A number of glutamate receptor ligands have been tested in humans and are in clinical trials as analgesics. A main focus has been on their therapeutic potential for neuropathic pain, a condition that continues to be notoriously difﬁcult to treat. Clinically tested compounds include antagonists for NMDA receptors and AMPA/kainate receptors. None of the mGluR compounds appear to have been studied clinically as potential analgesics. However, a group II agonist (LY354740) has been successfully tested in clinical trials for generalized anxiety disorders, and mGluR5 antagonists such as MPEP are on their way to clinical trials as well (Swanson et al. 2005).
NMDA Receptor Antagonists
Commercially available NMDA receptor antagonists, including ketamine, me- mantine, dextromethorphan, and its main metabolite dextrorphan, showed antinociceptive effects in preclinical studies (Sect. 3) and are analgesic in hu- mans (Parsons et al. 1999; Fisher et al. 2000; Weinbroum et al. 2000; Fundytus
2001; Parsons 2001; Kilpatrick and Tilbrook 2002; Henriksson and Sorensen
2002; Hewitt 2003; Carlsson et al. 2004). Evidence from preclinical and earlier clinical studies further suggested that NMDA receptor antagonists enhanced opioid analgesia and prevented the development of tolerance (Fundytus 2001; Parsons 2001), but a recent report of three controlled clinical trials failed to demonstrate such effects of a combination of morphine and dextromethorphan (Galer et al. 2005).
The noncompetitive NMDA receptor antagonist ketamine has long been used to induce and maintain anesthesia. Ketamine, which has analgesic properties at subanesthetic doses, is one of the most extensively studied NMDA recep- tor antagonists. The antinociceptive effects of ketamine in preclinical models of inﬂammatory and neuropathic pain have been well documented (Sect. 3; Fundytus 2001; Parsons 2001; Hewitt 2003). Ketamine is also analgesic in ex- perimental pain induced in healthy human subjects by intradermal capsaicin, ischemia, or burn (Fundytus 2001; Hewitt 2003). Interestingly, burn-induced pain was also inhibited by peripherally injected ketamine (Carlton 2001). In case reports and controlled studies, systemic (intravenous) ketamine produced relief of various forms of peripheral and central neuropathic pain (including postherpetic and posttraumatic neuralgias, phantom limb pain, and spinal cord injury), cancer pain, and pain in ﬁbromyalgia patients (Fisher et al.
2000; Fundytus 2001; Parsons 2001; Henriksson and Sorensen 2002; Hewitt
2003). Oral and epidural administrations of ketamine have also been reported to be effective. There is little evidence to support long-term treatment of chronic pain with ketamine. The usefulness of perioperative ketamine for the treatment of postoperative pain is controversial. Some studies reported ben- eﬁts of preemptive ketamine analgesia such as reduced postoperative pain and/or decreased analgesic consumption, but others found little or no evi- dence (Fundytus 2001; McCartney et al. 2004; Ong et al. 2005). Likewise, post- operative treatment has produced mixed results. Interestingly, intravenous administration of ketamine may be used as a diagnostic test to predict the analgesic response of neuropathic pain patients to dextromethorphan, which has a better side effect proﬁle (Cohen et al. 2004). Undesirable side effects of ketamine include dizziness, sedation, perceptual changes, and dissocia- tive symptoms, which typically disappear fairly quickly (Fisher et al. 2000; Fundytus 2001; Parsons 2001; Henriksson and Sorensen 2002). In summary, ketamine is effective in the treatment of neuropathic pain and perhaps ﬁ- bromyalgia but its usefulness is limited by the well-documented albeit tempo- rary side effects, abuse potential, and lack of evidence for long-term treatment beneﬁts.
A noncompetitive NMDA receptor antagonist memantine has been used for many years in the treatment of Parkinson’s disease and, more recently, de- mentia (Parsons et al. 1999; Kilpatrick and Tilbrook 2002). Memantine is clinically well tolerated and shows little of the side effects typically seen
with NMDA receptor antagonists, possibly because of the strong voltage de- pendence and rapid kinetics of its NMDA channel-blocking effects. Psy- chotomimetic effects, agitation/excitation, increased motor activity, and in- somnia have been reported but appear to be sporadic and at doses outside the recommend range (Parsons et al. 1999; Kilpatrick and Tilbrook 2002). Preclinical studies showed antinociceptive effects in animal models of in- ﬂammatory and neuropathic pain (see Sect. 3) although relatively high doses were used (Parsons et al. 1999). However, clinical trials have yielded less con- clusive data on its usefulness in the treatment of neuropathic pain (Parsons
2001; Kilpatrick and Tilbrook 2002). Memantine reduced nocturnal pain in patients with diabetic neuropathy but was not effective on spontaneous or evoked pain in patients with nerve injury from amputation or surgery (Kil- patrick and Tilbrook 2002). In summary, memantine may still have a place in the treatment of neuropathic pain because it is well tolerated and neu- ropathic pain represents one of the therapeutically most challenging pain conditions.
An orally available non-competitive NMDA receptor antagonist, dextromethor- phan, has a long history as a safe cough suppressant with few side effects (Fisher et al. 2000; Weinbroum et al. 2000; Fundytus 2001; Carlsson et al.
2004). Dextromethorphan showed antinociceptive effects in preclinical mod- els of inﬂammatory and neuropathic pain (Sect. 3). In clinical studies, pre-
treatment with dextromethorphan provided preemptive analgesia for acute postoperative pain, although others observed no such effect (Weinbroum
et al. 2000; Fundytus 2001; McCartney et al. 2004; Ong et al. 2005). There is, however, little evidence for analgesic effects of dextromethorphan in exper- imental pain in healthy human volunteers and in neuropathic pain, although a relatively high dose was analgesic in patients with diabetic or posttrau-
matic neuropathy (Weinbroum et al. 2000; Fundytus 2001; Henriksson and Sorensen 2002; Hewitt 2003; Carlsson et al. 2004). Intravenous ketamine has been proposed to have predictive value for the effectiveness of dextromethor-
phan in neuropathic pain (Cohen et al. 2004). Dextromethorphan appeared to be effective in a subgroup of patients with ﬁbromyalgia and again, intra- venous ketamine may help to select these patients (Henriksson and Sorensen
2002). It has also been suggested that dextromethorphan may enhance the analgesic effects of opioids and prevent opioid tolerance, but a recent study failed to show any clinical beneﬁt of MorphiDex (Endo Pharmaceuticals, Chadds Ford, PA), a combination of morphine and dextromethorphan (Galer et al. 2005).
Commercially available NR2B receptors antagonists such as CP-101606 (traxo- prodil) and ifenprodil were antinociceptive in preclinical models of inﬂamma- tory and neuropathic pain (Sect. 3; Parsons 2001; Chizh et al. 2001; Nakazato et al. 2005). CP-101606 showed analgesic effects in pain patients with peripheral neuropathy and spinal cord injury (Nakazato et al. 2005). CP-101606 and ifen- prodil have also been tested in clinical trials for the treatment of ischemic brain injury or stroke. They were well tolerated and did not cause ataxia, sedation, or impaired learning and memory at therapeutic concentrations, suggesting a better therapeutic index than other NMDA receptor antagonists (Chizh et al.
2001; Wang and Shuaib 2005). Unfortunately, some NR2B-selective antago- nists produced electrocardiographic abnormalities such as a prolonged Q-T interval through blockade of potassium channels (Parsons 2001). There is not enough clinical evidence yet to suggest any therapeutic advantages of NR2B antagonists in the treatment of pain.
GlycineB Site Antagonists
The presence of glycine at the strychnine-insensitive recognition site (glycineB) of the NMDA receptor is required for channel activation by glutamate or NMDA (Michaelis 1998; Dingledine et al. 1999; Wollmuth and Sobolevsky 2004; Mayer and Armstrong 2004; Mayer 2005). GlycineB site antagonists have been re- ported to lack the typical side effects of NMDA receptor antagonists (Par- sons 2001; Wallace et al. 2002). A glycineB antagonist (GV196771) showed antinociceptive effects in preclinical models of inﬂammatory (formalin and carrageenan evoked) and neuropathic pain (Parsons 2001; Wallace et al. 2002). However, a clinical study found no signiﬁcant effects of GV196771 on sponta- neous and mechanically or thermally evoked pain in patients with neuropathic pain associated with diabetic neuropathy, postherpetic neuralgia, complex re- gional pain syndrome, or peripheral nerve injury (Wallace et al. 2002). A sig- niﬁcant decrease of the size of the allodynic area was observed but did not translate into a reduction of pain (Wallace et al. 2002). The therapeutic value of glycineB antagonists for pain relief remains to be determined.
Side effects typically associated with NMDA receptor antagonists include mem- ory impairment, psychomimetic effects, ataxia, and motor incoordination as well as neurotoxicity (Parsons 2001; Haberny et al. 2002; Low and Roland
2004). Noncompetitive NMDA receptor antagonists such as memantine and
dextromethorphan are better tolerated and have fewer side effects, which has been explained by their relatively low afﬁnity, fast kinetics, and strong voltage dependence (Parsons et al. 1999; Parsons 2001). GlycineB- and NR2B-selective antagonists show a much better proﬁle in animal models of chronic pain than high-afﬁnity channel blockers and competitive NMDA receptor antago- nists (Parsons 2001; Chizh et al. 2001), but their clinical value as analgesics is not yet clear. Neurotoxic effects of NMDA receptor antagonists such as vac- uolizations were observed with competitive and some noncompetitive NMDA receptor antagonists whereas glycineB and NR2B antagonists did not show neurotoxicity (Low and Roland 2004). They appear to be dose and time de- pendent. Most studies, however, were done in animals (rats) and it is not clear if NMDA receptor antagonists induce neurotoxic effects in humans. Fur- thermore, several compounds have been shown to prevent NMDA antagonist
induced neurotoxicity; these include GABAergic, α2-adrenergic, anticholiner- gic, serotonergic, antipsychotic, and antiepileptic drugs (Haberny et al. 2002; Low and Roland 2004).
Non-NMDA Receptor Antagonists
Non-NMDA receptor antagonists produce antinociceptive effects in preclinical pain models but also inhibit normal nociceptive and non-nociceptive process- ing, which would suggest a narrow therapeutic window and make them unlikely drug targets for pain relief. Still, a systemically active non-NMDA receptor an- tagonist (LY293558, see the following section) has been tested in phase II clinical trials for pain. A selective AMPA receptor antagonist (YM872, zonam- panel) is being studied in phase II clinical trials for potentially neuroprotective effects in stroke. In preclinical studies, YM872 inhibited normal nociceptive responses and formalin-induced pain behavior (Nishiyama et al. 2004).
LY293558 is a competitive AMPA(GluR2)/kainate(GluR5) antagonist that show- ed antinociceptive effects in a preclinical study of formalin-induced pain be- havior (Simmons et al. 1998; Gilron 2001) but also inhibited nociceptive base- line responses (Von Bergen et al. 2002). In healthy human volunteers, systemic LY293558 signiﬁcantly reduced capsaicin-evoked pain and mechanical hyper- algesia and allodynia but had no effect on mechanical or heat pain evoked from normal skin (Sang et al. 1998). In clinical studies, LY293558 reduced spontaneous and movement-evoked postoperative pain following oral surgery (Gilron et al. 2000) and inhibited headache pain in patients with acute moder- ate or severe migraine attacks (Sang et al. 2004). It has been hypothesized that the analgesic effects of LY293558 involve its GluR5 antagonist action whereas GluR2 antagonism accounts for the side effects (see the following section).
In preclinical tests, LY293558 produced ataxia and motor deﬁcits, but no serious side effects were observed in clinical studies involving humans. Side effects included a mild transient visual impairment (hazy vision), dizziness, and sedation (Gilron 2001). Although LY293558 was reported to be well tolerated, the therapeutic range, duration of drug action, and effect of repeated dosing remain to be determined.
Ionotropic and metabotropic glutamate receptors (iGluRs and mGluRs) regu- late and ﬁne-tune neurotransmission and neuronal excitability throughout the nervous system and along the pain neuraxis through a variety of mechanisms. They are involved in a number of physiological and pathological processes and are emerging as promising therapeutic targets for the treatment of certain pain conditions in preclinical and some clinical studies. Despite the diversity of iGluR and mGluR mediated effects, it appears that antagonists for iGluRs and group I mGluRs and agonists for group II and III mGluRs can inhibit nociceptive processing in various parts of the nervous system. Among the iGluRs, NMDA and kainate receptors are more useful targets than AMPA re- ceptors. It has also been suggested that group I mGluR1 antagonists may be better candidates for pain relief than mGluR5 antagonists particularly for the treatment of neuropathic pain. Group II agonists appear to hold more promise than group III agonists, but the latter have been not been studied extensively and subtype-selective agonists are still largely missing.
Although some evidence suggests distinct antinociceptive effects of subtype- selective agents on mechanical versus thermal and inﬂammatory versus neu- ropathic pain and in different parts of the nervous system, a consistent pattern
has yet to emerge. The continued development of subtype-selective agents and analysis of their antinociceptive effects in different parts of the nervous system and in different pain models will provide answers in the near future. Of partic-
ular importance may be the fact that in the brainstem (PAG) some iGluRs and mGluRs produce effects opposite to those observed in the periphery and spinal cord. The role of glutamate receptors in higher brain centers in preclinical pain
models is still understudied and awaits a systematic and comparative analysis. The increasing availability of orally active compounds that can be tested and used clinically is extremely promising and suggests that certain iGluR and mGluR ligands can become novel pharmacological therapeutics for pain relief.
Glutamate receptor ligands may provide new alternatives or supplements to currently available analgesics.
Acknowledgements Work in the author’s laboratory is supported by National Institutes of
Health (NIH) grants NS38261 and NS11255.