Nerve growth factor and chronic daily headache: a potential implication for therapy

Transcription

1 Expert Review of Neurotherapeutics ISSN: (Print) (Online) Journal homepage: Nerve growth factor and chronic daily headache: a potential implication for therapy Paola Sarchielli & Virgilio Gallai To cite this article: Paola Sarchielli & Virgilio Gallai (2004) Nerve growth factor and chronic daily headache: a potential implication for therapy, Expert Review of Neurotherapeutics, 4:1, , DOI: / To link to this article: Published online: 10 Jan Submit your article to this journal Article views: 100 View related articles Citing articles: 16 View citing articles Full Terms & Conditions of access and use can be found at

2 Review For reprint orders, please contact CONTENTS Peripheral & central sensitization & CDH Neurotransmitters & intracellular transduction mechanisms involved in nociception Evidence of neurotransmitter alterations in CDH NGF levels in CDH Expert opinion Five-year view Key issues References Affiliations Nerve growth factor and chronic daily headache: a potential implication for therapy Paola Sarchielli and Virgilio Gallai The pivotal role of nerve growth factor in inducing hyperalgesia and central sensitization has been emphasized in experimental pain models. Higher nerve growth factor levels have recently been found in the cerebrospinal fluid of patients with chronic daily headache. These levels were significantly correlated with the cerebrospinal fluid levels of substance P and calcitonin gene-related peptide, supporting the involvement of this neurotrophin in enhancing the production of the two sensory neuropeptides of the trigemino-vascular system in chronic daily headache. This may, in part, account for the long-lasting sensitization and activation of this system, which could contribute to headache chronicity. More recent research has shown a significant correlation between the higher cerebrospinal fluid levels of nerve growth factor and those of another neurotrophin, the brain-derived neurotrophic factor, as well as glutamate in chronic daily headache patients. These findings suggest the potential involvement of nerve growth factor-mediated upregulation of brain-derived neurotrophic factor in persistent head pain. Therefore, nerve growth factor appears to indirectly exert its effect through enhancing glutamatergic transmission involved in the processing of head pain via brain-derived neurotrophic factor. Based on these data, a potential application can be hypothesized for novel strategies targeting neurotrophins (nerve growth factor and brain-derived neurotrophic factor) and their receptors to chronic daily headache. To date, the majority of the molecules discovered in this regard have been scarcely or never proved in animal pain models and are far from clinical use in chronic pain, including chronic daily headache. If this approach is to be developed in the near future, research should be focused on identifying strategies with few central side effects and specific selective action on central sites involved in chronic head pain and more generally in chronic pain conditions. This will represent a very difficult challenge, taking into account the pleiotropic effect of nerve growth factor and the wide range of intracellular signalling pathways activated by this neurotrophin which are not limited to the nociceptive system. Author for correspondence Department of Neuroscience, Neurologic Clinic, Via E Dal Pozzo 06126, Perugia, Italy Tel.: Fax: neuro.pg@tiscalinet.it KEYWORDS: brain-derived neurotrophic factor, central sensitization, chronic daily headache, clinical application, glutamate, nerve growth factor, therapy targeting NGF Expert Rev. Neurotherapeutics 4(1), (2004) Chronic daily headache (CDH) affects approximately 4 5% of the general population and is one of the more frequently encountered syndromes at tertiary headache centers [1]. This term encompasses a number of different primary headache disorders presenting daily, or almost daily and lasting more than 4 h per day over at least 3 months [2]. The diagnosis of these forms is still a matter of debate and the current International Headache Society (IHS) criteria do not allow adequate classification [3 5]. Several epidemiological studies were carried out with the aim of identifying the clinical features of CDH [6 8]. Based upon the variability of the clinical characteristics, some authors proposed a revision of the current criteria, emphasizing the need to categorize the most frequent subgroup disorders, chronic tension-type headache and the so-called transformed migraine, not only on the basis of their prevailing clinical characteristics but also by taking into account their evolution [9 11]. Future Drugs Ltd. All rights reserved. ISSN

3 Sarchielli & Gallai According to this diagnostic approach, other less frequent forms are new-onset persistent headache and hemicrania continua. Most CDH patients overuse symptomatic medication, particularly analgesics, ergotamine (Cafergot, Novartis Pharmaceuticals Corp., Basel, Switzerland) and more recently, triptans. In this case the terms of rebound headache or drug-induced headache were currently used [8,12,13]. CDH patients often respond to drug withdrawal with a significant reduction of headache frequency, intensity and duration. Other authors suggest classifying a drug-induced headache on the basis of the amount of drug consumption and without mandatory demands for withdrawal [14,15]. The new classification proposal reconfirms chronic tension-type headache as a separate entity and introduces the concept of chronic migraine as a complication of episodic migraine. This entity does not include chronic migraine with drug abuse or overuse [201]. Based on the new IHS classification proposal, if drug overuse is present in the majority of cases, migraine chronicity is caused by this overuse. The default rule is therefore to codify such patients according to the primary headache diagnosis and as headache attributed to chronic drug overuse as second diagnosis. If chronicity persists after drug withdrawal, chronic migraine should then replace headache attributed to drug overuse. The biochemical basis of CDH is poorly understood and few studies have been performed in this field. Several hypotheses have been formulated on the basis of experimental pain models and less consistently from data obtained from humans affected but none of them is completely exhaustive. The dominating concept which attempts to explain the development of a daily headache from a paroxysmal migrainous or tension-type headache pattern, is that the so-called headache transformation recognizes a common pathogenetic substrate in both forms. Peripheral & central sensitization & CDH Different lines of evidence suggest that chronic head pain (evolving from episodic migraine and tension-type headache) is the consequence of central sensitization following peripheral sensitization and/or an impaired descending inhibitory or increased facilitatory control on upper spinal dorsal horn and trigeminal nucleus caudalis neurons by supraspinal sites [16]. Sensitization of peripheral nociceptors (peripheral sensitization) is the first event which leads to headache chronicity [17]. It is well known that noxious stimuli in the periphery induce the release of algogenic or inflammatory substances from various sources, including nociceptors and this so-called inflammatory soup induces changes in nociceptors, including phosphorylation of target membrane proteins, particularly tetrodotoxin-resistant (TTX) sodium and calcium channels [18]. Sensitized nociceptors become more sensitive to suprathreshold stimuli, causing an exaggerated pain response to noxious stimuli (hyperalgesia) and may respond to low threshold non-noxious stimuli (allodynia) [20]. Central nociceptive neurons, both in the spinal dorsal horn and trigeminal nucleus caudalis, undergo a series of physiological and molecular events, which enhance their sensitivity and lead to central sensitization and the expansion of nociceptive fields. Repeated stimulation of primary C fibers is responsible for the long-lasting changes in synaptic potentials of nociceptive neurons, which over time can be summated [20]. Synaptic summation underlies the so-called phenomenon of wind-up, a nonlinear and activitydependent response consisting of an increased responsiveness of nociceptive neurons after repeated activation of primary afferents [21]. This phenomenon can involve both upper spinal dorsal horn and trigeminal nucleus caudalis neurons. The maintenance of molecular and functional changes in the central neurons may contribute to the development of central sensitization characterized by increased sensitivity to suprathreshold stimuli and responsiveness to low intensity subthreshold stimuli. The mechanism of central sensitization is believed to underlie chronic pain in general and also chronic head pain. Chronic migraine Burstein and colleagues tested and confirmed the hypothesis that sensitization of peripheral nociceptors that innervate the intracranial blood vessels and meninges is responsible for the pulsating/throbbing head pain during a migraine attack and the increased intracranial sensitivity associated with migraine attacks [22]. This results in the aggravation of pain by mechanical stimuli, such as those produced by the small increases in intracranial pressure due to coughing or bending. The same author demonstrated that sensitization of second-order nucleus caudalis neurons receiving convergent intracranial input from cerebral blood vessel meninges and extracranial input from meningeal nociceptors during a migraine attack can induce cutaneous allodynia within the referred pain area (V1, periorbital region), as well as outside the referred pain area (limbs). This finding adds further evidence to the concept of central sensitization as the basis of chronic migraine [23]. In the case of persistent activation, hyperexcitable peripheral nociceptors are spontaneously activated, even in the absence of peripheral stimuli. The impulses that they generate will travel orthodromically, reaching second-order nociceptive neurons in the upper dorsal horn of the spinal cord and trigeminal nucleus caudalis and this continuous nociceptive input on the second-order neurons can contribute to maintaining their long-lasting state of hyperexcitability. They are chronically and spontaneously activated and also become responsive to subthreshold stimuli. This body of evidence, therefore, supports the possibility that frequent migraine attacks and in particular chronic migraine, may be the consequence of a tonic state of increased excitability and neuroplastic changes within upper spinal horn neurons and the trigeminal nucleus caudalis, and perhaps supraspinal structures involved in head pain processing that results in spontaneous pain and a reduced threshold for activation. A model that has been extrapolated to chronic migraine and is strictly related to that of sensitization is the neurophysiological phenomenon named kindling [24]. According to this model, repeated and persistent crises induce a modification in 116 Expert Rev. Neurotherapeutics 4(1), (2004)

4 Nerve growth factor and chronic daily headache the excitability of head nociceptive pathways. Their reiterative and spontaneous firing, even in the absence of adequate trigger stimuli, may be responsible for chronic migraine. Chronic tension-type headache In chronic tension-type headache (CTTH) patients, the abnormal pericranial tenderness has been attributed to the increased sensitivity to persistent nociceptive stimuli from pericranial myofascial tissues [25]. The lower pressure pain detection and tolerance thresholds is also suggestive of a condition of hyperalgesia and allodynia in these patients [26]. They have also been found to be hypersensitive, not only to pressure stimuli but also to thermal or electrical stimuli, both at cephalic (symptomatic) and extracephalic locations, the latter ascribed to a widespread, nonspecific facilitation of spinal neurons by supraspinal structures [27]. Neuroplastic changes due to central sensitization in CTTH patients can increase the drive to motor neurons, both at the supraspinal and segmental levels, resulting in the increased electrical activity and hardness of pericranial muscles [28]. Therefore, both in migraine and tension-type headache, the mechanism of central sensitization may be maintained, even after the initial eliciting factors have been normalized and in both cases, individuals will experience CDH. Supraspinal modulation of nociceptive inputs Other potential mechanisms contributing to the maintenance of central sensitization may include impaired supraspinal inhibition, or increased supraspinal facilitation of nociceptive inputs by supraspinal structures [29]. Functional and/or structural changes, recently demonstrated in one of the most relevant structures of the antinociceptive network, the periacqueductal gray (PAG) matter, may render trigeminal nociceptive pathways more prone to paroxysmal, even frequent activation (migraine attacks) or sensitization (CDH) in susceptible subjects [30,31]. Chronic analgesic abuse, often encountered in CDH, appears to strongly influence the antinociceptive serotonergic pathways, whose neurons are located in the midbrain medial nuclei. A decrease in serotonin content and changes in receptor and transporter regulation following chronic exposure to symptomatic drugs may be responsible for the lack of their analgesic efficacy and contribute to headache chronicity [32]. Neurotransmitters & intracellular transduction mechanisms involved in nociception: putative targets for the treatment of chronic pain, including CDH Several lines of experimental evidence support the involvement of sensory peptides, substance P (SP) and calcitonin gene-related peptide (CGRP), as well as excitatory amino acids (in particular glutamate), but also other neuroactive substances, such as galanin and somatostatin, in the peripheral and central mechanisms underlying sensitization in painful conditions. The majority of data have been obtained in animal models [34]. Substance P, a neurokinin which acts principally on natural killer cell (NK) 1 receptors and CGRP, acting on its specific receptors, are released from nociceptive afferents during repeated stimulation [34]. These peptides elicit long-lasting changes in synaptic potentials, which can be summated over time, thus contributing to the development of the wind-up phenomenon [35]. A variety of NK 1 receptor antagonists have been synthesized but not all investigators reported significant antihyperalgesic effects of these compounds in animal models of pain [36]. Their usefulness was also denied in migraine patients during attacks [37]. Another proposed approach is to use CGRP antagonists and more promising results were obtained in this regard in migraine and could be of potential application in CDH [38]. Several experimental findings suggest an interaction between nociceptive neurons producing and releasing SP and CGRP and glutamate receptors, in particular N-methyl-D-aspartate (NMDA) receptors. The latter have been localized to the presynaptic terminals of small diameter nociceptors in the spinal cord and trigeminal nucleus caudalis, where they are believed to facilitate and prolong nociceptive input by increasing the release of both sensory neuropeptides [39]. The pivotal role of the phosphorylation of NMDA glutamate receptors has clearly been established in the development and maintenance of hyperalgesia and central sensitization in animal models [40]. The molecular mechanisms following NMDA receptor activation by glutamate involve calcium entry into the dorsal horn neurons and neurons of the trigeminal nucleus caudalis, which induces the activation of protein kinases II, A and C [41]. Neuronal proteins, which are substrates for these kinase receptors, include neurotransmitter receptors, ion channels and transcriptional factors. The results of these intracellular events are an increased membrane hyperexcitability and genetic transcription, including the expression of early genes [42]. The use of NMDA receptor antagonists has been reported to be effective in experimentally-induced mechanical and thermal skin hyperalgesia, in models of joint and visceral inflammation and neuropathic pain [43]. Among NMDA antagonists, noncompetitive NMDA receptor channel antagonists, such as ketamine and magnesium chloride, seem to be less effective than competitive NMDA receptor antagonists in preclinical animal studies [44]. All of the available data suggest that noncompetitive NMDA receptor antagonists could be advanced as antihyperalgesic but not analgesic compounds. The latter property was instead attributed to the competitive NMDA receptors [45]. Besides the above considerations and based upon the assumption that CDH, in particular, is associated with analgesic drug abuse, due to an increased glutamatergic transmission, Nicolodi and colleagues suggested the potential usefulness of the noncompetitive NMDA receptor ketamine or of the inhibitor of excitatory amino acid release gabapentin (Neurontin, Pfizer Inc., NY, USA), to overcome analgesic drug abuse without any physical abstinence signs [46]. Likewise, Fusco and colleagues demonstrated that dextromethorphan (Touro DM, PharmaFab, TX, USA), a weak NMDA antagonist, reduced the increased temporal summation of secondary pain in patients with transformed migraine and this effect was further augmented by magnesium administration [48]

5 Sarchielli & Gallai NonNMDA receptor antagonists are currently available and they have been documented to attenuate hyperalgesia in some experimental pain models and in human volunteers [49]. The above evidence, together with the observation of fewer side effects, suggests the development of novel drugs with nonnmda receptor antagonistic activity, which would lead to useful antihyperalgesic compounds for chronic pain, including CDH. Based on the observations that different genes encoded subunits variably assembling to form ion channels for NMDA and nonnmda receptors, the development of novel, highly selective drugs acting specifically on the different glutamatergic receptor subunits has been hypothesized for a potential application in humans. Calcium entry into nociceptive neurons in the dorsal horns of the spinal cord due to glutamate receptor activation induces the activation of the neuronal nitric oxide (NO ) synthase (NOS) enzyme responsible for the synthesis of NO and the activation of soluble guanylate cyclase, which results in the increase in the soluble intracellular messenger cyclic guanylate monophosphate (cgmp) [49]. NO diffuses back to presynaptic terminals and enhances neurotransmitter release, therefore contributing to the expression of long-term potentiation and long-term facilitation of synaptic flow [50]. Due to its crucial role here, NO could therefore be another target of a therapeutic approach for controlling chronic pain, including CDH. Based on the findings obtained in experimental models of pain and on the demonstration that NOS inhibition reduces central sensitization in models of chronic pain, a nonselective NOS inhibitor, N(G) monomethyl-l-arginine (L-NMMA) was used in CTTH patients and shown to significantly reduce pain scores compared with placebo [51]. The same NOS inhibitor was also able to significantly reduce the hardness and tenderness of pericranial muscles, reflecting the sensitization of second-order neurons due to prolonged myofascial input and together these data support its potential role as an analgesic in these patients [52]. The effectiveness of NOS inhibitors for the treatment of CDH evolving from a previous history of migraine remains to be established. The use of selective inhibitors of neural, endothelial or inducible NOS may also be helpful in discriminating the intervention of the different enzyme isoforms in peripheral and central sensitization in this pathological condition. Evidence of neurotransmitter alterations in CDH Few studies have been carried out to define the biochemical alterations underlying chronic head pain in CDH patients and the majority of the available data concern CDH evolving from a previous history of migraine, the so-called transformed migraine, according to the definition of Silberstein and colleagues [10]. A functional impairment of the serotonergic pathway and intracellular transduction mechanisms emerged in the platelet model in these patients [53,54]. It consisted of an upregulation of the 5-HT 2A receptors, a decreased content of serotonin and dilatation of the canalicular system implying an excessive release of 5-HT and therefore, resulting in a hyposerotonergic state, particularly evident in the case of analgesic abuse [54,56]. The impairment of antinociceptive modulatory pain pathways was also supported by the finding of reduced β-endorphin levels in CDH and increased levels of met-enkephalins in the CSF of CTTH [56,57]. The authors group showed an increased platelet NOS activity in patients with CDH evolving from a previous history of migraine without aura, as evidenced by the high values of NO and cgmp produced, the increased basal and collagen-stimulated cytosolic calcium and the decreased serotonin secretion and content [58]. These alterations were particularly evident in patients with analgesic abuse. A similar pattern of NO hyperproduction was more recently demonstrated in patients with tension-type headache, particularly in those with analgesic abuse [59]. A recent study carried out by the authors group in this regard demonstrated high levels of the two excitatory amino acids glutamate and aspartate in the CSF of CDH patients, which were significantly correlated with the values of the nitrites [60]. These findings support the hypothesis of the involvement of the glutamatergic system activation in CDH patients, which can be responsible for the activation and increase in neuronal NOS and could explain the increase in the stable end products of nitric oxide metabolism. Other mechanisms have been recently advocated to explain central sensitization in patients with CDH, such as those mediated by neurotrophins. In particular, experimental evidence suggests the involvement of nerve growth factor (NGF) in acute hyperalgesia and the maintenance of chronic pain, this prompted us in recent years to investigate its role in CDH. NGF & nociception Neurotrophins (NTs) are a group of structurally-related proteins, which include NGF, brain-derived neurotrophic factor (BDNF), NT-3, -4/5, whose action is mediated by specific receptors [61]. All neurotrophins bind to a low-affinity receptor called p75 [62]. However, it has been discovered that NTs can selectively bind to high-affinity receptors, which have now been identified as tyrosine kinases, that is, TrkA, B and C. NGF binds selectively to TrkA, BDNF and NT-5 to TrkB and NT-3 and NT-4 to TrkC. NGF displays a high affinity for the TrkA receptor, BDNF and NT-4/5 demonstrate a high affinity for the TrkB receptor, whereas NT-3 specifically acts on the TrkC receptor [63]. It has been clearly defined that NGF is involved in the embryonic development of sensory neurons and it has also been demonstrated that nociceptors need this neurotrophin to maintain their phenotype in the postnatal period [64]. Trk receptors moreover, continue to be expressed in some populations of sensory and particularly nociceptive neurons [65] and synthesize the neurotrophins after embryonic development [66]. These observations lead several authors to clarify the intervention of NGF and other neurotrophins in nociception in adulthood. NGF & peripheral sensitization In addition to being a survival factor during the development of sympathetic and sensory neurons, NGF has been 118 Expert Rev. Neurotherapeutics 4(1), (2004)

6 Nerve growth factor and chronic daily headache demonstrated to interact with pain-signalling systems in adult animals, where it regulates the specification of the nociceptive phenotype [64,65,67]. NGF is released by several cell types in response to tissue inflammation and is responsible for hyperalgesia when administered either locally or systemically in many species [68 71]. NGF-induced thermal hyperalgesia occurs within an hour after its administration and both this short latency and the local effect of NGF suggest the involvement of a prevalent peripheral mechanism. NGF-induced mechanical hyperalgesia occurs, in contrast, with a latency of some hours, suggesting more complex central mechanisms [72]. The role played by NGF in peripheral nociception is supported by the evidence of the upregulation and increased delivery of this neurotrophin and overexpression of high-affinity TrkA receptors in the nociceptive terminals in the injured areas after inflammatory insults [73]. The hyperalgesic action of NGF in inflammatory pain models may in part be the consequence of the increased sensitivity of peripheral terminals of high threshold nociceptors, either through a direct action of NGF on TrkA-expressing sensory fibers or via the release of sensitizing mediators from postganglionic sympathetic neurons and TrkA-expressing cells [74,75]. The TrkA receptor seems to mediate acute hyperalgesia in A δ and C fibers via the enhancement of SP and CGRP release [77]. NGF is also retrogradely transported in sensory neurons to the dorsal root ganglia, where it alters the transcription of several proteins and peptides, including that of sensory neuropeptides [77]. Both in vivo and in cultured sensory neurons, the exogenous administration of NGF induces, in fact, an overexpression of the sensory peptides SP and CGRP in cell bodies [78,79]. This overexpression is a slow process, requiring from hours to days to manifest itself, confirming the involvement of transcriptional mechanisms not related to its rapid effect on the sensory threshold [80]. Several lines of evidence support the role of mast cells in NGF-induced hyperalgesia. They display TrkA receptors on their membranes and may be degranulated by NGF, which functions as a chemoattractant for these cells through both mitogen-activated protein kinase and phosphatidyl inositol 3-kinase signalling pathways [67,80,82]. Mast cell involvement in neurogenic inflammation, which has been used as an experimental model of acute migraine attack is well known but whether its activation may be related to NGF release by these cells remains to be established [83]. Recent experimental data suggest that NGF activates mast cells through the collaborative interaction with lysophosphatidyl serine expressed on the membrane of activated platelets and this could be of further relevance for the model of neurogenic inflammation where the occurrence of activated platelets has been demonstrated [85]. Pretreatment with compound 48/80 inducing mast cell degranulation blocks short latency hyperalgesia due to NGF but does not modify its long latency effects [85]. Mast cells contain histamine, serotonin and NGF, which can contribute to the activation of peripheral nociceptors. In particular, serotonin has been demonstrated to sensitize polymodal nociceptors to the painful stimulus, and 5-HT 2 but not 5-HT 1 antagonists have been shown to block NGF-induced hyperalgesia [86 88]. NGF seems to be necessary in the model of hyperalgesia induced by Freund adjuvant, which is abolished by antingf antibodies or TrkA-IgG immunoadhesion [90,91]. The release of proinflammatory cytokines, such as tumor necrosis factor (TNF)-α and interleukin (IL)-1β, is responsible for the release of NGF by keratinocytes and fibroblasts, which then promote mast cell degranulation [91]. The capsaicin-induced pain model has also been used to investigate the role of NGF in peripheral sensitization. Capsaicin directly applied on primary sensory fibers induces their depolarization by acting on its specific receptors VR1 and increasing their permeability to calcium [92]. It has been demonstrated that NGF acutely sensitizes the response of sensory neurons to capsaicin in adult rats and exerts this effect by influencing the expression of VR1 receptors and inducing their phosphorylation [93,94]. Further studies have shown that NGF binding to the low-affinity neurotrophin receptor p75 activates the sphingomyelin signalling pathway causing the liberation of the second messenger molecule ceramide [95]. This additional NGF-mediated effect seems to be involved in the sensitization of sensory neurons consequent to noxious inflammatory stimulation. Ceramide exerts this effect by modulating the TTX-resistant sodium current in sensory neurons in the animal model [96]. Future studies exploring the role of the sphingomyelin signalling pathway in relation to NGF are needed for understanding its relevance in experimental models of acute and chronic pain. NGF & central sensitization Experimental data show that NGF is able to induce a pain-like response when infused into the CSF in rats, suggesting a central mechanism of action in addition to the above evidence of peripheral effects [97]. Moreover, NGF has been demonstrated not only to determine acute hyperalgesia but also to maintain chronic pain. Other than to immediately reduce the pain threshold, this neurotrophin contributes to the development of mechanical allodynia occurring 8 12 h later and to the secondary pain response [70]. NGF could provide a prolonged pain sensation through the potential redistribution of the sodium channels and through the production of PN1 sodium channels in the small neurons of the dorsal root ganglia [89,99]. By this mechanism and via direct activation of TrkA receptors, as well as by increasing the production of sensory neuropeptides in nociceptive neurons, NGF contributes to central sensitization [99,100]. The contribution of each of the above mechanisms of action has not been defined in the trigeminal nucleus caudalis and upper spinal dorsal horn neurons. In experimental animals, trigeminal nociceptors differ, in fact, in the expression of the TrkA receptor from dorsal root ganglia (DRG) nociceptors [101]. In particular, 10 16% of CGRP+ trigeminal neurons express TrkA receptors versus 40% found in DRG neurons but this different distribution cannot be extrapolated 119

7 Sarchielli & Gallai to humans and does not exclude an NGF high-affinity receptor upregulation in the condition of central sensitization underlying chronic head pain [102]. Signalling due to NGF involves intracellular responses through the small G-protein Ras and MAPK intervening in the regulation of gene expression through phosphorylation, the activation of various transcription factors (c-myc, elk-1, c-fos, c-jun) and of many downstream genes via the association with response elements in the gene promoter response regions [80]. This may be responsible for the transcriptional and phenotypical changes in nociceptive neurons of the trigeminal system. The evidence available in the literature until now, supports a prevailingly peripheral role of NGF in nociception, whereas a pivotal central role is attributed to another neurotrophin, BDNF. On the contrary, the effect of the latter neurotrophin in enhancing NMDA evoked responses seems to be influenced by NGF [103]. BDNF is synthesized by TrkA-positive sensory neurons, acts through TrkB receptors and has been demonstrated to be produced in experimental models of hyperalgesia [104]. Its role in sustained nociception is supported by the finding of the blockade of inflammation-induced changes in sensory processing pathways by TrkB-IgG fusion protein administration sequestering BDNF [105]. Experimental data suggest that the synthesis of BDNF appears to be enhanced by NGF [106]. NGF treatment, dramatically increases BDNF levels in TrkA nociceptive neurons where this neurotrophin is anterogradely transported, localized in dense vesicles and presumably released, at least at the spinal level [107,108]. The targets of TrkB-mediated modulation of central sensitization by BDNF are glutamate NMDA receptors, which are strongly recruited in activated nociceptive pathways and mediate much of the polysynaptic C-fibers-evoked discharge by BDNF [103]. The TrkB activation by this neurotrophin leads to the phosphorylation of NMDA receptor subunits 1 and 2B and increases the open probability of these channels [109]. NGF levels in CDH All data obtained in experimental animal pain models support the role of NGF as a putative candidate intervening in the pathogenesis of chronic pain, including CDH. Only a few studies have been performed to establish its role in maintaining pain states in humans. A study in this field demonstrated the increase of NGF in the CSF of patients with fibromyalgia [110]. Research was recently carried out by the authors group to investigate NGF levels in the CSF of patients suffering from CDH and to compare them with the values of this neurotrophin in the CSF of an age-matched control group, in which CSF and blood examinations, as well as adequate instrumental investigations excluded CNS or systemic diseases [111]. Values of SP and CGRP were also determined in the CSF of CDH patients and controls and correlated with NGF levels in the CSF. Significantly higher levels of NGF and sensory neuropeptides SP and CGRP in the CSF of CDH patients were found without significant differences between those with simple and combination analgesic abuse and those without. A significantly positive correlation emerged between NGF values in the CSF and the duration of chronic headache and number of days with headache per month, but not with visual analogical scale (VAS) values for pain intensity. NGF and SP, as well as CGRP values in the CSF of CDH patients were also significantly correlated. A similar significantly positive correlation emerged between CGRP and NGF levels in the CSF of the CDH patient group. The CSF levels of the two sensory neuropeptides, like those of NGF, were significantly correlated with the duration of chronic headache and number of days with headache per month. This can reflect the association found between the values of NGF and those of SP and CGRP in the CSF of patients with CDH. Another study recently performed by the author s group focused on the relationship between NGF and BDNF, and glutamate in the CSF of CDH patients [112]. In this study, patients with CDH showed significantly higher NGF and BDNF levels compared with control subjects. Even in this research, the CSF levels of NGF and to a lesser extent BDNF, appeared to be significantly correlated with the duration of chronic headache and the number of days with headache per month but not with VAS values for pain intensity. Moreover, a significantly positive correlation emerged between NGF and BDNF values in the CSF of CDH patients but not in control subjects. As in the author s previous study involving NGF, the abuse of simple or combination analgesics did not seem to influence the neurotrophin levels, which appeared to be more strictly related to the chronic pain status, rather than to symptomatic drug misuse. In this latter research, significantly higher values of glutamate were also found in the CSF of CDH patients when compared with control subjects, and no significant difference emerged in these levels between CDH patients with analgesic abuse and those without. Moreover, there was a significant correlation between the levels of NGF and BDNF in the CSF of CDH patients but this was not evident in controls. Expert opinion The reviewed data support the involvement of both NGF and BDNF in chronic head pain evolving from a previous history of migraine. Since NGF has been demonstrated to be involved in acute hyperalgesia, it can be hypothesized that an increase in its levels could occur during migraine attacks and that its levels tended to reverse and return to basal levels interictally. It can be hypothesized that higher levels are maintained in patients with CDH, particularly in those with a CDH evolving from a previous history of migraine attacks and therefore in those who can be classified as affected by transformed migraine, as seen in our study. It cannot be excluded that NGF, and perhaps BDNF upregulation, could also be present in CTTH, whose pathogenetic mechanisms have been related to a central sensitization. Further research should be carried out in this regard. Other neurotrophins, such as NT-3 and -4/5, have been recognized to cause hyperalgesia and also act as intermediates in nociception and could 120 Expert Rev. Neurotherapeutics 4(1), (2004)

8 Nerve growth factor and chronic daily headache contribute to both peripheral and central sensitization in CDH [ ]. They should be investigated in future research involving CDH patients. The effects of therapeutic approaches, especially prophylactic interventions on NGF and BDNF levels evidenced in the authors study of CDH patients, should be investigated in future research. Moreover, novel strategies targeting neurotrophins (NGF and BDNF) and their receptors, recently proposed for the treatment of chronic pain, should also be useful in CDH patients. Among these strategies, blocking the effect of NGF using neutralizing antingf antibodies or tyrosine kinase (Trk)A-Ig fusion proteins has been suggested as a new therapeutic approach with a specific action on chronic pain states, based upon the results obtained in animal pain models [119,120]. In the reviewed models, the effects of neutralizing antingf have been reported in several pain states. These include the attenuation of mechanical and thermal hyperalgesia following spinal cord injury or chronic sciatic nerve constriction, and at the molecular level, the decrease in CGRP densities in laminae I to IV, as well as the reversal of inflammation-induced SP and CGRP upregulation, early c-fos gene expression in the dorsal horn neurons, and reduction in the intensity of sodium channel labelling [99,122,123]. Furthermore, the administration of TrkA-Ig fusion proteins sequestering NGF was shown to produce thermal and mechanical hyperalgesia in the carrageenan inflammation model [121]. In addition, both approaches have very limited therapeutic potential in humans. Some antingf antibodies demonstrated a cross-reactivity to other neurotrophins; on the contrary, the supply of TrkA-Ig fusion protein and any recombinant protein may require a complex manufacturing and purification process [119]. Moreover, the problem of delivery only to specific CNS sites has not been solved. Antagonizing NGF binding and biological function is another approach which results in the inhibition of TrkA phosphorylation by NGF, a key initial event in the signalling transduction pathways mediated by this receptor. Compound PD90780 and kynurenic acid derivatives have been demonstrated to prevent the binding of NGF to low-affinity receptors p75 but their use has not been assessed in animal pain models [124,125]. ALE-0560 molecule has also been demonstrated to block NGF binding and NGF-induced neurite outgrowth [126]. This first nonpeptidergic NGF receptor antagonist molecule lacks both interaction with known analgesic targets including α1, H1, endothelin A, 5-HT 2, cannabinoids and opioid (µ, δ and k) receptors and central side effects. Although one suggested mode of action of this antagonist is the blockade of NGF-dependent sprouting of sensory and sympathetic axons coincident with behavioral signs of neuropathic pain, other potential targets, and therefore, its potential effectiveness in chronic pain including chronic head pain, should not be excluded and should be investigated in future research. In recent years, different neurotrophin mimetics with antagonistic action have been produced with potential clinical applications. In particular, small monomeric cyclic analogs that mimic the β-turn regions of NGF were designed and synthesized. Among them, potent competitive antagonists were derived from the NGF β-turn C D, which inhibits NGF binding to TrkA receptors and neurite outgrowth in PC12 cells [128]. These analogs were used to study the biological and receptor binding properties of NGF but they have never been tested in animal models of pain. Among NGF antagonists, a small peptide named C (92 96) that blocks NGF TrkA interaction was recently tested on cortex cholinergic synapses and has been demonstrated to induce a significant decrease in the size of vesicular acetylcholine transporter-immunoreactive (IR) sites [128]. The use of this approach for modulating and reducing molecular and neuronal events underlying chronic pain can be hypothesized from a theoretical point of view but raises the question of the dependency on NGF for the correct functioning of the CNS and the potential impact of this treatment on the maintenance of synaptic contacts in the adult CNS, which may not be limited to cholinergic neurons. It has been well-established that even in the mature CNS, exogenously applied NGF is capable of generating new cortical synapses and that TrkA receptors are able to modulate the neural phenotype in the adult CNS [ ]. Moreover, there is strong experimental evidence of an activity-dependent synthesis of neurotrophins in the adult CNS, acting on neurotrophin-sensitive nerve terminals impinging on NGF-secreting neurons. The result of this neuron neuron communication is the reinforcement of the synaptic connections, which could be involved in many physiological (i.e., increase in cortical synapses due to sensory stimulation) or pathological conditions (amplification of the molecular mechanisms in the nociceptive pathways in chronic pain). Based upon these observations, the idea of neutralizing NGF or blocking TrkA receptors could result in the withdrawal or loss of pre-existing synaptic contacts. The chronic administration of drugs targeting NGF or TrkA receptors could have as a consequence, therefore, a deleterious effect on several CNS functions, in particular, cognitive. Another approach which was developed to study the NGF TrkA interaction included the functional blockade of tyrosine kinase A by monoclonal antibodies against this receptor [133]. This was tested in the rat basal forebrain, but to our knowledge, it was not proven in experimental animal pain models [134]. In addition, the potential role of a recently-developed antitrka monoclonal antibody has been emphasized for a future therapeutic application in the painful condition [134], although the potential widespread effects of the receptor blocking intracellular signalling should be carefully evaluated before its use in humans. There is a great interest in new targets for pain therapy among intracellular signalling pathways. The activation of protein kinase C (PKC) could be one of these potential targets even in NGF-induced hyperalgesia. A PKC epsilon selective inhibitor peptide or related compounds have been proposed for the treatment of chronic pain states but needs to be tested in animals before their use in painful human conditions [136]. From these considerations it can be argued that much research is needed to define the putative role of strategies against NGF and NGF-mediated intracellular changes. If this 121

9 Sarchielli & Gallai approach is developed, research should be focused on discovering drugs with few central side effects and that have a specific selective action on central sites involved in chronic head pain. Finally, strategies against NGF-dependent BDNF production should be regarded with particular attention. BDNF intervenes, not only in enhancing glutamate transmission which plays a pivotal role in central sensitization but also in exerting an inhibitory action on SP release therefore modulating primary sensory neuron synaptic efficiency via the facilitation of the potassium stimulated release of γ-aminobutyric acid (GABA) from dorsal horn interneurons [137]. The role of these two opposing BDNF effects should be investigated in animal pain models before developing novel molecules targeting this neurotrophin, with the potential application to chronic pain and particularly CDH. Five-year view The evidence of increased NGF levels in the CSF of CDH patients confirms the role of this neurotrophin in maintaining central sensitization in this pathological condition. Sensitization can involve trigeminal nociceptive neurons resulting in the increased and maintained activation of the trigemino-vascular system. To substantiate this hypothesis, increased levels of neuropeptides from the trigemino-vascular system have been concurrently found in the CSF of CDH patients, without significant differences between those with simple and combined analgesic abuse and those without. The increased levels of these two sensory neuropeptides were significantly associated with the increased levels of NGF in the CSF of patients affected by CDH, suggesting the putative role of NGF in enhancing the synthesis, transport, content and release of SP and CGRP in sensory neurons of the trigemino-vascular system as in animal pain models [113]. The finding of increased levels of neuropeptides associated with NGF is further supported by experimental data showing the remodelling of CGRP fibers surrounding NGF-immunoreactive cell bodies and increased CGRP and SP, as well as NK1 receptor expression due to NGF. This may, at least in part, account for the long-lasting sensitization and activation of this system, which may contribute to maintaining head pain. A persistent upregulation of NGF can be hypothesized, not only in the peripheral trigeminal endings but also in the trigeminal nucleus caudalis in CDH. The upregulation of NGF has been scarcely investigated in supraspinal structures in experimental pain models. NGF in these structures could potentially be involved in CDH. Some recent data suggests that NGF levels could influence the dorsal horn neurons via the release of BDNF from primary afferents [107]. The results of one more recent study performed Key issues Chronic daily headache (CDH) affects approximately 4 5% of the general population and is one of the most frequently encountered syndromes at tertiary headache centers. This term encompasses a number of different diagnoses, in particular chronic migraine and chronic tension-type headache. The pathogenesis of CDH is poorly understood. However, a derangement of the serotonergic system and an increase in nitric oxide and glutamate production, as well as substance P and calcitonin gene-related peptide release, have been demonstrated in patients complaining of chronic migraine with and without symptomatic drug abuse. Current therapeutic strategies for CDH include withdrawal from drug abuse before deciding the most appropriate prophylactic treatment. This was aimed to reduce the frequency, severity and duration of headache attacks, to reduce disability and to improve responsiveness to acute therapies. Several pharmacological options exist and include antidepressants, anticonvulsants, muscle relaxants, serotonin antagonists, antianxiety agents and other miscellaneous drugs. In the last few years, experimental evidence has suggested the involvement of nerve growth factor (NGF) in acute hyperalgesia and the maintenance of chronic pain. NGFs pivotal role has been emphasized in neuroplastic changes underlying peripheral and central sensitization in persistent pain. Recent data by our group suggests an increase in cerebrospinal fluid NGF levels in CDH patients compared with control subjects. These levels were significantly correlated with brain-derived neurotrophic factor, glutamate, substance P and calcitonin gene-related peptide CSF levels. The above findings support the involvement of this neurotrophin in enhancing the production of the neuropeptides from sensory neurons of the trigemino-vascular system and glutamatergic transmission via brain-derived neurotrophic factor (BDNF) in CDH patients. Novel strategies targeting neurotrophins (NGF and BDNF) and their receptors have been proposed for the treatment of chronic pain. Blocking the effect of NGF using neutralizing antingf antibodies or TrkA-Ig fusion proteins has been suggested for the treatment of chronic pain states. Moreover, different neurotrophin mimetics with antagonistic action have been produced with potential clinical applications. The putative effects of the above strategies on intracellular signalling should, however, be carefully evaluated before their use in humans. If this approach is developed, research should focus on discovering drugs with few central side effects and selective mechanisms of action on pain processing pathways. 122 Expert Rev. Neurotherapeutics 4(1), (2004)

10 Nerve growth factor and chronic daily headache by our group support the NGF-dependent production of BDNF as in experimental models of pain [113]. In these models, the nociceptive effect of NGF, whose expression is increased in basal and stimulus-induced hyperalgesia appears to be exerted through the upregulation of BDNF mrna and protein and thus leads to the increased sensitivity and action potentials in nociceptive neurons. The above mechanism, clearly shown in animal pain models, may also occur in CDH and sustain central sensitization as the basis of chronic head pain. An increase of glutamate transmission has been demonstrated to underlie central sensitization and chronic pain in experimental animal models. The demonstration of increased glutamate levels in the CSF of CDH patients demonstrated by our group is entirely compatible with experimental evidence that bound central sensitization to the NMDA-mediated nociceptive-responses and concur with further recent data from our group [60]. Experimental findings suggest that the neurotrophin NGF may be an endogenous modulator of synaptic activity via the enhancement of NMDA-mediated responses. In particular, BDNF is responsible for the induction of NR2B subunit expression, which defines the pharmacological and biophysical properties of the receptors and also induces the phosphorylation of NR1 and NR2B subunits [109,114]. Through these changes, BDNF is able to enhance the NMDA response via a threefold increase in NMDA receptor open time [115]. Information resources Markenson JA. Mechanisms of chronic pain. Am. J. Med. 101(1A), S6 S18 (1996). Woolf CJ, Salter MW. Neuronal plasticity: increasing the gain in pain. Science 288(5472), (2000). Bendtsen L. Sensitization: its role in primary headache. Curr. Opin. Investig. Drugs 3(3), (2002). Srikiatkhachorn A. Pathophysiology of chronic daily headache. Curr. Pain Headache Rep. 5(6), (2001). Lake AE 3rd, Saper JR. Chronic headache: New advances in treatment strategies. Neurology 59 5(Suppl. 2), S8 S13 (2002). Mendell LM, Munson JB, Arvanian VL. Neurotrophins and synaptic plasticity in the mammalian spinal cord. J. Physiol. 533(Pt 1), (2001). Bennett DL. Neurotrophic factors: important regulators of nociceptive function. Neuroscientist 7(1), (2001). Acknowledgements The authors express their gratitude to John Toomey for editing the English and Marisa Morson for technical assistance. References Papers of special note have been highlighted as: of interest of considerable interest 1 Pascual J, Colas R, Castillo J. Epidemiology of chronic daily headache. Curr. Pain Headache Rep. 5(6), (2001). 2 Welch KM, Goadsby PJ. Chronic daily headache: nosology and pathophysiology. Curr. Opin. Neurol. 15(3), (2002). 3 Headache Classification Committee of the International Headache Society. Classification and diagnostic criteria for headache disorders, cranial neuralgias and facial pain. Cephalalgia 8(Suppl. 7), 1 96 (1988). 4 Olesen J, Rasmussen BK. The International Headache Society classification of chronic daily and neardaily headaches: a critique of the criticism. Cephalalgia 16(6), (1996). 5 Manzoni GC, Granella F, Sandrini G, Cavallini A, Zanferrari C, Nappi G. Classification of chronic daily headache by International Headache Society criteria: limits and new proposals. Cephalalgia 15(1), (1995). 6 Rothrock J, Patel M, Lyden P, Jackson C. Demographic and clinical characteristics of patients with episodic migraine versus chronic daily headache. Cephalalgia 16(1), (1996). 7 Srikiatkhachorn A, Phanthumchinda K. Prevalence and clinical features of chronic daily headache in a headache clinic. Headache 37(5), (1997). 8 Mathew NT, Stubits E, Nigam MP. Transformation of episodic migraine into daily headache: analysis of factors. Headache 22(2), (1982). 9 Silberstein SD, Lipton RB, Solomon S, Mathew NT. Classification of daily and near-daily headaches: proposed revisions to the IHS Criteria. Headache 34, 1 7 (1994). 10 Silberstein SD, Lipton RB, Sliwinski M. Classification of daily and near-daily headaches : field trial of revised IHS Criteria. Neurology 47(4), (1996). 11 Bigal ME, Sheftell FD, Rapoport AM, Lipton RB, Tepper SJ. Chronic daily headache in a tertiary care population: correlation between the International Headache Society diagnostic criteria and proposed revisions of criteria for chronic daily headache. Cephalalgia 22(6), (2002). 12 Mathew NT, Kurman R, Perez F. Drug induced refractory headache clinical features and management. Headache 30(10), (1990). 13 Stewart J, Tepper MD. Debate: analgesic overuse is a cause, not consequence, of chronic daily headache. Headache 42, (2002). 14 Linton-Dahlöf P, Linde M, Dahlöf C. Withdrawal therapy improves chronic daily headache associated with long-term misuse of headache medication: a retrospective study. Cephalalgia 20(7), (2000). 15 Katsarava Z, Fritsche G, Muessig M, Diener HC, Limmroth V. Clinical features of withdrawal headache following overuse of triptans and other headache drugs. Neurology 57(9), (2001). 16 Srikiatkhachorn A. Chronic daily headache: a scientist s perspective. Headache 42(6), (2002). Treats molecular mechanisms underlying central sensitization and biochemical abnormalities found in chronic daily headache patients. 17 Millan MJ. The induction of pain: an integrative view. Prog. Neurobiol. 57(1), (1999). 18 Sessle BJ. Neural mechanisms and pathways in craniofacial pain. Can. J. Neurol. Sci. 26(Suppl. 3), S7 S11 (1999). 19 Urban MO, Gebhart GF. Central mechanisms in pain. Med. Clin. North Am. 83(3), (1999)