- Email: [email protected]
NK1 (substance P) receptor antagonists – why are they not analgesic in humans?
VIEWPOINT
NK1 (substance P) receptor antagonists – why are they not analgesic in humans? Ray Hill Tachykinin NK1 receptor antagonists have failed to exhibit efficacy in clinical trials of a variety of clinical pain states. By contrast, in preclinical studies in animals NK1 receptor antagonists have been shown to attenuate nociceptive responses sensitized by inflammation or nerve damage, although they exhibit little effect on baseline nociception. Other agents with this profile of activity in animal tests, typically nonsteroidal anti-inflammatory drugs (NSAIDs), are analgesic in humans. Thus, NK1 receptor antagonists appear able to block behavioural responses to noxious and other stressful sensory stimuli at a level detectable in animal tests but fail to provide the level of sensory blockade required to produce clinical analgesia in humans. Clinical trials have now shown that drugs that block the actions of substance P by acting as antagonists at the tachykinin NK1 receptor fail to produce analgesia in a variety of clinical pain states. The observation that NK1 receptor antagonists are not effective against pain or migraine headache is interesting and important (see Ref. 1 for a review based on a workshop on this topic held at the World Congress on Pain, August 1999). The interest is increased by the parallel observation that several NK1 receptor antagonists do have clinical efficacy in chemotherapy-induced emesis2. The data now emerging with NK1 receptor antagonists form an important watershed in neuropeptide research and its significance extends beyond substance P. Several pharmacological antagonists for other neuropeptide receptors [e.g. corticotropin-releasing factor (CRF)] are currently in, or entering, clinical evaluation and the experience with antagonism of the actions of substance P will be a benchmark. Evidence supporting a principal role for substance P in mammalian nociception
R.G. Hill, Executive Director, Department of Pharmacology, Neuroscience Research Centre, Merck, Sharp and Dohme Research Laboratories, Harlow, Essex, UK CM20 2QR. E-mail: [email protected]
244
Much of the early evidence that suggested a role for substance P in mammalian nociception was circumstantial. Substance P (at that time identified by bioassay) had been found to be widespread in the CNS in the 1950s (Ref. 3). The observation that substance P was more abundant in dorsal rather than ventral roots led to the suggestion by Lembeck that substance P was a primary sensory neurotransmitter3. This, in turn, led to the association of substance P with pain because immunocytochemical studies by Hunt and colleagues4 showed that substance P was found in the smaller, unmyelinated sensory fibres. In addition, Henry showed that exogenous substance P, when applied to dorsal horn sensory neurones, had a slow onset and prolonged excitatory action that resembled the pattern of excitation observed after peripheral noxious stimuli4. Other workers have shown that the most important action of substance P might be to modulate the action of other transmitters, including other neuropeptides with similar actions on neurones, that are also found in unmyelinated sensory fibres. Multiple messengers often co-exist in the same neurones. In the same way as the activation of peripheral nociceptors is a result of an ‘inflammatory soup’ of algogens, the activation of
TiPS – July 2000 (Vol. 21)
secondary nociceptive circuits within the dorsal horn of the spinal cord is the result of parallel release of several transmitters and modulators from the terminals of primary afferent fibres4. Only if one of the released substances has a dominant role will the specific pharmacological blockade of its effects result in analgesia. At present, blockade of the effects of glutamate acting at NMDA receptors by drugs such as ketamine is the best example of a dominant role, in this case, for glutamate, which is probably contained in all sensory fibres. It is difficult to predict from clinical studies which transmitter might be the most important in a particular disease state. In patients with trigeminal neuralgia measurements of cerebrospinal fluid (CSF) showed that levels of somatostatin, noradrenaline, dopamine, 5-HT or metabolites were reduced compared with control subjects. However, in the same patients, substance P concentrations in CSF were increased5. Does this point to a pivotal role for substance P in this extremely painful condition? Alternatively, are those agents whose levels were reduced (suggesting increased turnover) more likely to be involved? Similarly, it has been argued that substance P has an important role in pain because treatment with the sensoritoxin capsaicin (after the initial painful stimulant effect has desensitized) depletes sensory fibres of substance P, reduces sensitivity to noxious stimuli in rats and produces analgesia in humans4. This result could equally well be a result of depletion of either co-existing calcitonin gene-related peptide (CGRP) (the most abundant peptide in sensory fibres) or other neuropeptides, or be a result of effects on the release of all of the diverse transmitters contained by fibres carrying the vanilloid VR1 (capsaicin) receptor. The result could also be explained by a direct lesioning effect of the sensoritoxin on nerve terminals. It has been argued that because the synthesis of substance P and upregulation of NK1 receptors are driven by nerve growth factor (NGF) acting at trkA receptors, there is a specific role for substance P in NGF-induced hyperalgesia. For example, it has been assumed that hyperalgesia observed in a transgenic NGFoverexpressing mouse is evidence for a role of substance P in nociception. The objection to the capsaicin experiment (above) can be raised in this case, in that many other substances contained in trkA-receptor-expressing nociceptive
0165-6147/00/$ – see front matter © 2000 Elsevier Science Ltd. All rights reserved.
PII: S0165-6147(00)01502-9
VIEWPOINT fibres could be implicated in addition to substance P. Indeed, there is a better correlation between the presence of CGRP in primary afferents and trkA receptors than there is between the presence of substance P and trkA receptors6. It is increasingly unlikely that any single peptide (or, for that matter, any other individual neurotransmitter) can be confidently associated with a specific sensory function4,7. Additionally, it is now accepted that sensory neurotransmission is plastic, that synthesis of transmitters and their receptors, including that of substance P, can be upregulated by conditions such as inflammation, which cause pain, and that synaptic connections often change after nerve injury8. We are thus faced with a nociception system that uses numerous transmitter substances in parallel and that adjusts its sensitivity up and down in response to tissue injury. In this respect, it is significant to note that when those lamina I dorsal horn neurones that carry the NK1 receptor are killed by intrathecal injection of substance P conjugated with the toxin saporin the antinociceptive effect observed is greater than that produced by NK1 receptor antagonists presumably because all input via these cells is removed9. Studies with knockout mice and selective receptor antagonists
The best way to reliably establish whether any one transmitter in this complex neurotransmission process has a pivotal role is to either block its receptor with a pharmacological antagonist or to remove the transmitter itself or the target receptor using transgenic gene deletion. Both approaches have now been used for NK1 (substance P) receptors and thus we have a body of objective data from which to draw conclusions. The phenotype of the knockout mice provides some, not totally convincing, evidence that NK1 receptor antagonists should be analgesic in humans. The preprotachykinin knockout mouse (which lacks both substance P and neurokinin A) has a substantially normal baseline nociception with attenuation of responses to intense noxious stimuli. The NK1 receptor knockout mice also exhibit no changes in acute nociception tests but display a reduction in responses to inflammatory stimuli and reduced wind up in dorsal horn neurones10. This is substantially the same picture that is observed in many of the published behavioural studies with the non-peptide, systemically active NK1 receptor antagonists that have been developed in the past decade. Antinociceptive effects were seen with a variety of NK1 receptor antagonists in the presence of nerve injury or inflammation but no unequivocal effects on nociception were observed in acute tests such as the hot plate2,11. In total, there were sufficient well conducted studies in which an NK1 receptor antagonist could be shown to attenuate the behavioural or electrophysiological response to a noxious stimulus to justify performing clinical trials for analgesia. After all, the profile of the compounds across the behavioural tests used was not dissimilar to that of nonsteroidal anti-inflammatory drugs (NSAIDs), which are known to be analgesic in humans. There was also enough negative evidence, however, to suggest that NK1 receptor antagonists were not going to be strong analgesics like morphine10,11. It is interesting to speculate whether the search for such compounds as analgesics would have been so widely
followed by the pharmaceutical industry if the knockout data had been available before the discovery of the non-peptide antagonists rather than ten years afterwards! Preclinical antinociceptive data in animals can be rationalized with the lack of effect in the clinic
It was a surprise to many of us when most of the clinical trials of NK1 receptor antagonists failed to show any analgesic effects whatsoever1 (but see Ref. 12 for a description of studies with CP99994 that showed some analgesia in dental pain). Obvious explanations for a lack of effect in humans are species differences in the physiology of substance P or differences between clinical pain and the type of noxious stimulus and response studied experimentally10. It is interesting to note that most of the differences in NK1 receptor distribution between species are expressed at supraspinal sites (and possibly the importance of these sites in pain perception has been underestimated), with remarkable similarities in the distribution of both the peptide and NK1 receptors in the dorsal horn of the spinal cord (the initial relay in the nociceptive process but not the level of sensation). In addition, the animal tests that showed antinociceptive activity with NK1 receptor antagonists have recently been reliably predictive of analgesic activity in humans with other novel strategies (e.g. in the case of cyclooygenase 2 inhibitors13). It is possible to reconcile the animal and clinical data if one includes the noxious stimuli used in the animal studies under the general heading of stressful stimuli. The published work on nociception then links up well with other animal experimental work on, for example, isolation-induced stress with NK1 receptor antagonists and NK1 receptor knockout mice2. This cumulative data set tells us that NK1 receptor antagonists can reliably attenuate the response to a stressful stimulus (e.g. whether produced by intraplantar formalin, nerve injury or by separation from a family group2). This effect is apparently not sufficient by itself to result in clinical analgesia. In support of this idea, the studies of Baulmann et al.14 on the response to peripheral injection of formalin as a noxious and stressful stimulus in rats are helpful. These workers showed that blockade of tachykinin receptors reduces the behavioural response to formalin and also reduces the induction of the immediate early gene Fos in brain areas linked to nociception and also in some of the brain areas that have been specifically associated with central stress responses. It has been noted that in veterinary practice it is difficult to distinguish the factors that might cause an animal distress. Pain is just one of these factors and clinically it might combine with other stressors to alter an animal’s behaviour15. This blurring of the distinction between pain, emotion and stress also extends to pharmacological treatment. Before the advent of effective drugs for treating depression, one accepted treatment, for what was then generically called melancholia, was opioid drugs such as morphine. A recent clinical trial in a small number of subjects showed relief of the symptoms of unipolar depression with the mu opioid peptide receptor partial agonist, buprenorphine, which until that time was only thought to be useful for the treatment of pain16. We are thus left with the working hypothesis that attenuating sensory neurotransmission, in a subtle way by interfering
TiPS – July 2000 (Vol. 21)
245
VIEWPOINT with the role of a single neuropeptide, might be adequate for reducing the behavioural response to a stressor such as inappropriate or intolerable sensory input. A more complete block of sensory input (such as can be produced by morphine) is needed to also achieve clinical pain relief. In terms of clinical benefit to individual patients it might be difficult to determine which factor is the more important. Loeser and Melzack17 wrote recently, in their introduction to a series of articles in the Lancet on pain, ‘Injury does not only produce pain, it also leads to stress […]. This process involves neural, hormonal, and behavioural activities.’ Concluding remarks
Acknowledgements I would like to thank all my colleagues, especially Nadia Rupniak, Sue Boyce, Mike Cumberbatch, Qing-Ping Ma, Jenny Laird, Jenny Longmore, Chris Swain and Richard Hargreaves for their data, discussions and ideas about the functions of substance P over the past ten years.
There is clearly still much to learn about the clinical effects of NK1 receptor antagonists. It is possible that there might be particular pain states in which the role of substance P is dominant and where NK1 receptor antagonists might be effective analgesics in humans, although I suspect such a situation will be infrequent. The possibility for using these drugs as adjuvants to existing analgesics, especially when there are other stress-related factors to consider, is a more attractive prospect. Additionally, it is relevant to ask whether the lack of ability to predict the presence or absence of analgesic properties in humans in the case of NK1 receptor antagonists has an impact on the discovery and clinical evaluation of other putative analgesics. What preclinical criteria should be used to determine whether clinical trials of a new analgesic are likely to be successful? It has recently been suggested that positron emission tomography (PET) studies in volunteers should be used to establish that adequate blocking doses are achieved in the brain, in humans before efficacy studies of a new analgesic are carried out18. It was suggested that the failure of a new inhibitor of enkephalin metabolism to exhibit analgesia in humans was a result of inadequate brain penetration even though it had shown antinociceptive effects in animal tests18. In fact, this criticism cannot be made, because some of the NK1 receptor antagonists that were found to be ineffective in humans had been subjected to PET studies and doses of drug that were adequate to saturate the brain receptors were used in the clinical efficacy studies (Ref. 2 and D. Burns et al., unpublished). Furthermore, similar doses of the same drugs were found to be effective in emesis studies, which are known to require both high levels of receptor occupancy and adequate brain penetration2. The final range of clinical uses of the NK1 receptor antagonists remains to be defined.
Selected references 1 Boyce, S. and Hill, R.G. (2000) Discrepant results from preclinical and clinical studies on the potential of substance P–receptor antagonist compounds as analgesics. In Proc 9th World Congress on Pain (Devor, M. et al., eds), pp. 313–324, IASP Press 2 Rupniak, N.M.J. and Kramer, M.S. (1999). Discovery of the antidepressant and anti-emetic efficacy of substance receptor (NK1) antagonists. Trends Pharmacol. Sci. 20, 485–490 3 Von Euler, U.S. (1985) The history of substance P. In Neurotransmitters in Action (Bousfield, D., ed.), pp. 143–150, Elsevier Biomedical Press 4 Salt, T.E. and Hill, R.G. (1983) Neurotransmitter candidates of somatosensory primary afferent fibres. Neuroscience 10, 1083–1103 5 Strittmatter, M. et al. (1997) Cerebrospinal fluid neuropeptides and monoaminergic transmitters in patients with trigeminal neuralgia. Headache 37, 211–217 6 Averill, S. et al. (1995) Immunocytochemical localization of trkA receptors in chemically identified subgroups of adult rat sensory neurons. Eur. J. Neurosci. 7, 1484–1494 7 Levine, J.D. et al. (1993) Peptides and the primary afferent nociceptor. J. Neurosci. 13, 2273–2286 8 Woolf, C.J. and Mannion, R.J. (1999) Neuropathic pain: aetiology, symptoms, mechanisms and management. Lancet 353, 1959–1964 9 Mantyh, P.W. et al. (1997) Inhibition of hyperalgesia by ablation of lamina-I spinal neurons expressing the substance P receptor. Science 278, 275–279 10 Ma, Q-P. and Hill, R.G. (1999). Neurokinin antagonists as potential agents for use in pain management. Curr. Opin. Invest. Drugs 1, 65–71 11 Rupniak, N.M.J. and Hill, R.G. (1999). Neurokinin antagonists. In Novel Aspects of Pain Management: Opioids and Beyond (Sawynok, J. and Cowan, A., eds), pp. 135–155, Wiley-Liss 12 Dionne, R.A. (1999) Clinical analgesic trials of NK1 antagonists. Curr. Opin. Invest. Drugs 1, 82–85 13 Chan, C.C. et al. (1999) Rofecoxib [Vioxx, MK-0966, 4-(49methylsulfonylphenyl)-3-phenyl-2-(5H )-furanone]: a potent and orally active cyclo-oxygenase2 inhibitor. Pharmacological and biochemical profiles. J. Pharmacol. Exp. Ther. 290, 551–560 14 Baulmann, J. et al. (2000) Tachykinin receptor inhibition and c-fos expression in the rat brain following formalin induced pain. Neuroscience 95, 813–820 15 Headley, P.M. and Livingston, A. (1989) Pain and stress in animals: problems of assessment and treatment. Front. Pain 1, 1–4 16 Bodkin, J.A. et al. (1995) Buprenorphine treatment of refractory depression. J. Clin. Psychopharmacol. 15, 49–57 17 Loeser, J.D. and Melzack, R. (1999) Pain: an overview. Lancet 353, 1607–1609 18 Langley, G. et al. (2000) Volunteer studies replacing animal experiments in brain research. Alternatives to Laboratory Animals 28, 315–331
Chemical name CP99994: (2S 3S)cis-3(2-methoxybenzylamino)-2-phenyl piperidine
Trends in Pharmacological Sciences : a forum for comment Controversial? Thought-provoking? If you have any thoughts or comments on articles published in Trends in Pharmacological Sciences, then why not write a Letter to the Editor at [email protected]? Letters should be no more than 500 words. Please state clearly whether you wish the letter to be considered for publication. The letter will be sent to the author of the original article for their response; both the letter and reply will be published together. Please note: submission does not guarantee publication.
246
TiPS – July 2000 (Vol. 21)
NK1 (substance P) receptor antagonists – why are they not analgesic in humans? Ray Hill Tachykinin NK1 receptor antagonists have failed to exhibit efficacy in clinical trials of a variety of clinical pain states. By contrast, in preclinical studies in animals NK1 receptor antagonists have been shown to attenuate nociceptive responses sensitized by inflammation or nerve damage, although they exhibit little effect on baseline nociception. Other agents with this profile of activity in animal tests, typically nonsteroidal anti-inflammatory drugs (NSAIDs), are analgesic in humans. Thus, NK1 receptor antagonists appear able to block behavioural responses to noxious and other stressful sensory stimuli at a level detectable in animal tests but fail to provide the level of sensory blockade required to produce clinical analgesia in humans. Clinical trials have now shown that drugs that block the actions of substance P by acting as antagonists at the tachykinin NK1 receptor fail to produce analgesia in a variety of clinical pain states. The observation that NK1 receptor antagonists are not effective against pain or migraine headache is interesting and important (see Ref. 1 for a review based on a workshop on this topic held at the World Congress on Pain, August 1999). The interest is increased by the parallel observation that several NK1 receptor antagonists do have clinical efficacy in chemotherapy-induced emesis2. The data now emerging with NK1 receptor antagonists form an important watershed in neuropeptide research and its significance extends beyond substance P. Several pharmacological antagonists for other neuropeptide receptors [e.g. corticotropin-releasing factor (CRF)] are currently in, or entering, clinical evaluation and the experience with antagonism of the actions of substance P will be a benchmark. Evidence supporting a principal role for substance P in mammalian nociception
R.G. Hill, Executive Director, Department of Pharmacology, Neuroscience Research Centre, Merck, Sharp and Dohme Research Laboratories, Harlow, Essex, UK CM20 2QR. E-mail: [email protected]
244
Much of the early evidence that suggested a role for substance P in mammalian nociception was circumstantial. Substance P (at that time identified by bioassay) had been found to be widespread in the CNS in the 1950s (Ref. 3). The observation that substance P was more abundant in dorsal rather than ventral roots led to the suggestion by Lembeck that substance P was a primary sensory neurotransmitter3. This, in turn, led to the association of substance P with pain because immunocytochemical studies by Hunt and colleagues4 showed that substance P was found in the smaller, unmyelinated sensory fibres. In addition, Henry showed that exogenous substance P, when applied to dorsal horn sensory neurones, had a slow onset and prolonged excitatory action that resembled the pattern of excitation observed after peripheral noxious stimuli4. Other workers have shown that the most important action of substance P might be to modulate the action of other transmitters, including other neuropeptides with similar actions on neurones, that are also found in unmyelinated sensory fibres. Multiple messengers often co-exist in the same neurones. In the same way as the activation of peripheral nociceptors is a result of an ‘inflammatory soup’ of algogens, the activation of
TiPS – July 2000 (Vol. 21)
secondary nociceptive circuits within the dorsal horn of the spinal cord is the result of parallel release of several transmitters and modulators from the terminals of primary afferent fibres4. Only if one of the released substances has a dominant role will the specific pharmacological blockade of its effects result in analgesia. At present, blockade of the effects of glutamate acting at NMDA receptors by drugs such as ketamine is the best example of a dominant role, in this case, for glutamate, which is probably contained in all sensory fibres. It is difficult to predict from clinical studies which transmitter might be the most important in a particular disease state. In patients with trigeminal neuralgia measurements of cerebrospinal fluid (CSF) showed that levels of somatostatin, noradrenaline, dopamine, 5-HT or metabolites were reduced compared with control subjects. However, in the same patients, substance P concentrations in CSF were increased5. Does this point to a pivotal role for substance P in this extremely painful condition? Alternatively, are those agents whose levels were reduced (suggesting increased turnover) more likely to be involved? Similarly, it has been argued that substance P has an important role in pain because treatment with the sensoritoxin capsaicin (after the initial painful stimulant effect has desensitized) depletes sensory fibres of substance P, reduces sensitivity to noxious stimuli in rats and produces analgesia in humans4. This result could equally well be a result of depletion of either co-existing calcitonin gene-related peptide (CGRP) (the most abundant peptide in sensory fibres) or other neuropeptides, or be a result of effects on the release of all of the diverse transmitters contained by fibres carrying the vanilloid VR1 (capsaicin) receptor. The result could also be explained by a direct lesioning effect of the sensoritoxin on nerve terminals. It has been argued that because the synthesis of substance P and upregulation of NK1 receptors are driven by nerve growth factor (NGF) acting at trkA receptors, there is a specific role for substance P in NGF-induced hyperalgesia. For example, it has been assumed that hyperalgesia observed in a transgenic NGFoverexpressing mouse is evidence for a role of substance P in nociception. The objection to the capsaicin experiment (above) can be raised in this case, in that many other substances contained in trkA-receptor-expressing nociceptive
0165-6147/00/$ – see front matter © 2000 Elsevier Science Ltd. All rights reserved.
PII: S0165-6147(00)01502-9
VIEWPOINT fibres could be implicated in addition to substance P. Indeed, there is a better correlation between the presence of CGRP in primary afferents and trkA receptors than there is between the presence of substance P and trkA receptors6. It is increasingly unlikely that any single peptide (or, for that matter, any other individual neurotransmitter) can be confidently associated with a specific sensory function4,7. Additionally, it is now accepted that sensory neurotransmission is plastic, that synthesis of transmitters and their receptors, including that of substance P, can be upregulated by conditions such as inflammation, which cause pain, and that synaptic connections often change after nerve injury8. We are thus faced with a nociception system that uses numerous transmitter substances in parallel and that adjusts its sensitivity up and down in response to tissue injury. In this respect, it is significant to note that when those lamina I dorsal horn neurones that carry the NK1 receptor are killed by intrathecal injection of substance P conjugated with the toxin saporin the antinociceptive effect observed is greater than that produced by NK1 receptor antagonists presumably because all input via these cells is removed9. Studies with knockout mice and selective receptor antagonists
The best way to reliably establish whether any one transmitter in this complex neurotransmission process has a pivotal role is to either block its receptor with a pharmacological antagonist or to remove the transmitter itself or the target receptor using transgenic gene deletion. Both approaches have now been used for NK1 (substance P) receptors and thus we have a body of objective data from which to draw conclusions. The phenotype of the knockout mice provides some, not totally convincing, evidence that NK1 receptor antagonists should be analgesic in humans. The preprotachykinin knockout mouse (which lacks both substance P and neurokinin A) has a substantially normal baseline nociception with attenuation of responses to intense noxious stimuli. The NK1 receptor knockout mice also exhibit no changes in acute nociception tests but display a reduction in responses to inflammatory stimuli and reduced wind up in dorsal horn neurones10. This is substantially the same picture that is observed in many of the published behavioural studies with the non-peptide, systemically active NK1 receptor antagonists that have been developed in the past decade. Antinociceptive effects were seen with a variety of NK1 receptor antagonists in the presence of nerve injury or inflammation but no unequivocal effects on nociception were observed in acute tests such as the hot plate2,11. In total, there were sufficient well conducted studies in which an NK1 receptor antagonist could be shown to attenuate the behavioural or electrophysiological response to a noxious stimulus to justify performing clinical trials for analgesia. After all, the profile of the compounds across the behavioural tests used was not dissimilar to that of nonsteroidal anti-inflammatory drugs (NSAIDs), which are known to be analgesic in humans. There was also enough negative evidence, however, to suggest that NK1 receptor antagonists were not going to be strong analgesics like morphine10,11. It is interesting to speculate whether the search for such compounds as analgesics would have been so widely
followed by the pharmaceutical industry if the knockout data had been available before the discovery of the non-peptide antagonists rather than ten years afterwards! Preclinical antinociceptive data in animals can be rationalized with the lack of effect in the clinic
It was a surprise to many of us when most of the clinical trials of NK1 receptor antagonists failed to show any analgesic effects whatsoever1 (but see Ref. 12 for a description of studies with CP99994 that showed some analgesia in dental pain). Obvious explanations for a lack of effect in humans are species differences in the physiology of substance P or differences between clinical pain and the type of noxious stimulus and response studied experimentally10. It is interesting to note that most of the differences in NK1 receptor distribution between species are expressed at supraspinal sites (and possibly the importance of these sites in pain perception has been underestimated), with remarkable similarities in the distribution of both the peptide and NK1 receptors in the dorsal horn of the spinal cord (the initial relay in the nociceptive process but not the level of sensation). In addition, the animal tests that showed antinociceptive activity with NK1 receptor antagonists have recently been reliably predictive of analgesic activity in humans with other novel strategies (e.g. in the case of cyclooygenase 2 inhibitors13). It is possible to reconcile the animal and clinical data if one includes the noxious stimuli used in the animal studies under the general heading of stressful stimuli. The published work on nociception then links up well with other animal experimental work on, for example, isolation-induced stress with NK1 receptor antagonists and NK1 receptor knockout mice2. This cumulative data set tells us that NK1 receptor antagonists can reliably attenuate the response to a stressful stimulus (e.g. whether produced by intraplantar formalin, nerve injury or by separation from a family group2). This effect is apparently not sufficient by itself to result in clinical analgesia. In support of this idea, the studies of Baulmann et al.14 on the response to peripheral injection of formalin as a noxious and stressful stimulus in rats are helpful. These workers showed that blockade of tachykinin receptors reduces the behavioural response to formalin and also reduces the induction of the immediate early gene Fos in brain areas linked to nociception and also in some of the brain areas that have been specifically associated with central stress responses. It has been noted that in veterinary practice it is difficult to distinguish the factors that might cause an animal distress. Pain is just one of these factors and clinically it might combine with other stressors to alter an animal’s behaviour15. This blurring of the distinction between pain, emotion and stress also extends to pharmacological treatment. Before the advent of effective drugs for treating depression, one accepted treatment, for what was then generically called melancholia, was opioid drugs such as morphine. A recent clinical trial in a small number of subjects showed relief of the symptoms of unipolar depression with the mu opioid peptide receptor partial agonist, buprenorphine, which until that time was only thought to be useful for the treatment of pain16. We are thus left with the working hypothesis that attenuating sensory neurotransmission, in a subtle way by interfering
TiPS – July 2000 (Vol. 21)
245
VIEWPOINT with the role of a single neuropeptide, might be adequate for reducing the behavioural response to a stressor such as inappropriate or intolerable sensory input. A more complete block of sensory input (such as can be produced by morphine) is needed to also achieve clinical pain relief. In terms of clinical benefit to individual patients it might be difficult to determine which factor is the more important. Loeser and Melzack17 wrote recently, in their introduction to a series of articles in the Lancet on pain, ‘Injury does not only produce pain, it also leads to stress […]. This process involves neural, hormonal, and behavioural activities.’ Concluding remarks
Acknowledgements I would like to thank all my colleagues, especially Nadia Rupniak, Sue Boyce, Mike Cumberbatch, Qing-Ping Ma, Jenny Laird, Jenny Longmore, Chris Swain and Richard Hargreaves for their data, discussions and ideas about the functions of substance P over the past ten years.
There is clearly still much to learn about the clinical effects of NK1 receptor antagonists. It is possible that there might be particular pain states in which the role of substance P is dominant and where NK1 receptor antagonists might be effective analgesics in humans, although I suspect such a situation will be infrequent. The possibility for using these drugs as adjuvants to existing analgesics, especially when there are other stress-related factors to consider, is a more attractive prospect. Additionally, it is relevant to ask whether the lack of ability to predict the presence or absence of analgesic properties in humans in the case of NK1 receptor antagonists has an impact on the discovery and clinical evaluation of other putative analgesics. What preclinical criteria should be used to determine whether clinical trials of a new analgesic are likely to be successful? It has recently been suggested that positron emission tomography (PET) studies in volunteers should be used to establish that adequate blocking doses are achieved in the brain, in humans before efficacy studies of a new analgesic are carried out18. It was suggested that the failure of a new inhibitor of enkephalin metabolism to exhibit analgesia in humans was a result of inadequate brain penetration even though it had shown antinociceptive effects in animal tests18. In fact, this criticism cannot be made, because some of the NK1 receptor antagonists that were found to be ineffective in humans had been subjected to PET studies and doses of drug that were adequate to saturate the brain receptors were used in the clinical efficacy studies (Ref. 2 and D. Burns et al., unpublished). Furthermore, similar doses of the same drugs were found to be effective in emesis studies, which are known to require both high levels of receptor occupancy and adequate brain penetration2. The final range of clinical uses of the NK1 receptor antagonists remains to be defined.
Selected references 1 Boyce, S. and Hill, R.G. (2000) Discrepant results from preclinical and clinical studies on the potential of substance P–receptor antagonist compounds as analgesics. In Proc 9th World Congress on Pain (Devor, M. et al., eds), pp. 313–324, IASP Press 2 Rupniak, N.M.J. and Kramer, M.S. (1999). Discovery of the antidepressant and anti-emetic efficacy of substance receptor (NK1) antagonists. Trends Pharmacol. Sci. 20, 485–490 3 Von Euler, U.S. (1985) The history of substance P. In Neurotransmitters in Action (Bousfield, D., ed.), pp. 143–150, Elsevier Biomedical Press 4 Salt, T.E. and Hill, R.G. (1983) Neurotransmitter candidates of somatosensory primary afferent fibres. Neuroscience 10, 1083–1103 5 Strittmatter, M. et al. (1997) Cerebrospinal fluid neuropeptides and monoaminergic transmitters in patients with trigeminal neuralgia. Headache 37, 211–217 6 Averill, S. et al. (1995) Immunocytochemical localization of trkA receptors in chemically identified subgroups of adult rat sensory neurons. Eur. J. Neurosci. 7, 1484–1494 7 Levine, J.D. et al. (1993) Peptides and the primary afferent nociceptor. J. Neurosci. 13, 2273–2286 8 Woolf, C.J. and Mannion, R.J. (1999) Neuropathic pain: aetiology, symptoms, mechanisms and management. Lancet 353, 1959–1964 9 Mantyh, P.W. et al. (1997) Inhibition of hyperalgesia by ablation of lamina-I spinal neurons expressing the substance P receptor. Science 278, 275–279 10 Ma, Q-P. and Hill, R.G. (1999). Neurokinin antagonists as potential agents for use in pain management. Curr. Opin. Invest. Drugs 1, 65–71 11 Rupniak, N.M.J. and Hill, R.G. (1999). Neurokinin antagonists. In Novel Aspects of Pain Management: Opioids and Beyond (Sawynok, J. and Cowan, A., eds), pp. 135–155, Wiley-Liss 12 Dionne, R.A. (1999) Clinical analgesic trials of NK1 antagonists. Curr. Opin. Invest. Drugs 1, 82–85 13 Chan, C.C. et al. (1999) Rofecoxib [Vioxx, MK-0966, 4-(49methylsulfonylphenyl)-3-phenyl-2-(5H )-furanone]: a potent and orally active cyclo-oxygenase2 inhibitor. Pharmacological and biochemical profiles. J. Pharmacol. Exp. Ther. 290, 551–560 14 Baulmann, J. et al. (2000) Tachykinin receptor inhibition and c-fos expression in the rat brain following formalin induced pain. Neuroscience 95, 813–820 15 Headley, P.M. and Livingston, A. (1989) Pain and stress in animals: problems of assessment and treatment. Front. Pain 1, 1–4 16 Bodkin, J.A. et al. (1995) Buprenorphine treatment of refractory depression. J. Clin. Psychopharmacol. 15, 49–57 17 Loeser, J.D. and Melzack, R. (1999) Pain: an overview. Lancet 353, 1607–1609 18 Langley, G. et al. (2000) Volunteer studies replacing animal experiments in brain research. Alternatives to Laboratory Animals 28, 315–331
Chemical name CP99994: (2S 3S)cis-3(2-methoxybenzylamino)-2-phenyl piperidine
Trends in Pharmacological Sciences : a forum for comment Controversial? Thought-provoking? If you have any thoughts or comments on articles published in Trends in Pharmacological Sciences, then why not write a Letter to the Editor at [email protected]? Letters should be no more than 500 words. Please state clearly whether you wish the letter to be considered for publication. The letter will be sent to the author of the original article for their response; both the letter and reply will be published together. Please note: submission does not guarantee publication.
246
TiPS – July 2000 (Vol. 21)