Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Anticonvulsants: Clinical
SUMMARY
Anticonvulsants target a range of different molecular mechanisms, including voltage-gated sodium channels, the α 2 δ subunits of calcium channels, SV2A, K v 7 potassium channels, and AMPA. In some cases the exact mechanism of anticonvulsant action has not yet been elucidated or may involve multiple molecular mechanisms. A key attribute of anticonvulsants is that they directly or indirectly modulate neuronal excitability. This suggests that anticonvulsants efficacious in treating epilepsy may also be efficacious for neuropathic pain conditions, where neuronal hyperexcitability is also thought to play an important role. However, caution needs to be exercised because the pathophysiological mechanisms underlying hyperexcitability in epileptic conditions may be quite different from those underlying neuropathic pain conditions. Use of anticonvulsants for neuropathic pain has its origins in the empirical use of agents such as carbamazepine for trigeminal neuralgia and gabapentin for post-herpetic neuralgia, which ultimately led to these agents being licensed for these indications. Overall, however, the empirical use of a range of different anticonvulsant agents has not resulted in the eventual licensed use of most anticonvulsants for neuropathic pain conditions. This is primarily because, as in the case of lamotrigine, early evidence of efficacy from case studies and open-label investigations has not translated into an efficacy signal in adequately powered, randomized controlled trials. The aim of this chapter is to review evidence of efficacy from published or otherwise publically available placebo-controlled, double-blind, randomized clinical trial results of the use of commonly available anticonvulsants for a range of neuropathic pain conditions comprehensively.
Introduction
Anticonvulsants have been used off-label for a variety of non-epileptic conditions, including migraine headache, chronic neuropathic pain (NP), mood disorders, schizophrenia, and various neuromuscular syndromes ( ). In fact, the efficacy of carbamazepine for trigeminal neuralgia (TN) is almost diagnostic of the condition. As discussed in Chapter 35 , anticonvulsants act on a number of different molecular targets, all of which play a role in modulating neuronal excitability ( , ). Most notably these targets include voltage-gated sodium channel (VGSC) blockers, modulation of the α 2 δ subunits of calcium channels, γ-aminobutyric acid type A receptor (GABA A ) agonists, blockade of GABA transporter 1, inhibition of GABA transaminase, synaptic vesicle 2A (SV2A) modulators, K v 7 potassium channels activators, α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) blockers, and anticonvulsants that act through multiple mechanisms ( Table 36-1 ). It is important to recognize that even anticonvulsants that act predominantly via a single mechanism may have multiple modes of action that may contribute to their efficacy and/or tolerability profile. Overall, the fundamental attribute of modulation of neuronal hyperexcitability has helped confer the therapeutic utility of anticonvulsants not only for the treatment of epilepsy but also for a wide range of neurological conditions, including NP. In this respect these drugs may better be termed “neuromodulators” rather than “anticonvulsants” given their utility in a variety of conditions.
Table 36-1
Targeted Mechanisms of Licensed Anticonvulsants
| TARGETED MECHANISM | MEDICINES |
|---|---|
| Voltage-gated sodium channels | Carbamazepine, oxcarbazepine, lamotrigine, phenytoin, lacosamide |
| Modulation of α 2 δ subunits of calcium channels | Gabapentin and pregabalin |
| GABA A agonists | Phenobarbital, benzodiazepines |
| Blockade of GABA transporter 1 | Tiagabine |
| Inhibition of GABA transaminase | Vigabatrin |
| SV2A modulators | Levetiracetam |
| K v 7 potassium channels activators | Retigabine/ezogabine |
| AMPA blockers | Perampanel (undergoing regulatory review Dec 2011) |
| Multiple mechanisms | Topiramate, valproic acid |
AMPA, α-amino3-hydroxy-5-methyl-4-isoxazolepropionic acid; GABA, γ-aminobutyric acid; SV2A, synaptic vesicle 2A.
NP conditions are complex biopsychosocial phenomena that continue to be an enigma and a challenge in clinical practice. The current International Association for the Study of Pain definition of NP ( http://www.iasp-pain.org/AM/Template.cfm?Section=Pain_Definitions ) is “pain caused by a lesion or disease of the somatosensory nervous system.” It may be peripheral or central in origin. In some cases, however, it may not be possible to identify a lesion or a disease that satisfies the established neurological diagnostic criteria. As detailed in a recent review by Baron and co-authors (2010), pain is frequently the initial symptom, but there are often other positive symptoms associated with the pain, including hyperalgesia, allodynia, paresthesias, and shooting, electric shock–like symptoms. As well as these positive symptoms, there are negative symptoms, including reductions in mechanical, vibratory, and thermal sensations. Interestingly, these positive and negative symptoms may co-exist in adjacent or even the same areas. Current therapy is focused on the treatment of positive symptoms rather than negative symptoms. As stated by Baron and co-authors (2010), NP is associated with a number of disease states, such as post-herpetic neuralgia (PHN), diabetic painful neuropathy (DPN), drug-induced neuropathy (e.g., antiretroviral therapy), infection (e.g., human immunodeficiency virus [HIV] neuropathy), trauma (e.g., thoracotomy), and nerve entrapment (e.g., carpal tunnel syndrome). These states may be primarily due to a focal injury or a more generalized neuropathy. What is remarkable is that in general, there is no defined pattern of symptoms associated with any one clinically defined NP condition. Indeed, many NP conditions share a similar pattern of symptoms, albeit with a large range of variability (see Baron and colleagues 2010 for detailed review).
Given the overlap in clinical symptoms between different NP conditions, it is possible that no single pathophysiological mechanism is likely to be associated with an individual clinically defined condition. Consistent with this statement, therapeutic interventions that are efficacious for one NP condition, such as pregabalin, are also efficacious in a number of other NP conditions and yet have little reported efficacy in managing acute or chronic nociceptive pain conditions. This suggests that these NP-specific interventions are not acting as general analgesics but somehow target a common mechanism associated with positive symptoms in each of these conditions.
It is, however, unlikely that one mechanism will explain the pain and sensory changes that accompany NP conditions. Indeed, as has recently been reviewed by , Baron and co-authors (2010), and elsewhere in this volume, there is evidence of increases in ectopic activity in nerve fibers associated with a nerve lesion or disease in the peripheral and/or central nervous system (CNS). Furthermore, sensitization of the CNS, as well as longer-term plastic changes in the CNS and peripheral nervous system, may also contribute to the pain and associated sensory changes ( ). Evidence suggests that a potentially complex interplay between multiple molecular, neuro-immune, and synaptic mechanisms contributes to the manifestation of ectopic activity, central sensitization, and plasticity. Despite this complexity, a common theme is the relative importance of targeting neuronal hypersensitivity. Many of the molecular targets identified as being contributors to the neuronal hyperexcitability in NP conditions, such as voltage-gated sodium and calcium channels, are similar to those thought to contribute to the neuronal hyperexcitability in epilepsy and targeted by anticonvulsant drugs. As a result of this better understanding of the pathophysiological mechanisms and on the basis of a number of case studies, as well as open-label clinical studies, randomized, placebo-controlled trials have been carried out in recent years to investigate the efficacy of newer anticonvulsants in treating a number of different NP conditions. The remainder of this chapter comprehensively reviews the clinical evidence generated in investigations of the analgesic efficacy of anticonvulsants for a variety of NP states. The focus is on results from published, randomized, placebo-controlled, double-blind clinical studies.
Anticonvulsants that Act by Blocking Voltage-Gated Sodium Channels
As detailed in Chapter 35 , given the potential importance of neuronal hyperexcitability in the peripheral nervous system and central sensitization in the CNS in contributing to the pain, there is strong rationale for investigating the analgesic efficacy of anticonvulsants with sodium channel–blocking properties in treating a variety of NP conditions. In the past there was considerable confusion regarding the efficacy of these medications for NP. This has been partly due to a number of case reports and/or open-label studies reporting efficacy in small groups of patients, only for this efficacy signal to all but disappear in randomized controlled trials (RCTs). In recent years, the quality of clinical evidence supporting the effectiveness of anticonvulsants with sodium channel–blocking properties, including lamotrigine ( ) and carbamazepine ( ), has been evaluated in a number of systematic reviews. In this section we review double-blind, randomized, placebo-controlled clinical trials that evaluate the efficacy of anticonvulsants with sodium channel–blocking properties.
Carbamazepine
Carbamazepine acts primarily by blocking VGSCs but also has pharmacological activity at calcium channels, as well as other molecular targets of anticonvulsant drugs, and is used as a specific treatment of the pain associated with TN. The most common adverse events associated with carbamazepine, particularly during the initial phases of therapy, are dizziness, drowsiness, unsteadiness, nausea, and vomiting. As a result, carbamazepine requires titration from 200 mg daily to the therapeutic dose, which is usually 400–800 mg daily for the treatment of TN. Carbamazepine has a “boxed warning” regarding the risk for severe dermatological reactions, including toxic epidermal reaction and Stevens–Johnson syndrome, as well as actions on the hematopoietic system, including a low risk for aplastic anemia and agranulocytosis.
Four randomized, placebo-controlled trials investigating TN were conducted in the 1960s, all of which demonstrated the superior analgesic efficacy of carbamazepine over placebo ( Table 36-2 ; , , , ). One important consideration is that the studies had relatively small numbers of subjects and did not benefit from modern clinical trial methodologies for assessing analgesic efficacy, with some studies not using pain measurement scales, not measuring quality-of-life end points, or not carrying out the investigation for a long enough duration. In addition, the use of high doses for a few days carries a risk that subjects were functionally unblinded to the study treatment. Regardless, the continued clinical use of carbamazepine for the treatment of TN and the head-to-head clinical comparisons of newer treatments with carbamazepine (see for review) confirm the utility of this anticonvulsant for the treatment of TN. It is recommended as the first-line treatment of TN in the guidelines for the treatment of TN by the American Academy of Neurology and the European Federation of Neurological Societies Taskforce ( ).
Table 36-2
Summary of Published Randomized, Placebo-Controlled Clinical Trials of Carbamazepine for Neuropathic Pain Conditions
| PATIENT POPULATION | TREATMENTS | DURATION | DESIGN | NUMBER OF SUBJECTS | PRIMARY END POINT | TRIAL OUTCOME | SUMMARY |
|---|---|---|---|---|---|---|---|
| TN ( ) | CZP, 200 mg Placebo |
3 days | Crossover | 9 | Patient preference | Positive | 8/9 patients expressed a preference for CZP; 1/9 found both treatments equally effective |
| TN, PHN, other ( ) | CZP, 400 mg–1 g Placebo |
5 days | Crossover | TN: n = 30 PHN: n = 6 Other: n = 6 |
Positive | 19/27 TN patients with a complete or very good response, placebo response minimal or absent in all cases | |
| TN ( ) | CZP (flexible dosing) Placebo |
2 wk | Crossover | 70 | 4-point scale | Positive | CZP more effective than placebo in the treatment of TN |
| CRPS (type 1) ( ) | First phase of study: CZP, 600 mg/day Placebo |
8 days | Parallel group | 38 | NRS | Positive | After inactivation with a spinal cord stimulator, a significant number of subjects receiving CZP had a delay in increases in pain in comparison to the placebo-treated group |
| DPN ( ) | CZP, 200–600 mg Placebo |
2 wk | Crossover | 30 randomized | Pain intensity | Positive | 28/30 patients receiving CZP improved versus 19/30 with placebo 0/30 worsened with carbamazepine versus 11/30 with placebo (27% dropout rate) |
| TN, facial pain ( ) | CZP, 2.4 g/day Placebo |
2–42 mo | Crossover | Facial pain: n = 64 TN: n = 54 |
4-point scale | Mixed | Results presented only on 44 with TN because of insufficient follow-up 15/20 starting CZP had a good or excellent response 12/17 switched from placebo to CZP 6/7 taking placebo only had a good or excellent response |
| DPN ( ) | CZP, 300–600 mg Nortriptyline, 10 mg, plus fluphenazine, 0.5 mg |
30 days | Crossover | 16 | 100-mm VAS | Negative | No significant difference between treatments |
CRPS, complex regional pain syndrome; CZP, carbamazepine; DPN, diabetic painful neuropathy; NRS, numerical rating scale; PHN, post-herpetic neuralgia; TN, trigeminal neuralgia; VAS, visual analog scale.
Only one fully randomized, placebo-controlled trial of carbamazepine for a non-orofacial pain condition has been published. In this study, subjects in whom a spinal cord stimulator was implanted and complex regional pain syndrome (CRPS type 1) was diagnosed had the stimulator switched off before receiving carbamazepine or placebo ( ; see Table 36-2 ). There was a significant delay in increases in pain in subjects who received carbamazepine compared with subjects who received placebo, which led the authors to conclude that carbamazepine had an analgesic effect. It is yet to be seen whether these results can be replicated in a larger and broader population of NP patients. In a non–placebo-controlled, crossover study design, reported that neither carbamazepine nor nortriptyline–fluphenazine had a significant impact on pain ratings in patients with DPN.
Oxcarbazepine
Oxcarbazepine is an antiepileptic drug indicated for use as monotherapy or adjunctive therapy for the treatment of partial seizures. The proposed mechanism of action is blockade of VGSCs, inhibition of high-threshold calcium channels, and enhancement of the potassium rectifier ( ). Though structurally related to carbamazepine, evidence suggests that differences in the pharmacological mode of action of oxcarbazepine may result in a different clinical profile ( ).
Oxcarbazepine is used as monotherapy or adjunctive therapy for the treatment of partial seizures in adults, and the most common adverse events when used as monotherapy are nausea, dizziness, somnolence, and headache. Oxcarbazepine has been associated with a risk for severe dermatological reactions.
Three randomized, placebo-controlled studies investigated the analgesic efficacy of oxcarbazepine for DPN ( Table 36-3 ). One of the three studies met the primary end point: a significantly greater reduction in pain ratings at the end of treatment with 1800 mg oxcarbazepine in comparison to placebo ( ). Furthermore, some of the secondary end points, including the proportion of subjects with greater than a 50% reduction in pain, as well as quality-of-life measures and the durability of the analgesic effect, were consistent with a greater analgesic effect of oxcarbazepine than placebo. However, in a second DPN study ( ) that was of similar size and design, there was no indication of analgesic efficacy. The target dose in this study, however, was lower (1200 mg), and the magnitude of placebo effect was double that of the above study (see Table 36-3 ). In the third study ( ), the primary end point was not met with doses of up to 1800 mg even though the magnitude of the placebo response was consistent with that in the first study ( ; Table 36-3 ). Therefore, although oxcarbazepine has shown a positive signal of efficacy for DPN, it has not been demonstrated consistently. Even though double-blind, randomized, placebo-controlled studies have not yet been carried out with oxcarbazepine for TN, open-label studies have demonstrated analgesic efficacy in patients refractory to carbamazepine treatment ( ; Gomez-Arguelles et al 2008), and it is recommended as second-line treatment ( ).
Table 36-3
Summary of Published Randomized, Placebo-Controlled Clinical Trials of Oxcarbazepine for Neuropathic Pain Conditions
| PATIENT POPULATION | TREATMENTS | DURATION | DESIGN | NUMBER OF SUBJECTS | PRIMARY END POINT | TRIAL OUTCOME | NOTES |
|---|---|---|---|---|---|---|---|
| DPN ( ) | OXC, 1800 mg/day max Placebo |
4-wk titration 12-wk maintenance |
Parallel group | OXC: n = 69 Placebo: n = 7 |
VAS | Positive | Significant reduction in pain from baseline in last week of treatment 1800 mg: −24.3; placebo: 14.7; P = 0.0108 50% responder rate significant: 35.2% for OXC, 18.4% for placebo |
| DPN ( ) | OXC, 1200 mg/day Placebo |
4-wk titration 12-wk maintenance |
Parallel group | OXC: n = 71 Placebo: n = 70 |
VAS | Negative | No significant difference between active and placebo in change from baseline in pain score to last week of treatment 1200 mg: −27.9; placebo: −31.1 |
| DPN ( ) | Placebo OXC, 600 mg/day OXC, 1200 mg/day OXC, 1800 mg/day |
4-wk titration 12-wk maintenance |
Parallel group | OXC, 600 mg: n = 83 OXC, 1200 mg: n = 87 OXC, 1800 mg: n = 88 Placebo: n = 89 |
VAS | Negative | No significant difference between active and placebo in change from baseline in pain score to last week of treatment Primary end point: 1800 mg, −26.5; 1200 mg, −29.0; 600 mg, −25.9; placebo, −19.1 |
DPN, diabetic painful neuropathy; OXC, oxcarbazepine; VAS, visual analog scale.
Phenytoin
Phenytoin (phenytoin sodium) is an anticonvulsant that is indicated in the United States for the control of generalized tonic–clonic (grand mal) and complex partial (psychomotor, temporal lobe) seizures and for the prevention and treatment of seizures occurring during or following neurosurgery. The most common adverse events associated with phenytoin are nystagmus, ataxia, slurred speech, decreased coordination, and mental confusion. The exact mechanism of anticonvulsant action of phenytoin has not yet been fully determined, but it is thought to act primarily by blocking VGSCs. Relatively few studies have been conducted to determine the efficacy of phenytoin for NP. In a randomized, double-blind, placebo-controlled study in a mixed NP population totaling 20 subjects, intravenous infusion of phenytoin, 15 mg/kg, over a 2-hour period resulted in significantly greater relief in comparison to the placebo group ( ). Specifically, although placebo infusion was associated with a reduction in the reported sensations of sensitivity and numbness, infusion with phenytoin was reported to additionally achieve a reduction in the sensations of burning pain, shooting pain, and overall pain as measured on a visual analog scale (VAS). In other double-blind, placebo-controlled studies of phenytoin, conflicting results have been reported regarding the efficacy of phenytoin in patients with DPN ( , ). Overall, there are surprisingly limited good-quality clinical trial data to support the utility of phenytoin for NP.
Lamotrigine
Lamotrigine, a phenyltriazine chemically unrelated to other antiepileptic drug treatments, is indicated for adult adjunctive therapy, pediatric adjunctive therapy, and adult monotherapy for epilepsy in Europe. In the United States and other countries, lamotrigine is indicated for use as adjunctive therapy for partial seizures in adults and the pediatric population older than 2 years, for generalized seizures of the Lennox–Gastaut syndrome, and for primary tonic–clonic seizures in adults and the pediatric population older than 2 years. Lamotrigine is also licensed for the maintenance and treatment of bipolar disorder. The most common adverse reactions (incidence ≥10%) in adult epilepsy clinical studies were dizziness, headache, diplopia, ataxia, nausea, blurred vision, somnolence, rhinitis, and rash.
A number of randomized, placebo-controlled trials investigating the efficacy of lamotrigine for NP have been conducted on the basis of preclinical evidence and following the publication of a number of case reports. These provided a compelling argument for exploring the efficacy of lamotrigine in patients with NP, such as a report of sustained efficacy of 300–600 mg lamotrigine in four patients with central NP refractory to other treatments ( ). As detailed in Table 36-4 , randomized, placebo-controlled trials have been carried out for a wide range of NP conditions, including TN ( ), DPN ( , ), HIV neuropathy–associated pain ( , ), chemotherapy-associated pain ( ), and central pain conditions ( , ). Although some studies have had a limited number of subjects ( , ), other studies have been powered to maximize the chance of detecting a treatment effect ( , , ).
Table 36-4
Summary of Published Randomized, Placebo-Controlled Clinical Trials of Lamotrigine for Neuropathic Pain Conditions
| PATIENT POPULATION | TREATMENTS | DURATION | DESIGN | NUMBER OF SUBJECTS | PRIMARY END POINT | TRIAL OUTCOME | NOTES |
|---|---|---|---|---|---|---|---|
| TN ( ) | LTG, 400 mg Placebo |
14 days | Crossover | 14 randomized | Composite measure | Positive | LTG was significantly superior to placebo ( P = 0.011) based on analysis of a composite efficacy index. Efficacy for one treatment over another was determined according to a hierarchy of (1) use of escape medication, (2) total pain scores, or (3) global evaluations. Eleven of the 13 patients eligible for inclusion in the composite efficacy index showed better efficacy with LTG than with placebo (all patients stable with carbamazepine or phenytoin) |
| DPN ( ) | LTG, 200 mg LTG, 300 mg LTG, 400 mg Placebo |
7-wk titration 12-wk maintenance |
Parallel group | LTG, 200 mg: n = 88 LTG, 300 mg: n = 89 LTG, 400 mg: n = 90 Placebo: n = 88 (60 centers) |
11-point NRS | Negative | Study 1: No significant difference in adjusted mean change from baseline at week 19 in any treatment group: 200 mg, −1.93; 300 mg, −2.87; 400 mg, −2.49; placebo, −2.22 No significant effects on any secondary end points Note: Large placebo effect |
| DPN ( ) | LTG, 200 mg LTG, 300 mg LTG, 400 mg Placebo |
7-wk titration 12-wk maintenance |
Parallel group | LTG, 200 mg: n = 86 LTG, 300 mg: n = 85 LTG, 400 mg: n = 84 Placebo: n = 84 (62 centers) |
11-point NRS | Mixed | Study 2: Significant difference in adjusted mean change from baseline at week 19 in 400-mg LTG group versus placebo only ( P < 0.023) in analysis of observed cases LOCF analysis was not significant for any treatment arms: 200 mg, −2.32; 300 mg, −2.30; 400 mg, −2.72; placebo, −1.61 Note: Large placebo effect |
| Mixed NP ( ) | LTG Placebo |
8-wk titration 6-wk maintenance |
Parallel group | LTG: n = 112 Placebo: n = 111 |
11-point NRS | Negative | No significant difference in adjusted mean change from baseline to last week of treatment: LTG, −2.13; placebo, −2.11 (LTG adjunctive to gabapentin, tricyclic antidepressants, or non-opioid analgesics) |
| DPN ( ) | LTG, 400 mg Placebo |
8-wk titration | Parallel group | LTG: n = 29 Placebo: n = 30 |
11-point NRS | Positive | Change from baseline daily pain in LTG-treated group was significantly reduced ( P < 0.001): LTG, −2.2; placebo, −1.2 Results of the MPQ, PDI, and BDI remained unchanged in both groups Global assessment of efficacy favored LTG treatment over placebo |
| Central pain–SCI ( ) | LTG, 400 mg Placebo |
9 wk | Crossover | Randomized: n = 30 (22 completed) | VAS | Mixed | No significant difference between LTG and placebo in change in median pain score from baseline to last week of treatment In patients with incomplete SCI, LTG significantly reduced pain at or below the level of SCI—8 responded to LTG and 3 responded to placebo, but to a lesser extent. Responder defined as reduction in pain of 2 or more Dropout rate, 27% |
| Central post-stroke pain ( ) | LTG, 200 mg Placebo |
8 wk | Crossover | Randomized: n = 30 (22 completed) | 11-point NRS | Mixed | Significant reduction in change from baseline pain score with LTG versus placebo: LTG, −1.0; placebo, 0 Dropout rate, 27% |
| Cancer pain ( ) | LTG, 300 mg Placebo |
8-wk titration 2-wk maintenance |
Parallel group | LTG: n = 63 Placebo: n = 64 |
11-point NRS | Negative | No statistically significant difference between LTG and placebo in change from baseline: LTG, −0.3; placebo, −0.5 |
| General NP ( ) | LTG, 200 mg Placebo |
8-wk titration | Parallel group | Completers: LTG: n = 36 Placebo n = 38 |
VAS | Negative | No significant change from baseline in average pain levels with LTG or placebo |
| HIV neuropathy ( ) | LTG, 300 mg Placebo |
7-wk titration 7-wk maintenance |
Parallel group | LTG: n = 20 Placebo n = 22 |
Gracey Pain Scale score | Positive | Significant reduction in average pain in final week versus baseline in LTG group ( P = 0.03): LTG, −0.55; placebo, −0.18 The reduction in pain in the LTG group is equivalent to a decrease from moderate to less than very mild pain and is clinically significant |
| HIV neuropathy ( ) | LTG, 400 mg Placebo |
7-wk titration 4-wk maintenance |
Parallel group | LTG: n = 150 Placebo: n = 77 |
Gracey Pain Scale score | Mixed | Mean change from baseline in Gracely Pain Scale score for average pain was not different between LTG and placebo at the end of the maintenance phase However, in subjects receiving neurotoxic ART, the slope of the change in the Gracely Pain Scale score for average pain reflected greater improvement with LTG than with placebo ( P = 0.004), as did the mean change from baseline scores on the VAS for pain intensity and the MPQ and patient and clinician ratings of global impression of change in pain ( P ≤ 0.02) |
ART, antiretroviral therapy; BDI, Beck Depression Inventory; DPN, diabetic painful neuropathy; HIV, human immunodeficiency virus; LOCF, last observation carried forward; LTG, lamotrigine; MPQ, McGill Pain Questionnaire; NP, neuropathic pain; NRS, numerical rating scale; PDI, Pain Disability Index; PHN, post-herpetic neuralgia; SCI, spinal cord injury; TN, trigeminal neuralgia; VAS, visual analog scale.
The earliest evidence of the efficacy of lamotrigine in a randomized, placebo-controlled trial comes from an adjunctive study of refractory TN ( ). Despite the small size of the crossover study (see Table 36-4 ), a clear difference between placebo and lamotrigine treatment was seen across both periods, although this difference did decrease during the second period because of a higher than expected placebo response ( ). Consistent with this finding, in a non–placebo-controlled study ( ), 16 of 20 subjects reported complete remission of pain associated with TN following the administration of doses of between 100 and 400 mg. This study benefited from concurrent measurement of plasma lamotrigine concentrations, and it appeared that a wide variation in plasma levels and doses was required in different subjects to achieve the same level of pain relief. However, within each subject, a steady increase in the level of pain relief correlated with an increase in plasma concentration and dose.
The clinical development of lamotrigine for the treatment of NP in patients with non-orofacial pain conditions showed early promise when baseline daily pain in the group treated with 400 mg lamotrigine was significantly reduced in comparison to placebo in a small placebo-controlled trial of DPN ( ; Table 36-4 ). Subsequently, however, inconsistent results were generated in two larger placebo-controlled studies (Study 1 and Study 2; Table 36-4 ) with multiple dose levels and more than 80 subjects per treatment group ( ). In Study 1, no difference was seen between the three dose levels of lamotrigine and placebo at week 19, whereas study 2 reported a significant difference in adjusted mean change from baseline at week 19 in the 400 mg lamotrigine group versus placebo ( P < 0.023). This was true for the observed cases analysis only and not for the last observation carried forward, which was considered important by the authors given the high dropout rate. The results from Study 1 are consistent with those of a study conducted in a mixed NP population ( ; Table 36-4 ), in which no significant difference was found between placebo and lamotrigine even following subgroup analysis of subjects with DPN. This confirms that when taken together, evidence of the efficacy of lamotrigine in relieving painful DPN is at best variable and not easy to detect in randomized, placebo-controlled trials ( Table 36-4 ). It is also unlikely that a significant subgroup of DPN patients respond more favorably to lamotrigine inasmuch as no difference was seen in the 30 or 50% responder rate between the different dose levels of lamotrigine and placebo in either Study 1 or Study 2 ( ). Furthermore, secondary end point analysis did not consistently reveal a significant difference between lamotrigine and placebo ( ).
The potential for greater analgesic efficacy in subgroups of patients has been explored in HIV-associated painful neuropathies. In a study reported by , the overall mean change in average pain was not different between lamotrigine and placebo. However, in the subset of subjects receiving neurotoxic antiretroviral treatment, the slope of the change in the Gracely Pain Scale score for average pain reflected greater improvement with lamotrigine than with placebo ( P = 0.004), as did the mean change from baseline scores on the VAS for pain intensity and the McGill Pain Questionnaire and patient and clinician ratings of global impression of change in pain ( P ≤ 0.02). Similarly, in a non–placebo-controlled study of TN, results from the five patients in the symptomatic neuralgia group who had multiple sclerosis were less variable than the results from the group of 15 subjects with the “essential” form of TN ( ). The authors concluded that the more homogeneous nature of disease in the symptomatic neuralgia group may have resulted in a more consistent requirement for dosages and plasma levels of lamotrigine to achieve analgesic efficacy. At present, however, the balance of evidence does not form a compelling argument for subsets of NP patients who respond better to lamotrigine.
The analgesic efficacy of lamotrigine for central NP conditions is also not clear. For example, in a crossover study, a low 200-mg dose of lamotrigine led to a greater reduction in pain ratings than placebo did ( ; Table 36-4 ). However, in patients with spinal cord injury, no significant difference was found between lamotrigine and placebo ( ; Table 36-4 ), although there appeared to be a greater lamotrigine effect in subjects with incomplete spinal cord injury.
Overall, as detailed in Table 36-4 , little evidence from large RCTs supports the use of lamotrigine for the treatment of a range of different NP conditions at the doses administered in clinical trials (see also a recent systematic review by ). Finally, no evidence supports lamotrigine as add-on therapy to existing treatment since there was no evidence of efficacy for lamotrigine in a mixed NP population maintained on stable doses of gabapentin and/or amitriptyline ( ; Table 36-4 ).
Lacosamide
Lacosamide is a functionalized amino acid that is currently licensed as an antiepileptic drug for use as an adjunctive therapy in the treatment of partial onset seizures in adults. The most common adverse reactions (≥10% and greater than with placebo) are diplopia, headache, dizziness, and nausea. Although the exact mechanism of action of lacosamide has not yet been elucidated, the mechanism whereby lacosamide modulates VGSCs appears to differ from that of other anticonvulsants that also act on VGSCs, such as carbamazepine, phenytoin, and lamotrigine. Specifically, lacosamide appears to reduce VGSC availability by selective enhancement of slow inactivation but without the apparent interaction with fast inactivation gating ( ; see Chapter 35 ).
Even though the efficacy of lacosamide for PHN and mixed NP has been explored, the main focus has been on DPN, for which four randomized, placebo-controlled studies have been carried out, three of which have appeared as full publications ( , , ; Table 36-5 ). The key finding has been positive signs of analgesic efficacy with lacosamide, 400 mg or higher, in all three published studies, although in one of the three published studies the primary end point was not met ( ; Table 36-5 ). Furthermore, in one of the unpublished studies the primary end point was also not met (European Medicines Agency identifier: EMEA/CHMP/658067/2008). Overall, 400 mg lacosamide demonstrated superiority over placebo in all three published studies across a range of different end points, though not always consistently. With respect to the primary end point, treatment differences from placebo were less than 1 point on an 11-point numerical rating scale (NRS) ( Table 36-5 ), the established “gold standard” for an effective treatment.
Table 36-5
Summary of Published Randomized, Placebo-Controlled, Clinical Trials of Lacosamide for Neuropathic Pain Conditions
| PATIENT POPULATION | TREATMENTS | DURATION | DESIGN | NUMBER OF SUBJECTS | PRIMARY END POINT | TRIAL OUTCOME | SUMMARY |
|---|---|---|---|---|---|---|---|
| DPN ( ) | Placebo LCM, 400 mg/day or MTD |
6-wk titration 4-wk maintenance |
Parallel group | LCM: n = 60 Placebo: n = 59 |
11-point NRS | Positive | LCM afforded significantly ( P = 0.039) better pain relief than did placebo in comparison to baseline LOCF: LCM, −3.11; placebo, −2.21 Observed cases: LCM, −3.72; placebo, −2.28 |
| DPN ( ) | Placebo LCM, 200 mg/day LCM, 400 mg/day LCM, 600 mg/day |
6-wk titration 12-wk maintenance |
Parallel group | Placebo: n = 65 LCM, 200 mg: n = 141 LCM, 400 mg: n = 125 LCM, 600 mg: n = 137 |
11-point NRS | Mixed | Patients randomized to LCM, 400 mg/day, experienced a mean reduction in pain scale scores of 2.5 (39%), from 6.4 to 3.9. For patients in the placebo group, the mean reduction was 1.8 (29%), from 6.2 to 4.4. The estimated treatment difference in the LSMean value between LCM, 400 mg/day, and placebo (see Table 36-2 ) was −0.61 (95% CI; −1.23, 0.00) and was at the level of significance ( P = .0507) for the primary end point (last 4 wk of maintenance period) |
| DPN ( ) | Placebo LCM, 200 mg/day LCM, 400 mg/day LCM, 600 mg/day |
6-wk titration 12-wk maintenance |
Parallel group | Placebo: n = 93 LCM, 200 mg: n = 93 LCM, 400 mg: n = 91 LCM, 600 mg: n = 93 |
11-point NRS | Positive | For patients randomized to the LCM, 400 mg/day, group, the mean daily pain score decreased by 2.5 (38.5%), from 6.5 to 4. For placebo-treated patients, it decreased by 1.8 (27.3%), from 6.6 to 4.8. The estimated treatment difference value using least-square means was −0.74 (95% CI, −1.32, −0.16) between these 2 groups and was statistically significant ( P = 0.01) |
CI, confidence interval; DPN, diabetic painful neuropathy; LCM, lacosamide; LOCF, last observation carried forward; MTD, maximum tolerated dose; NRS, numerical rating scale; LS Mean, least squares mean.
Although 600 mg lacosamide had the potential for greater analgesic efficacy, this was not conclusively demonstrated because of poor tolerability in some subjects and the associated high withdrawal rates ( ). Specifically, 9 of 65 placebo subjects (14%), 17 of 141 receiving lacosamide, 200 mg/day (12%), 54 of 125 receiving lacosamide, 400 mg/day (43%), and 91 of 137 receiving lacosamide, 600 mg/day (66%), discontinued because of adverse events in this study. Adverse events reported following 400 and 600 mg lacosamide, respectively, were dizziness, 21.6 and 28.5%; headache, 8 and 13.1%; tremor, 9.6 and 14.6%; somnolence, 8 and 8.8%; and balance disorder, 4.8 and 9.5%. The incidence of these adverse events is broadly in line with that reported for these doses in clinical trials of epileptic subjects with partial onset seizures.
In the responder analysis in the study of , 58 and 44% of subjects receiving lacosamide, 400 mg/day, 58 and 30% of those receiving lacosamide, 600 mg/day, and 45 and 27% of subjects receiving placebo had greater than a 30 and 50% reduction in pain, respectively. In the study reported by , in an analysis of patient global impression of change, 81 and 37% of subjects randomized to 400 mg lacosamide versus 68 and 21% of placebo subjects reported feeling better and much better, respectively. Furthermore, 6% of 400 mg lacosamide and 17% placebo-treated subjects reported feeling much worse after treatment. Overall, these results support a primary end point analysis suggesting an improvement in pain ratings following 400 mg lacosamide in patients with DPN. It is also interesting to note that the responder rate results do not provide any indication that there are significant subgroups of subjects who respond more to lacosamide, although this would require further investigation.
The Committee for Medical Products for Human Use of the European Medicines Agency considered the application of lacosamide for the treatment of NP associated with DPN in adults as non-approvable. The major objections raised were “1) The clinical relevance of the observed efficacy has not been sufficiently demonstrated; 2) The safety profile, foremost the cardiac related effects, but also the CNS effects in perspective of the intended target population and the treatment length; 3) Benefit/risk relationship given the safety profile and questionable relevance of the observed effect” (European Medicines Agency identifier: EMEA/CHMP/658067/2008). Based on the view that the benefit versus adverse event profile of lacosamide for DPN is currently deemed unacceptable, it is unclear whether any future investigations are being planned with this anticonvulsant for this or other NP conditions.
Novel Sodium Channel Blockers
In the context of discussions regarding anticonvulsants mainly targeting VGSCs, it is appropriate to also consider selected novel oral sodium channel blockers at an earlier stage of development. These may not necessarily be considered for development as anticonvulsants at present.
Ralfinamide is a small-molecule α-amino amide derivative that is being developed for the treatment of NP conditions and is thought to act by blocking both tetrodotoxin (TTX)-resistant and TTX-sensitive sodium currents. In addition, ralfinamide is reported to block N -methyl- d -aspartate (NMDA) receptors, as well as modulate the release of neurotransmitters from primary afferent neurons. In a dose titration study of 80–320 mg in a mixed NP population (including patients with neuropathic back pain), the difference in pain ratings following ralfinamide ( n = 59) versus placebo ( n = 38) was −9.5 mm on a VAS scale ( P = 0.051). However, in the follow-up SERENA phase II/III 12-week study carried out in a total of 411 subjects with neuropathic back pain who received either 160 or 320 mg/day or placebo, the primary end point as measured on an 11-point NRS was not met ( http://www.newron.com/Ralfinamide.html ).
CNV1014802 is a small-molecule, state-dependent sodium channel blocker that is being developed for the treatment of NP. The efficacy of CNV1014802 is currently being examined in a randomized, double-blind, placebo-controlled study involving neuropathic back pain patients. In addition, the efficacy of this compound for TN is being explored ( http://www.convergencepharma.com/index.asp?page_id=14 ).
XEN402 is a novel sodium channel blocker with a reported 80-nM IC 50 (inhibitory concentration in 50%) potency of block at Na v 1.7 channels and is in development for the treatment of NP. The selectivity of XEN402 versus other sodium channel isoforms has not yet been reported. In a pilot placebo-controlled, randomized crossover study involving four subjects with inherited erythromelalgia as a result of mutations in Na v 1.7 ( ), 2 days of treatment with XEN402, 400 mg twice daily, increased the time to induce maximal pain by heating and reduced the amount of pain after induction ( P = 0.014). The efficacy of topical XEN402 in alleviating the pain of PHN has recently been reported, with clinically meaningful reductions in pain being achieved that were significantly greater for topical XEN402 than for placebo ( P = 0.049 for greater than 30% and P = 0.0078 for greater than 50%; http://www.xenon-pharma.com/product-candidates/pain/ ).
Sodium Channel Subtype–Selective Blockers in Early Clinical Development
PF-05089771 and PF-04531083 are novel small-molecule sodium channel blockers that are highly selective for the Na v 1.7 and Na v 1.8 sodium channel subtypes, respectively. To date, these compounds were found to be safe in phase I studies and well tolerated at systemic exposures reaching unprecedented high multiples of in vitro potency (Pfizer, unpublished data). These findings indicate that genuinely pharmacologically selective agents may be investigated in the clinic with the potential to enable pain relief without the CNS events observed with other less selective or non-selective compounds. The therapeutic utility of both these compounds for NP may be explored in the near future.
In summary, results from randomized, placebo-controlled trials of anticonvulsants with sodium channel–blocking properties in a variety of NP conditions have been disappointing (Wiffen et al 2011). Although there have been some encouraging data, such as results with lacosamide for DPN, the benefit–risk profile was insufficient to achieve registration of this compound for this condition. Similarly, studies of oxcarbazepine and lamotrigine have produced mixed results for DPN, as well as for some other NP conditions. One interpretation of these results is that these compounds did not achieve sufficient block of the sodium channels contributing to ongoing pain, allodynia, hyperalgesia, and other sensory disturbances in NP. This is a reasonable consideration since these compounds have relatively low potency for sodium channels and have been optimized for CNS disorders such as epilepsy. This is particularly the case because the maximal doses that can be administered and the requirements for slow dose titration are limited by the narrow therapeutic window with respect to the predominantly CNS adverse events associated with these treatments.
One interesting observation is that anticonvulsants with sodium channel–blocking properties do appear to offer therapeutic utility for TN, albeit with relatively few placebo-controlled data from RCTs to fully verify such utility. Indeed, carbamazepine, oxcarbazepine, and lamotrigine are recommended as first-, second-, and third-line treatments of TN in the guidelines for the treatment of TN from the American Academy of Neurology and the European Federation of Neurological Societies Taskforce ( ). One consideration is whether other focal neuropathies such as post-surgical NP and entrapment syndromes might be more responsive to sodium channel–blocking anticonvulsant activity than more generalized neuropathies such as DPN are.
You're Reading a Preview
Become a Clinical Tree membership for Full access and enjoy Unlimited articles
If you are a member. Log in here