTG/EZG at 600, 900, and 1,200 mg/day demonstrated incremental improvements in seizure control with increasing dose versus placebo (p < 0.001) in patients with inadequately controlled partial-onset seizures despite stable therapy with one or two AEDs (study 205, n = 399; Porter et al., 2007). The efficacy of RTG/EZG in partial-onset seizures was further demonstrated in two randomized, double-blind, placebo-controlled phase III trials (studies 301 and 302, n = 306 and 538, respectively; ClinicalTrials.gov identifiers NCT00232596 and NCT00235755) (Brodie et al., 2010; French et al., 2011). Efficacy, as measured by responder rate (50% reduction in seizures) and median reduction in seizure frequency, was significantly improved with all doses of RTG/EZG: 600 and 900 mg/day (study 302); and 1,200 mg/day (study 301), compared with placebo. Notably, RTG/EZG did not affect the clearance of other AEDs, and the concomitant AEDs allowed in studies 301 and 302 did not interact with the clearance of RTG/EZG, supporting the potential utility of RTG/EZG in combination with other AEDs. The efficacy of RTG/EZG in treating other seizure types or in combination with other specific AEDs has not been assessed.

In this review, we consider the results of studies that have examined the activity of RTG/EZG across a broad spectrum of in vitro and in vivo animal models of seizures.

Human Serum and Brain Concentrations

  1. Top of page
  2. Summary
  3. Human Serum and Brain Concentrations
  4. In Vitro Studies
  5. Models of Epileptiform Activity
  6. Human Epileptic Brain Tissue
  7. In Vivo Models of Partial Epilepsy (Partial and Secondarily Generalized Seizures)
  8. Temporal Lobe Epilepsy and Epileptogenesis
  9. Generalized Tonic–Clonic Seizures
  10. Generalized Epilepsy (Absence Seizures)
  11. Genetic Models
  12. Status Epilepticus
  13. Neuroprotection
  14. CNS Tolerability Profile of AEDs
  15. Comparison with Other AEDs
  16. Combination Treatments
  17. Conclusions
  18. Acknowledgments
  19. Disclosure
  20. References
  21. Supporting Information

The clinical relevance of efficacy or side effects observed in animal models in vitro or in vivo depends on whether concentration of the drug in the model lies within a range that might be achieved in humans. In clinical trials conducted with RTG/EZG, doses of 200, 300, or 400 mg were administered three times daily. The combined population-modeled average steady state serum concentration (Cave) data based upon the area under curve (AUC0–τ) dosing interval from studies 205, 301, and 302 showed that mean Cave levels across the treated population were 601, 914, and 1,209 ng/ml at 600, 900, and 1,200 mg/day doses, respectively [GlaxoSmithKline (GSK)/Valeant, personal communication]. Given that the plasma protein binding of RTG/EZG is 79–80%, mean free plasma concentrations of RTG/EZG, when adjusted for molar weight, were ∼0.4, 0.6, and 0.8 μm, respectively. Because this is the average concentration over the dosing interval, we can assume that exposures remained greater than these levels for a period of time. Furthermore, given that RTG/EZG can cross the blood–brain barrier and is not subject to P-glycoprotein active transport across this barrier (GSK/Valeant, personal communication, PR2008-017), free brain concentrations will closely follow free plasma concentrations. Consequently, maximum free brain concentrations of RTG/EZG achieved in the epilepsy trials at the 1,200 mg dose can be estimated at a minimum of ∼1 μm. Allowing for some margin of error, brain concentrations ≤2 μm may be considered to have clinical relevance.

In Vitro Studies

  1. Top of page
  2. Summary
  3. Human Serum and Brain Concentrations
  4. In Vitro Studies
  5. Models of Epileptiform Activity
  6. Human Epileptic Brain Tissue
  7. In Vivo Models of Partial Epilepsy (Partial and Secondarily Generalized Seizures)
  8. Temporal Lobe Epilepsy and Epileptogenesis
  9. Generalized Tonic–Clonic Seizures
  10. Generalized Epilepsy (Absence Seizures)
  11. Genetic Models
  12. Status Epilepticus
  13. Neuroprotection
  14. CNS Tolerability Profile of AEDs
  15. Comparison with Other AEDs
  16. Combination Treatments
  17. Conclusions
  18. Acknowledgments
  19. Disclosure
  20. References
  21. Supporting Information

Several studies have demonstrated a graded inhibitory effect of RTG/EZG on cortical neuron excitability, which is considered to be primarily mediated by KCNQ (Kv7) channel facilitation (see companion submission Gunthorpe et al., 2012). Given the observed concentrations of drug that are associated with doses in the efficacious range (see above), interaction of RTG/EZG with Kv7.2/7.3 channels is most likely to contribute to the clinical efficacy of the drug. In contrast, there is evidence that RTG/EZG at higher concentrations (10–50 μm), can enhance γ-aminobutyric acid (GABA)A–mediated inhibitory synaptic transmission in vitro, which, in theory, could also contribute to anticonvulsant efficacy (Otto et al., 2002). However, these concentrations are significantly higher than the plasma concentrations achieved in the pivotal clinical trials or the estimated brain concentrations in those studies. Therefore, it seems unlikely that significant facilitation of GABAA transmission contributes to the observed clinical efficacy or safety of RTG/EZG.

Models of Epileptiform Activity

  1. Top of page
  2. Summary
  3. Human Serum and Brain Concentrations
  4. In Vitro Studies
  5. Models of Epileptiform Activity
  6. Human Epileptic Brain Tissue
  7. In Vivo Models of Partial Epilepsy (Partial and Secondarily Generalized Seizures)
  8. Temporal Lobe Epilepsy and Epileptogenesis
  9. Generalized Tonic–Clonic Seizures
  10. Generalized Epilepsy (Absence Seizures)
  11. Genetic Models
  12. Status Epilepticus
  13. Neuroprotection
  14. CNS Tolerability Profile of AEDs
  15. Comparison with Other AEDs
  16. Combination Treatments
  17. Conclusions
  18. Acknowledgments
  19. Disclosure
  20. References
  21. Supporting Information

Several groups have studied the effects of RTG/EZG on epileptiform activity (EA) induced in rat brain slices in order to investigate its anticonvulsant mechanism of action. Armand et al. (2000) found that 10–50 μm of RTG/EZG could completely suppress, in a concentration-dependent manner, EA induced by a reduction in magnesium (Mg2+) concentration in entorhinal cortex (EC)–hippocampus slices. In comparison, only high concentrations of carbamazepine (CBZ, 75 μm), phenytoin (PHT, 50 μm), valproate acid (VPA, 1.5 mm), and phenobarbital (PB, 150 mm) were able to completely suppress seizure-like events (SLEs), and these drugs had no effect on the frequency of ensuing repetitive discharges. In a similar study, in which EA was driven by transmission through N-methyl-d-aspartate (NMDA) receptors following their release from Mg2+ block, concentrations as low as 1 μm of RTG/EZG reduced the frequency of discharges in hippocampal slices incubated in Mg2+-free artificial cerebrospinal fluid (ACSF) by approximately 20% after 30 min (Dost & Rundfeldt, 2000).

EA can also be induced in rat EC slices with the potassium (K+) channel blocker, 4-aminopyridine (4-AP), which inhibits delayed rectifier K+ channels, but has no effect on KCNQ (Kv7) channels at concentrations ≤10 μm (Main et al., 2000; Robbins, 2001). Although RTG/EZG at concentrations as low as 2 μm inhibited 4-AP–induced SLEs, higher concentrations were required to inhibit the brief negative field potential shifts (“interictal events” or interictal epileptiform discharges) that occurred between SLEs, and which are thought to have GABAergic origins (Armand et al., 1999). Yonekawa et al. (1995) compared RTG/EZG 20 μm with PHT, CBZ, VPA, and losigamone, and found RTG/EZG to be the most effective at preventing 4-AP–induced excitation without depressing physiologically evoked activity.

Human Epileptic Brain Tissue

  1. Top of page
  2. Summary
  3. Human Serum and Brain Concentrations
  4. In Vitro Studies
  5. Models of Epileptiform Activity
  6. Human Epileptic Brain Tissue
  7. In Vivo Models of Partial Epilepsy (Partial and Secondarily Generalized Seizures)
  8. Temporal Lobe Epilepsy and Epileptogenesis
  9. Generalized Tonic–Clonic Seizures
  10. Generalized Epilepsy (Absence Seizures)
  11. Genetic Models
  12. Status Epilepticus
  13. Neuroprotection
  14. CNS Tolerability Profile of AEDs
  15. Comparison with Other AEDs
  16. Combination Treatments
  17. Conclusions
  18. Acknowledgments
  19. Disclosure
  20. References
  21. Supporting Information

The ability of RTG/EZG to inhibit spontaneous sharp waves in slices of cortex removed from patients with intractable, drug-resistant seizures was investigated by Straub et al. (2001). A concentration of 10 μm RTG/EZG reduced spontaneous sharp waves by 50–80%, whereas complete inhibition was achieved with 50 μm RTG/EZG. A majority of the resected tissues did not display spontaneous EA, but this could be induced through incubation in Mg2+-free ACSF, with or without the GABAA antagonist bicuculline. Under these conditions, similar concentration-dependent inhibition of EA was achieved with RTG/EZG. In contrast, CBZ and PHT (25 μm in each case, which is close to their therapeutic free concentrations) achieved only minor reductions (30–40%) in spontaneous EA (Kohling et al., 1998). These results indicate that RTG/EZG is able to inhibit EA in pathologic brain tissue from the presumed epileptogenic zone resected in these patients with pharmacoresistant partial epilepsy. This also suggests that RTG/EZG is effective against a pathologic state that does not respond to sodium (Na+) channel blockers (Remy et al., 2003). Therefore, the expression and function of neuronal KCNQ (Kv7) channels is not significantly compromised in pathologic tissue from patients with severe, intractable epilepsy. As far as we are aware, only one study has examined the expression of KCNQ (Kv7) channels in the human epileptic brain directly. Yus-Najera et al. (2003) observed a slight reduction in KCNQ5 (Kv7.5) immunoreactivity in the hippocampus from patients with temporal lobe epilepsy (TLE), although they attributed this to the dispersion of granule cell neurons in the sclerotic tissue, rather than a reduction in expression by individual neurons. Unfortunately, they were unable to study the expression of either KCNQ2 (Kv7.2) or KCNQ3 (Kv7.3).

In Vivo Models of Partial Epilepsy (Partial and Secondarily Generalized Seizures)

  1. Top of page
  2. Summary
  3. Human Serum and Brain Concentrations
  4. In Vitro Studies
  5. Models of Epileptiform Activity
  6. Human Epileptic Brain Tissue
  7. In Vivo Models of Partial Epilepsy (Partial and Secondarily Generalized Seizures)
  8. Temporal Lobe Epilepsy and Epileptogenesis
  9. Generalized Tonic–Clonic Seizures
  10. Generalized Epilepsy (Absence Seizures)
  11. Genetic Models
  12. Status Epilepticus
  13. Neuroprotection
  14. CNS Tolerability Profile of AEDs
  15. Comparison with Other AEDs
  16. Combination Treatments
  17. Conclusions
  18. Acknowledgments
  19. Disclosure
  20. References
  21. Supporting Information

The most common models of complex partial and secondary generalized seizures are the kindling models. These involve electrical stimulation of the limbic system to produce a progressive sensitization to the stimulus, leading to the evolution of focal seizures and, subsequently, secondary generalized seizures. These models reproduce many of the features of human partial epilepsies (Bertram, 2007; McIntyre & Gilby, 2009), including treatment resistance (Löscher, 2002), but differ in three key respects: There is no clear evidence that a process similar to kindling occurs in humans. Repeated kindling can lead to the occurrence of spontaneous seizures (Sutula, 2007); however, only after repeated long-term stimulation. Lastly, kindling is imposed on a normal brain over a relatively short period of time (days to weeks), and thus may not reflect some of the chronic pathological changes that accompany human epilepsy. Despite these differences, kindling models provide an important opportunity to assess the potential antiepileptogenic and anticonvulsant efficacy of novel drugs in partial-onset seizures (Sankar et al., 2010).

RTG/EZG has been shown to inhibit amygdala-kindled focal and secondarily generalized seizures in adult rats (Tober et al., 1996). An increase in the threshold for afterdischarge induction within the amygdala was achieved with doses as low as 0.01 mg/kg orally (p.o.), although much higher doses (10 mg/kg p.o.) were required to reduce the behavioral seizure. A similar dose relationship was observed with VPA, with 30 mg/kg, intraperitoneal (i.p.), able to increase the threshold for afterdischarge induction, but 200 mg/kg, i.p., required to reduce behavioral seizure severity and duration (Tober et al., 1996). Notably, afterdischarge duration was reduced significantly but only at the higher doses of RTG/EZG and VPA required to reduce the behavioral seizure. This may be a common feature of the model, since a similar observation was made with the Na+ channel blocker, lamotrigine (LTG)(Stratton et al., 2003). These results confirm the efficacy of RTG/EZG in the amygdala kindling model, although the dose required to reduce seizure severity (10 mg/kg, p.o.) may have greater clinical relevance than the dose associated with increased threshold for afterdischarge induction by electrical stimulus (0.01 mg/kg, p.o.).

Rapid kindling of the hippocampus is a model that has been used to profile a number of new anticonvulsant drugs. The model not only differs in the site stimulated, but also in the time frame with which a fully kindled state is achieved—typically within a day (Lothman et al., 1985). An advantage of this model is that it can be applied to immature rats in order to evaluate AED efficacy in the developing brain. Mazarati et al. (2008) found that RTG/EZG was able to raise afterdischarge threshold, reduce duration, and reduce behavioral severity and duration in rats at postnatal days 14 (2.5 mg/kg, i.p.), 21 (5 mg/kg, i.p.), and 35 (5 mg/kg, i.p.). These ages correspond roughly to the human developmental equivalents of neonatal, childhood, and adolescence, respectively. Although there was a suggestion of increased effectiveness of RTG/EZG in the younger rats, it is not possible to draw firm conclusions in the absence of pharmacokinetic data and further investigation with additional doses. However, the study does serve to highlight the anticonvulsant efficacy of RTG/EZG in the developing brain, which may predict efficacy in pediatric epilepsy. The occurrence of side effects in these models will be considered in more detail below, but the doses of RTG/EZG chosen for the hippocampal kindling model were associated with mild-to-moderate motor impairments that were most marked in the younger rats (Mazarati et al., 2008). A range of other AEDs were examined in a model of rapid kindling (Sankar et al., 2010) in neonatal animals (P14-P35), including levetiracetam (LEV), LTG, and topiramate (TPM)(Mazarati et al., 2008). Here, RTG/EZG reduced baseline hippocampal excitability across all ages, and was quantitatively comparable or superior to the other AEDs in terms of reducing kindling acquisition and effects on kindled seizures at all ages. For example, TPM inhibited rapid kindling epileptogenesis in 2-, 3-, and 5-week-old animals, but the drug was most effective in 5-week-old animals, and was least effective in P14 rats. Unlike TPM, RTG/EZG was equally effective across all three ages, and delayed the occurrence of focal seizures selectively in 2-week-old rats (Mazarati et al., 2007, 2008).

RTG/EZG has also been evaluated in two models of pharmacoresistant epilepsy: the LTG-resistant amygdala kindled rat and the 6-Hz psychomotor seizure test. The former is a variant of the amygdala kindling model, wherein kindling is conducted in the presence of low-dose LTG (5 mg/kg, i.p.). Once fully kindled, rats are tested to confirm they do not respond to treatment with a higher dose of LTG (Postma et al., 2000). RTG/EZG (10 or 20 mg/kg, i.p.) reduced behavioral seizures and afterdischarge duration in fully kindled, LTG-resistant animals (Srivastava & White, 2005). VPA, clonazepam, and LEV were also found to be effective, whereas CBZ, PHT, and TPM were not (White HS, personal communication).

In the 6-Hz psychomotor seizure model, RTG/EZG blocked seizures induced by either 32 or 44 mA current stimulation in a dose-dependent manner (Table 1). The median effective dose to block seizures in 50% (ED50) of the mice tested was 26 mg/kg at 32 mA and 33 mg/kg at 44 mA. Notably, whereas the ED50 for RTG/EZG reduced by less than one-half as current intensity increased from 32 to 44 mA, the ED50 for LEV and VPA increased. Thus potency of these AEDs was reduced by 50- and 3-fold, respectively. At the 44 mA current intensity, RTG/EZG was clearly more potent than either of the older AEDs, and was more efficacious than PHT, LTG, and ethosuximide (ESM) (Bialer et al., 2009).

Table 1.   Mouse 6-Hz psychomotor seizure model: efficacy of RTG/EZG and other AEDs in partial and secondarily generalized seizures
Drugs (i.p. administration)Corneal stimulation intensities (ED50 values, mg/kg)
32 mA44 mA
  1. AEDs, antiepileptic drugs; ED50, median effective dose to block seizures in 50% of the study population; i.p., intraperitoneal; RTG/EZG, retigabine (ezogabine).

  2. bHistorical data published in Barton et al. (2001).

Retigabine (ezogabine)a2633
Lamotrigineb>60>60
Phenytoinb>60>60
Ethosuximideb167>600
Levetiracetamb191089
Valproateb126310

Temporal Lobe Epilepsy and Epileptogenesis

  1. Top of page
  2. Summary
  3. Human Serum and Brain Concentrations
  4. In Vitro Studies
  5. Models of Epileptiform Activity
  6. Human Epileptic Brain Tissue
  7. In Vivo Models of Partial Epilepsy (Partial and Secondarily Generalized Seizures)
  8. Temporal Lobe Epilepsy and Epileptogenesis
  9. Generalized Tonic–Clonic Seizures
  10. Generalized Epilepsy (Absence Seizures)
  11. Genetic Models
  12. Status Epilepticus
  13. Neuroprotection
  14. CNS Tolerability Profile of AEDs
  15. Comparison with Other AEDs
  16. Combination Treatments
  17. Conclusions
  18. Acknowledgments
  19. Disclosure
  20. References
  21. Supporting Information

Related to kindling models are other models that target the limbic brain in rodents to recreate a chronic pathology similar to that observed in patients with TLE. Electrical stimulation of the perforant path (Sloviter, 1991), and local or peripheral administration of the chemoconvulsant kainic acid (Nadler, 1981), or similar methods, are used to induce status epilepticus (SE), a severe and prolonged convulsive seizure that produces a lasting epileptic state. A consequence of this severe seizure is widespread neuronal loss, particularly affecting the hippocampus, and reorganization of granule cell axons within the dentate gyrus (“mossy fiber sprouting”). This pathology closely resembles that observed in resected or postmortem tissue from patients with TLE (Buckmaster & Dudek, 1997; Pitkänen et al., 2007). Mossy fiber sprouting may provide a structural basis for enhanced excitation and epileptogenesis in the hippocampus, which may lead to recurrent seizures and may be the basis for forming intractable TLE. A similar pathology has also been observed after repeated kindled seizures in rats, suggesting that even short-duration seizures (lasting only minutes) can, if repeated, lead to neurodegeneration (Kotloski et al., 2002). It is notable that 1–2 months after SE, spontaneous partial and secondary generalized seizures emerge in many of the animals. These seizures are sensitive to drugs used to treat human focal-onset epilepsy (Grabenstatter et al., 2005, 2007; Nissinen & Pitkänen, 2007; Grabenstatter & Dudek, 2008), and also show evidence of the drug tolerance observed in the humans (van Vliet et al., 2008, 2010). Therefore, these models recapitulate some of the essential characteristics of human TLE.

Protection against the neurodegenerative consequences of SE in animal models has not been evaluated with RTG/EZG and has been demonstrated with only a few AEDs (Pitkänen & Lukasiuk, 2009). However, the ability of RTG/EZG to prevent the expression of spontaneous seizures following SE has been studied indirectly. Ex vivo studies of EC brain slices taken 1–2 weeks after kainic acid administration were conducted by Smith et al. (2007). Slices were incubated in high-K+ saline with the GABAA-receptor blocker, picrotoxin, to increase the occurrence of spontaneous EA. Parallel slices were taken to confirm neuronal loss and reorganization characteristic of hippocampal sclerosis. Under these conditions, RTG/EZG decreased the burst rate with a half-inhibitory concentration (IC50) of 3.1 μm. Values for CBZ and PHT were >50 μm in each case. Although the use of artificial means to increase burst rate perhaps reduces the relevance of these results for human TLE, the apparent potency of RTG/EZG in the model suggests that the KCNQ (Kv7) mechanism remains viable and effective despite significant pathology within the hippocampus and EC. These findings are consistent with the efficacy of RTG/EZG on human brain tissue resected during surgery from patients with intractable epilepsy (see above).

Generalized Tonic–Clonic Seizures

  1. Top of page
  2. Summary
  3. Human Serum and Brain Concentrations
  4. In Vitro Studies
  5. Models of Epileptiform Activity
  6. Human Epileptic Brain Tissue
  7. In Vivo Models of Partial Epilepsy (Partial and Secondarily Generalized Seizures)
  8. Temporal Lobe Epilepsy and Epileptogenesis
  9. Generalized Tonic–Clonic Seizures
  10. Generalized Epilepsy (Absence Seizures)
  11. Genetic Models
  12. Status Epilepticus
  13. Neuroprotection
  14. CNS Tolerability Profile of AEDs
  15. Comparison with Other AEDs
  16. Combination Treatments
  17. Conclusions
  18. Acknowledgments
  19. Disclosure
  20. References
  21. Supporting Information

Generalized tonic–clonic seizures (GTCS), characterized by muscle rigidity interspersed with violent muscle contractions and loss of consciousness, arise when waves of excitation and hypersynchronization involve large parts of the cortex. These seizures may be triggered by focal events such as partial seizures, or occur in response to widespread increased neuronal excitability induced by drugs, other neurologic conditions, or extreme body temperatures (e.g., febrile seizures) during an infectious illness. GTCS that causes widespread increased neuronal excitation can be induced in animal models. The most common methods involve electroshock or administration of a chemoconvulsant, such as the GABAA blocker, pentylenetetrazole (PTZ). Given the relative simplicity of these models, they have been used extensively to study the anticonvulsant potency of novel AEDs (White, 2003). Variants of these models, when combined with an appropriate pharmacokinetic analysis, allow a quantitative assessment of drug efficacy that can be used to predict clinical dose (Large et al., 2009).

Rostock et al. (1996) studied the effects of RTG/EZG in a range of generalized-seizure models. In mice and rats, supramaximal electroshock-induced seizures were inhibited with RTG/EZG at ED50 values of 9.3 and 5.1 mg/kg, i.p. (15 and 10 min pretreatment time), respectively. In mice, seizures induced by subcutaneous (s.c.), PTZ were inhibited with an ED50 of 13.5 mg/kg i.p. (15 min pretreatment time), and those induced by intracerebroventricular NMDA were inhibited with an ED50 of 9.1 mg/kg, i.p (15 min pretreatment time). RTG/EZG also increased the threshold for seizures induced by intravenous (i.v.) infusion of PTZ. This manner of administration first evokes a brief myoclonic, followed by clonic, twitch. Significant elevation of the myoclonic threshold was observed with doses of 7.5 mg/kg, i.p.; a slightly lower dose (5 mg/kg, i.p.) increased the clonic threshold (Rostock et al., 1996). These results confirm that RTG/EZG is a potent anticonvulsant against generalized seizures induced by disturbance of both excitatory and inhibitory transmission. However, RTG/EZG was unable to prevent clonic seizures induced by bicuculline, or motor seizures induced by blocking spinal glycinergic transmission using strychnine (ED50 > 30 mg/kg, i.p., in each case) (Rostock et al., 1996). These authors also examined potential tolerance to the drug’s anticonvulsant efficacy, and showed that at 14 days posttreatment, there was no diminished effect in the electroshock model (Rostock et al., 1996).

In a separate study conducted by the National Institute of Neurological Disorders and Stroke (NINDS), RTG/EZG was effective against maximal electroshock-induced seizures in mice and rats, with ED50 values of 9.3 and 2.9 mg/kg, p.o., respectively (Kupferberg HJ: Valeant, data on file). RTG/EZG also afforded protection against GTCS induced by s.c. PTZ (ED50 values of 13.5 mg/kg, i.p., in mice and 68 mg/kg, p.o., in rats). However, in contrast to the findings of Rostock et al. (1996), the same study found that the drug did not increase the threshold for a myoclonic twitch or subsequent clonus induced by intravenous PTZ.

Finally, confirmation that the efficacy of RTG/EZG in these models is likely to be mediated by activation of KCNQ (Kv7) channels comes from studies where these channels have either been pharmacologically blocked or are genetically less functionally active. In a previously unpublished study conducted by Larsen and Cheney (GSK/Valeant, personal communication, PR2008-019), anticonvulsant efficacy of RTG/EZG in the electroshock seizure model in mice was reduced in a dose-dependent manner by the selective Kv7 blocker, XE-991 (Fig. 1). Similarly, XE-991 also reversed the antiepileptogenic effects of RTG/EZG in the rapid kindling model in rats (Sankar et al., 2009). The dependence of RTG/EZG on the activation of KCNQ (Kv7) channels was further demonstrated in the Szt1 mouse model, which features a C-terminal deletion in the mouse Kcnq2 gene, as well as deletions of the Chrn4 (nACh-receptor α4 subunit) and Arfgap1 (guanosine triphosphatase-activating protein that inactivates adenosine diphosphate-ribosylation factor 1) genes. In Szt1 mice, the functionality of the KCNQ2 channel is impaired, and the anticonvulsant action of RTG/EZG is reduced (Otto et al., 2004).

image

Figure 1.   The KCNQ inhibitor XE-991 blocks the efficacy of retigabine (RTG)/ezogabine (EZG) in the maximal electroshock model of epilepsy in mice. RTG/EZG was coadministered with XE-991, 15 min prior to testing. Values indicate the percentage of animals experiencing a seizure in response to a supramaximal stimulation. Veh, vehicle-treated control.

Download figure to PowerPoint

Generalized Epilepsy (Absence Seizures)

  1. Top of page
  2. Summary
  3. Human Serum and Brain Concentrations
  4. In Vitro Studies
  5. Models of Epileptiform Activity
  6. Human Epileptic Brain Tissue
  7. In Vivo Models of Partial Epilepsy (Partial and Secondarily Generalized Seizures)
  8. Temporal Lobe Epilepsy and Epileptogenesis
  9. Generalized Tonic–Clonic Seizures
  10. Generalized Epilepsy (Absence Seizures)
  11. Genetic Models
  12. Status Epilepticus
  13. Neuroprotection
  14. CNS Tolerability Profile of AEDs
  15. Comparison with Other AEDs
  16. Combination Treatments
  17. Conclusions
  18. Acknowledgments
  19. Disclosure
  20. References
  21. Supporting Information

PTZ administered to Wistar rats has been suggested to model electroencephalographic (EEG) features of absence seizures, specifically the appearance of spike-wave discharges (SWDs; field-potential spikes at 6–11 Hz lasting ≥1 s) (Marescaux et al., 1984). RTG/EZG was tested in this model to determine whether it would reduce EEG events, but doses ≤10 mg/kg, p.o., had no effect on either incidence or duration (Rostock A, personal communication, FP4295). Higher doses begin to affect the behavior/locomotor activity in rats. In contrast, ethosuximide (ESM; 25–300 mg/kg, p.o.), diazepam (DZP; 0.3–3 mg/kg, p.o.), and VPA (200 and 400 mg/kg, p.o.) dose-dependently inhibited the SWDs in these animals (Micheletti et al., 1985).

Selective breeding of Wistar rats has produced a line that exhibits spontaneous absence seizures, characterized by SWDs and brief episodes of behavioral arrest (Danober et al., 1998). This line, known as the Genetic Absence Epilepsy Rats of Strasbourg (GAERS) has been extensively studied with a range of AEDs. Drugs active against human absence epilepsy were active in the model, and contraindicated drugs such as CBZ were found to be inactive (Marescaux et al., 1992). In a previously unpublished study using the GAERS model, RTG/EZG had no effect on seizure expression (both cumulative duration and number of SWDs) at doses ≤8 mg/kg, i.p. or p.o. (Nehlig A, personal communication, PR2007-032; for more detailed information, see Supporting information).

Therefore, at doses that have demonstrated to be effective in GTCS models, RTG/EZG appeared to be ineffective in two generalized absence seizure models: the GAERS, and seizures acutely induced by low-dose PTZ. However, bursts are often triggered by the initial hyperpolarization of nucleus reticularis thalami neurons followed by the slow depolarization of low-threshold Ca2+ channels (Crunelli & Leresche, 2002). Therefore, augmenting Kv7 currents in these models may not be effective. In the GAERS model, dissipation of SWDs over the daytime recording period naturally occurs in rats due to an increasing state of restfulness and reduced behavioral activity. SWDs occur only in a state of calm wakefulness (Danober et al., 1998). Against this background of SWD reduction over time, and absence of a positive control, it cannot be ruled out that a potential effect of RTG/EZG may have been masked. However, given the very different mechanistic basis for absence seizures compared with other seizure types, available results from the two absence models suggest that positive modulation of KCNQ (Kv7) channels is not effective. SWDs are characteristic of absence seizures, and are generated by thalamocortical circuits (Steriade, 2005). KCNQ (Kv7) channels are found at high levels in the cortex and thalamus (Geiger et al., 2006). Therefore, the potential lack of action of RTG/EZG on absence seizures and epilepsy does not reflect a low density of KCNQ (Kv7) channels in relevant brain areas, but rather, may be linked to the underlying difference in seizure networks. Absence seizures constitute a particular type of epilepsy that may be related to excess GABAergic transmission in the thalamus, whereas in focal or convulsive epilepsy an excess of excitation is prevalent (Danober et al., 1998). Further studies will be necessary to clarify this point and whether there is any potential efficacy of KCNQ (Kv7) positive modulators in absence epilepsy.

Genetic Models

  1. Top of page
  2. Summary
  3. Human Serum and Brain Concentrations
  4. In Vitro Studies
  5. Models of Epileptiform Activity
  6. Human Epileptic Brain Tissue
  7. In Vivo Models of Partial Epilepsy (Partial and Secondarily Generalized Seizures)
  8. Temporal Lobe Epilepsy and Epileptogenesis
  9. Generalized Tonic–Clonic Seizures
  10. Generalized Epilepsy (Absence Seizures)
  11. Genetic Models
  12. Status Epilepticus
  13. Neuroprotection
  14. CNS Tolerability Profile of AEDs
  15. Comparison with Other AEDs
  16. Combination Treatments
  17. Conclusions
  18. Acknowledgments
  19. Disclosure
  20. References
  21. Supporting Information

Many idiopathic epilepsies are thought to arise from genetic variation or mutation; approximately 20 genetic variants have been identified that give rise to specific heritable forms of the disease, and it is likely that a complex interplay between multiple genes accounts for many more (Frankel, 2009). Mouse models that mimic the genetic mutations of Mendelian epilepsy syndromes have been generated, and include a Kcnq2 mutant mouse that mimics human benign familial neonatal convulsions (BFNC), an autosomal dominantly inherited idiopathic epilepsy of newborns characterized by partial or generalized clonic convulsions that start on days 2–4 of life. Gene loci for this inherited epilepsy were originally mapped to chromosomes 20q13.3 and 8q24, and the associated genes were subsequently identified as KCNQ2 and KCNQ3, respectively (Biervert et al., 1998; Singh et al., 1998; Charlier et al., 1998). Variants in KCNQ2 or KCNQ3 genes associated with BFNC have been shown to give rise to KCNQ (Kv7) channels with reduced conductance and alterations in their voltage-dependence of activation (Wuttke et al., 2007; Volkers et al., 2009). Where specific mutated channels have been expressed in oocytes, the depolarizing shift of activation observed for the mutant channels could be reversed by RTG/EZG (Wuttke et al., 2007), suggesting that seizures resulting from these mutations might also respond to RTG/EZG. This hypothesis remains to be tested.

In addition, different rodent strains that display increased seizure susceptibility or an epilepsy phenotype have been identified, including the dibenzylamine and Frings mouse models of sensory-evoked seizures that display TCS in response to a strong auditory stimulus (Hall, 1947; Skradski et al., 2001); genetically epilepsy-prone rats (GEPRs) that also display audiogenic seizures (Reigel et al., 1986); and the GAERS model mentioned earlier. RTG/EZG completely protected GEPRs from sound-induced TCS at doses of approximately 3 mg/kg, i.p., for moderate-seizure (GEPR-3) and 10 mg/kg, i.p., for severe-seizure (GEPR-9) animals (60 min posttreatment in each case) (Dailey et al., 1995). The relevance of these results to human familial epilepsy syndromes is unclear, since the genetic variations underlying the GEPR phenotype are unknown, although reductions in noradrenaline and serotonin transmission have been implicated (Dailey et al., 1995). RTG/EZG has not been shown to augment transmission of either of these monoamines, and it is likely that the efficacy of RTG/EZG relates to a reduction in excitability brought about by positive modulation of KCNQ (Kv7) channels. These results thus confirm the broad-spectrum efficacy of RTG/EZG in models of epilepsy characterized by seizures of idiopathic origin.

Candidate genes accounting for seizure susceptibility have been identified in some of the inbred rodent strains mentioned. For example, an inward rectifier K+ channel gene, KCNJ10 (Kir4.1), has been linked to seizure susceptibility in DBA/2 mice (Ferraro et al., 2007). The channels are predominantly found on astrocytes in the CNS and have a role in K+ transport; thus loss-of-function mutations could alter the balance of CNS excitability through changes in K+ homeostasis. Variants in the gene have also been linked to human epilepsy (Buono et al., 2004; reviewed in Frankel, 2009). Studies with RTG/EZG found that the drug was able to prevent audiogenic TCS in DBA/2 mice with ED50 values of 3–10 mg/kg. i.p. (De Sarro et al., 2001). The susceptibility gene Mass1 (monogenic audiogenic seizure-susceptible, also known as Mgr1) has been linked to audiogenic seizures in the Frings mouse (Skradski et al., 2001), and febrile seizures in humans (Nakayama et al., 2002). As part of the NINDS screen, RTG/EZG was shown to prevent auditory-evoked seizures in Frings mice with an ED50 of 2.1 mg/kg, i.p. (Kupferberg HJ: Valeant, data on file). These results suggest broad-spectrum anticonvulsant efficacy of RTG/EZG at doses likely to be achievable in clinical use, although to date RTG/EZG has not been demonstrated to have broad-spectrum efficacy in clinical epilepsy.

Status Epilepticus

  1. Top of page
  2. Summary
  3. Human Serum and Brain Concentrations
  4. In Vitro Studies
  5. Models of Epileptiform Activity
  6. Human Epileptic Brain Tissue
  7. In Vivo Models of Partial Epilepsy (Partial and Secondarily Generalized Seizures)
  8. Temporal Lobe Epilepsy and Epileptogenesis
  9. Generalized Tonic–Clonic Seizures
  10. Generalized Epilepsy (Absence Seizures)
  11. Genetic Models
  12. Status Epilepticus
  13. Neuroprotection
  14. CNS Tolerability Profile of AEDs
  15. Comparison with Other AEDs
  16. Combination Treatments
  17. Conclusions
  18. Acknowledgments
  19. Disclosure
  20. References
  21. Supporting Information

As mentioned earlier, SE is a prolonged seizure that can have very severe consequences and be fatal. Generalized convulsive SE can be induced by the peripheral injection of d,l-homocysteine thiolactone (d,l-HCT) to rats in which a cortical lesion has been generated by the application of cobalt (Walton & Treiman, 1988). Although the epileptogenic mechanism is not completely understood, it appears to be due to an activation of glutamatergic transmission. Several older anticonvulsants, such as PHT, PB, and VPA have been shown to reduce seizure severity at therapeutically relevant doses (Walton & Treiman, 1988, 1992), whereas LTG was ineffective (Walton et al., 1996). In a previously unpublished study, RTG/EZG 5 mg/kg, i.p., caused a statistically significant delay in the onset of first GTCS induced by injection of D,L-HCT to cobalt-lesioned rats (latency to first GTCS: 38.6 ± 10.9 s; n = 10 vs. 17.8 ± 7.3 s, n = 14, p < 0.05) (Voronina TA: Valeant, data on file). RTG/EZG also reduced the severity of EEG discharges, but did not prevent the onset of focal motor seizures. PHT 50 mg/kg, i.p., was parallel-tested in the study and provided similar protection. In a related study, the same group explored the ability of RTG/EZG and PHT to interrupt d,l-HCT–induced SE when the AEDs were administered immediately after the second GTCS. RTG/EZG 5 mg/kg, i.p., significantly reduced the number of subsequent GTCS and completely blocked further seizures in 62% of rats (Voronina TA: Valeant, data on file). At a higher dose (15 mg/kg, i.p.), RTG/EZG reduced seizure number and severity, but to a lesser degree than at the lower dose, and was associated with increased behavioral activity. In the same study, PHT 150 mg/kg, i.p., was unable to interrupt the ongoing SE and did not reduce the number of seizures or their severity. The efficacy of RTG/EZG in clinical SE has not been assessed.

Neuroprotection

  1. Top of page
  2. Summary
  3. Human Serum and Brain Concentrations
  4. In Vitro Studies
  5. Models of Epileptiform Activity
  6. Human Epileptic Brain Tissue
  7. In Vivo Models of Partial Epilepsy (Partial and Secondarily Generalized Seizures)
  8. Temporal Lobe Epilepsy and Epileptogenesis
  9. Generalized Tonic–Clonic Seizures
  10. Generalized Epilepsy (Absence Seizures)
  11. Genetic Models
  12. Status Epilepticus
  13. Neuroprotection
  14. CNS Tolerability Profile of AEDs
  15. Comparison with Other AEDs
  16. Combination Treatments
  17. Conclusions
  18. Acknowledgments
  19. Disclosure
  20. References
  21. Supporting Information

Consensus suggests that there is a relationship among seizures, neurodegeneration, and progressive worsening of epilepsy (Sutula et al., 2003). In humans, longitudinal studies that map seizure history with neuronal damage have been conducted using noninvasive imaging techniques. These studies suggest that progressive damage does occur (Fuerst et al., 2003; Bernasconi et al., 2005; Bernhardt et al., 2009a,b). Consequently, the ability of an AED to protect against seizure-induced neuronal damage is an important characteristic that may favor use of that drug (Pitkänen & Sutula, 2002). Indirect evidence of the potential neuroprotective properties of RTG/EZG can be obtained from in vitro and in vivo studies using models of neurotoxicity. Seyfried et al. (2000) found that RTG/EZG 10 μm could inhibit necrotic death of PC12 cells induced by l-glutamate, but did not prevent apoptotic cell death induced by l-DOPA. In a later study using organotypic hippocampal cultures, RTG/EZG protected against loss of dentate granule cells induced by serum withdrawal (IC50 0.4 μm) (Boscia et al., 2006). Protection was observed even in the presence of linopirdine or XE-991, two KCNQ (Kv7)–selective blockers, suggesting that efficacy was not mediated by KCNQ (Kv7) channels in this model. RTG/EZG 10 μm also protected these cultures against NMDA or oxygen/glucose deprivation (Boscia et al., 2006), consistent with the PC12 study by Seyfried et al. (2000). In contrast, Rekling (2003) found that RTG/EZG did not protect hippocampal slice cultures against oxygen/glucose deprivation, although several other AEDs tested in their model were effective. In an in vivo model of neurodegeneration induced by hippocampal administration of 4AP, RTG/EZG reduced cell loss from the cornu ammonis-1 region measured 24 h post-4-AP (Mora & Tapia, 2005). Both drugs were infused via a dialysis probe, which allowed simultaneous measurement of extracellular glutamate. 4-AP increased glutamate levels, which most likely contributed to pyramidal cell loss. The increase was not observed when 4-AP was codialyzed with RTG/EZG. Surprisingly, given earlier studies showing efficacy versus 4-AP–induced epileptiform activity (see above), RTG/EZG failed to reduce the epileptiform electrical activity induced by the chemoconvulsant. Further studies to examine potential neuroprotective properties of RTG/EZG will be required, in particular in models with greater relevance to the progression of epilepsy.

CNS Tolerability Profile of AEDs

  1. Top of page
  2. Summary
  3. Human Serum and Brain Concentrations
  4. In Vitro Studies
  5. Models of Epileptiform Activity
  6. Human Epileptic Brain Tissue
  7. In Vivo Models of Partial Epilepsy (Partial and Secondarily Generalized Seizures)
  8. Temporal Lobe Epilepsy and Epileptogenesis
  9. Generalized Tonic–Clonic Seizures
  10. Generalized Epilepsy (Absence Seizures)
  11. Genetic Models
  12. Status Epilepticus
  13. Neuroprotection
  14. CNS Tolerability Profile of AEDs
  15. Comparison with Other AEDs
  16. Combination Treatments
  17. Conclusions
  18. Acknowledgments
  19. Disclosure
  20. References
  21. Supporting Information

The clinical utility of an AED depends on its tolerability as much as on its anticonvulsant efficacy. In clinical trials conducted with RTG/EZG, a progressive, dose-dependent increase in adverse events (AEs), particularly dizziness and somnolence, was observed (Brodie et al., 2010; French et al., 2011). The threshold variant of the electroshock model was used to assess the potency of RTG/EZG in preventing TCS in rats, and, in parallel the effect of the drug on spontaneous locomotor activity (sLMA) was monitored as an indicator of potential AEs (Large CH, personal communication). RTG/EZG significantly increased seizure threshold in a dose-dependent manner from as low as 1 mg/kg, p.o. (Fig. 2A), similar to previous reports by Rostock et al. (1996). However, in this model, a dose of 10 mg/kg was required to raise the seizure threshold to a comparable degree as a clinically relevant dose of LTG (3.2 mg/kg, p.o., which achieves plasma concentrations that are clinically effective; Large et al., 2009). Therefore, a 10 mg/kg dose of RTG/EZG in rats may be the most relevant indicator to predict the minimal blood concentration required for efficacy in humans. At 10 mg/kg, p.o. (but not 3 mg/kg), RTG/EZG significantly reduced spontaneous locomotor activity (Fig. 2B), again consistent with a previous report (Roeloffs et al., 2008).

image

Figure 2.   Efficacy of retigabine (RTG)/ezogabine (EZG) in (A) the rat electroshock seizure model and (B) spontaneous locomotor activity in an open field. Data shown are mean ± standard error of the mean (SEM) for each group (n = 12 per group for seizure threshold or n = 6 for locomotor activity). Drugs were administered 60 min before seizure threshold determination, or 30 min before measurement of spontaneous activity over a 60 min test period. *p < 0.05, **p < 0.01 with respect to the vehicle group, Wilcoxon test. LTG, lamotrigine; Veh, vehicle-treated control.

Download figure to PowerPoint

Pharmacokinetic analysis of blood samples in these studies showed that the blood concentration of RTG/EZG 90 min after a 10 mg/kg, p.o., dose was 790 ng/ml (Large CH, personal communication). This is consistent with plasma levels associated with anticonvulsant efficacy in humans. These data show that doses of RTG/EZG that give plasma concentrations in rats similar to those found to be clinically effective are also effective in the animal model, but are associated with a significant impairment of locomotor activity. A caveat is that animal studies consider efficacy and AE liability in healthy animals after a single dose of drug. It is well known that dose titration can improve tolerability to AEDs in humans, and patients with epilepsy may tolerate AEDs better than healthy subjects (Schmidt, 2007). Thus the available animal studies may underestimate the potential therapeutic window for AEDs and for RTG/EZG in patients with epilepsy.

Comparison with Other AEDs

  1. Top of page
  2. Summary
  3. Human Serum and Brain Concentrations
  4. In Vitro Studies
  5. Models of Epileptiform Activity
  6. Human Epileptic Brain Tissue
  7. In Vivo Models of Partial Epilepsy (Partial and Secondarily Generalized Seizures)
  8. Temporal Lobe Epilepsy and Epileptogenesis
  9. Generalized Tonic–Clonic Seizures
  10. Generalized Epilepsy (Absence Seizures)
  11. Genetic Models
  12. Status Epilepticus
  13. Neuroprotection
  14. CNS Tolerability Profile of AEDs
  15. Comparison with Other AEDs
  16. Combination Treatments
  17. Conclusions
  18. Acknowledgments
  19. Disclosure
  20. References
  21. Supporting Information

Preclinical studies conducted with RTG/EZG demonstrate that this novel investigational AED possesses broad-spectrum anticonvulsant activity in a battery of established seizure and epilepsy models, including GTCS, primary generalized, and partial epilepsies (Kupferberg HJ: Valeant, data on file). Overall, RTG/EZG is mechanistically unique and possesses an anticonvulsant profile that most closely parallels VPA and felbamate (FBM), and is clearly differentiated from the Na+ channel modulators PHT, CBZ, and LTG (Table 2).

Table 2.   Comparative anticonvulsant profile of RTG/EZG, a dihydrochloride salt of RTG/EZG, and selected prototype anticonvulsants in mice and rats
Test substanceMiceRats
MESs.c. Mets.c. Bics.c. PicAGSMESs.c. MetKindled seizures
  1. Data were obtained from a study conducted on behalf by the National Institute of Neurological Disorders and Stroke, Epilepsy Branch, Anticonvulsant Screening Program (1995) (Kupferberg HJ: Valeant, data on file).

  2. +, protection at doses producing no behavioral toxicity; +/−, protection at doses producing some behavioral toxicity; −, <50% protection at highest dose tested; AGS, audiogenic seizure; Bic, bicuculline; MES, maximal electroshock seizure; Met, metrazole; N.D., not determined; Pic, picrotoxin; RTG/EZG, retigabine (ezogabine); s.c., subcutaneous.

RTG/EZG+/−+/−N.D.N.D.+++/−+
RTG/EZG dihydrochloride salt+++/−+++/−+
Valproate+++/−+/−+++/−+
Felbamate+++++++
Phenytoin++++/−
Lamotrigine+++N.D.
Carbamazepine++/−+++/−
Gabapentin++N.D.+N.D.
Ethosuximide++/−+++
Clonazepam+++++/−++

Combination Treatments

  1. Top of page
  2. Summary
  3. Human Serum and Brain Concentrations
  4. In Vitro Studies
  5. Models of Epileptiform Activity
  6. Human Epileptic Brain Tissue
  7. In Vivo Models of Partial Epilepsy (Partial and Secondarily Generalized Seizures)
  8. Temporal Lobe Epilepsy and Epileptogenesis
  9. Generalized Tonic–Clonic Seizures
  10. Generalized Epilepsy (Absence Seizures)
  11. Genetic Models
  12. Status Epilepticus
  13. Neuroprotection
  14. CNS Tolerability Profile of AEDs
  15. Comparison with Other AEDs
  16. Combination Treatments
  17. Conclusions
  18. Acknowledgments
  19. Disclosure
  20. References
  21. Supporting Information

Approximately 30% of patients with epilepsy have uncontrolled seizures that are resistant to available AEDs (Kwan & Brodie, 2006). This in itself is a strong motivation to continue the development of novel AEDs, particularly those with a unique mechanism of action such as RTG/EZG. However, there is also an opportunity to maximize efficacy through judicious use of AED combinations, and logic suggests that coadministration of AEDs with different pharmacologic action is more likely to confer improved efficacy than combinations of drugs with similar pharmacology. Clinical studies conducted with RTG/EZG already confirm its ability to provide additional efficacy as add-on therapy to existing AEDs in patients with residual seizures. There are too few clinical data thus far to determine whether specific combinations might be particularly effective, but two studies conducted in rodent models may provide a starting point for directed combination treatment in patients (De Sarro et al., 2001; Luszczki et al., 2009).

In the first of these, De Sarro et al. (2001) found that RTG/EZG, at a dose ineffective alone (0.5 mg/kg, i.p.), significantly augmented the efficacy of a range of AEDs in audiogenic seizures in DBA/2 mice. For example, add-on low-dose RTG/EZG lowered the apparent ED50 for VPA in clonic seizures from 43 to 20 mg/kg. A similar degree of augmentation was observed with DZP, PB, and PHT. Less marked, but still significant, augmentation was observed with the two Na+ channel blockers: CBZ and LTG. No significant augmentation was observed with FBM. RTG/EZG did not affect plasma levels of the AEDs, suggesting that the observed augmented efficacy was due to a pharmacodynamic interaction (De Sarro et al., 2001). Importantly, the study showed that combination with low-dose RTG/EZG did not alter the TD50 (dose of drug that causes a toxic response in 50% of the population) for CNS side effects in the mice, thus suggesting that the combination effectively improves therapeutic index.

A more detailed study evaluating the ability of RTG/EZG to augment efficacy of CBZ, LTG, or VPA compared protection against supramaximal electroshock afforded by monotherapy with each AED over various dose ranges, or in combination with RTG/EZG at various dose ranges (Luszczki et al., 2009). An isobolographic analysis of the results allowed an assessment of whether the combination provided additive or synergistic protection. Consistent with the findings of De Sarro et al. (2001), Luszczki et al. (2009) found that RTG/EZG combined with either LTG or CBZ provided additive protection against electroshock-induced TCS, and synergistic protection with VPA. The study confirmed an absence of pharmacokinetic interaction in each pair combination, with the exception of increased VPA concentrations when administered at a high dose in combination with low-dose RTG/EZG. Combination treatment did not affect various measures of CNS side effects, including motor performance (chimney test), long-term memory (light–dark, step-through passive avoidance task), and skeletal muscle strength (grip strength test). However, individual AED doses were well below what might cause these events when administered alone; therefore, strong conclusions about the potential for deleterious effects of RTG/EZG on AEs cannot be made.

Conclusions

  1. Top of page
  2. Summary
  3. Human Serum and Brain Concentrations
  4. In Vitro Studies
  5. Models of Epileptiform Activity
  6. Human Epileptic Brain Tissue
  7. In Vivo Models of Partial Epilepsy (Partial and Secondarily Generalized Seizures)
  8. Temporal Lobe Epilepsy and Epileptogenesis
  9. Generalized Tonic–Clonic Seizures
  10. Generalized Epilepsy (Absence Seizures)
  11. Genetic Models
  12. Status Epilepticus
  13. Neuroprotection
  14. CNS Tolerability Profile of AEDs
  15. Comparison with Other AEDs
  16. Combination Treatments
  17. Conclusions
  18. Acknowledgments
  19. Disclosure
  20. References
  21. Supporting Information

We have considered studies of RTG/EZG in seizure and epilepsy animal models in vitro and in vivo. The results of these studies confirm the broad-spectrum efficacy across different seizure types, and the notable potency of RTG/EZG even in models refractory to several other AEDs and/or where there is significant pathology, including resected human epileptic brain tissue. As a consequence of the drug’s ability to potentiate KCNQ (Kv7) channel activity, RTG/EZG is able to resist and block the occurrence of seizures induced by a range of stimuli across different regions of the brain, including those regions known to be the origin for partial seizures. RTG/EZG is, therefore, well placed to treat epilepsy, in which neuronal hyperexcitability leads to seizure.

Less clear at this stage in its clinical development is whether RTG/EZG will also be effective in controlling seizures characterized by hypersynchrony. For example, although idiopathic generalized epilepsies or absence seizures characterized by SWDs involve hyperexcitability within specific neural networks, these networks are driven by thalamic activity intimately linked to mechanisms underlying sleep (Hughes, 2009). Although KCNQ (Kv7) channels are present in the thalamus (Schroeder et al., 2000; Saganich et al., 2001), direct study of rat thalamocortical neurons suggest that the channels only marginally regulate excitability (Kasten et al., 2007). Studies on the effects of RTG/EZG on thalamocortical activity under conditions that lead to SWDs would be valuable. In addition, it remains to be determined whether RTG/EZG might possess disease-modifying or antiepileptogenic properties. In vitro studies that examined the neuroprotective properties of RTG/EZG are promising, but studies in chronic models allowing the assessment of emerging spontaneous seizures and the ensuing neuronal damage will be useful in providing information of greater potential clinical relevance.

Finally, studies examining combinations of RTG/EZG with established AEDs suggest that combination with VPA or other GABAergic AEDs may be of particular benefit and should be investigated further. It remains to be determined whether these combinations might be effective across different seizure types, and also whether they might serve to improve upon the therapeutic index of either drug alone.

Disclosure

  1. Top of page
  2. Summary
  3. Human Serum and Brain Concentrations
  4. In Vitro Studies
  5. Models of Epileptiform Activity
  6. Human Epileptic Brain Tissue
  7. In Vivo Models of Partial Epilepsy (Partial and Secondarily Generalized Seizures)
  8. Temporal Lobe Epilepsy and Epileptogenesis
  9. Generalized Tonic–Clonic Seizures
  10. Generalized Epilepsy (Absence Seizures)
  11. Genetic Models
  12. Status Epilepticus
  13. Neuroprotection
  14. CNS Tolerability Profile of AEDs
  15. Comparison with Other AEDs
  16. Combination Treatments
  17. Conclusions
  18. Acknowledgments
  19. Disclosure
  20. References
  21. Supporting Information

Charles H. Large, David M. Sokal, and Martin J. Gunthorpe are former employees of GlaxoSmithKline. Astrid Nehlig has received basic research support from Johnson & Johnson Pharmaceutical Research and Development, Astra Charnwood, Parke Davis, and Valeant Pharmaceuticals. Chris Crean is an employee of Valeant Pharmaceuticals North America. Raman Sankar has received basic research support from Valeant Pharmaceuticals and NTP, has participated in clinical trials sponsored by Pfizer, and has served as a speaker and/or paid consultant for GlaxoSmithKline, UCB, Lundbeck, NTP, and Sunovion. Kevan E. VanLandingham is an employee of GlaxoSmithKline. H. Steve White has served as a paid consultant to Johnson & Johnson Pharmaceutical Research and Development, GlaxoSmithKline, Valeant Pharmaceuticals, Eli Lilly & Co., and Upsher-Smith Laboratories, Inc., is a member of the UCB Pharma Speakers Bureau, the NeuroTherapeutics Pharma Scientific Advisory Board, and has received research funding from NeuroAdjuvants, Inc. We confirm that we have read the Journal’s position on issues involved in ethical publication and affirm that this article is consistent with those guidelines.

References

  1. Top of page
  2. Summary
  3. Human Serum and Brain Concentrations
  4. In Vitro Studies
  5. Models of Epileptiform Activity
  6. Human Epileptic Brain Tissue
  7. In Vivo Models of Partial Epilepsy (Partial and Secondarily Generalized Seizures)
  8. Temporal Lobe Epilepsy and Epileptogenesis
  9. Generalized Tonic–Clonic Seizures
  10. Generalized Epilepsy (Absence Seizures)
  11. Genetic Models
  12. Status Epilepticus
  13. Neuroprotection
  14. CNS Tolerability Profile of AEDs
  15. Comparison with Other AEDs
  16. Combination Treatments
  17. Conclusions
  18. Acknowledgments
  19. Disclosure
  20. References
  21. Supporting Information

Supporting Information

  1. Top of page
  2. Summary
  3. Human Serum and Brain Concentrations
  4. In Vitro Studies
  5. Models of Epileptiform Activity
  6. Human Epileptic Brain Tissue
  7. In Vivo Models of Partial Epilepsy (Partial and Secondarily Generalized Seizures)
  8. Temporal Lobe Epilepsy and Epileptogenesis
  9. Generalized Tonic–Clonic Seizures
  10. Generalized Epilepsy (Absence Seizures)
  11. Genetic Models
  12. Status Epilepticus
  13. Neuroprotection
  14. CNS Tolerability Profile of AEDs
  15. Comparison with Other AEDs
  16. Combination Treatments
  17. Conclusions
  18. Acknowledgments
  19. Disclosure
  20. References
  21. Supporting Information

Figure S1. Typical EEG recordings of SWDs occurring either in the presence or absence of retigabine.

Figure S2. Effects of increasing doses of retigabine given via i.p. administration on SWD duration in GAERS.

Figure S3. Effects of increasing doses of retigabine per os on SWD number in GAERS.

Data S1. Effects of retigabine on the expression of spike-and-wave discharges (SWDs) in GAERS.

FilenameFormatSizeDescription
EPI_3364_sm_FigS1.tif400KSupporting info item
EPI_3364_sm_FigS2.tif705KSupporting info item
EPI_3364_sm_FigS3.tif685KSupporting info item
EPI_3364_sm_Supporting-information.docx24KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.