Abstract
EEG-studies in migraine in the last decade has contributed modestly to the understanding of headache pathogenesis. Headache patient groups seem to have increased EEG responses to photic stimulation, but a useful biological marker for migraine in single patients has not been found. In future EEG and QEEG studies we recommend to use follow-up designs and record several EEGs across the migraine cycle. It is also important to use a blinded study design in order to avoid selection bias.
A clinical EEG should be performed in patients with acute headache attacks when either epilepsy, basilar migraine, migraine with prolonged aura or alternating hemiplegia is suspected. Unequivocal epileptiform abnormalities usually suggest a diagnosis of epilepsy. In children with occipital spike-wave activity the probable diagnosis is childhood epilepsy with occipital paroxysms (CEOP). The final diagnosis of either an epilepsy syndrome or migraine must be mainly based on a clinical judgement.
Introduction
In the last 15 years, the provision of animal models for migraine has produced impressive advances regarding data on the mechanisms and mediators underlying migraine attacks. The trigeminovascular system has emerged as a critical efferent component, and the mediators of its activity have been identified and characterized, as have some of the receptors involved. Studies with substances known to induce migraine-like attacks have provided interesting insights into the framework of central nuclei that are probably involved in the initiation and repetition of migraine attacks.
Neurogenic inflammation and the trigeminovascular system: the role of triptans
The neurogenic inflammation (NI) model during the past decades has helped the understanding of possible mechanisms underlying migraine attack pathophysiology and drug action. The NI model represents the vasodilatation and plasma protein extravasation that follows release of vasoactive peptides such as calcitonin gene-related peptide (CGRP), substance P (SP) and neurokinin A from nerve endings, induced by activation of sensory endings. In experimental animal models of migraine, NI is induced either by systemic administration of capsaicin (the pungent ingredient of hot pepper that is able to depolarize sensory fibres, or to abolish their functions when given to neonate animals) or by stimulating trigeminal sensory fibres using electrical stimulus in the rat (1). Unilateral electrical trigeminal ganglion stimulation (UETGS) induces oedema in ipsilateral tissues (either intra- or extra-cranial) receiving trigeminal innervation on vasculature, whereas systemic capsaicin depolarizes sensory fibres all over the body, producing diffuse tissue oedema, as well as cardiorespiratory side-effects. Administration of SP is also able to induce tissue oedema without direct involvement of sensory fibres (2).
Sumatriptan, the selective agonist of 5-HT1B/1D receptors that was designed to abolish migraine pain, has proven able to block NI in rat dura mater following UETGS or capsaicin administration via prejunctional receptor mechanism (2, 3). Following this observation, the NI model was used to test a number of drugs with activitity on 5-HT1B/1D receptors in order to predict efficacy in migraine pain. In this model, not only was plasma protein extravasation blocked by pretreatment with sumatriptan, but the drug also proved effective on single components of the neurogenic response to peripheral trigeminal ganglion stimulation, such as CGRP increase in plasma from the superior sagittal sinus (4) and mast cell activation and degranulation (3). Dihydroergotamine, the ergot derivative commonly used to abort migraine pain, which has, among other receptor activities, 5-HT1B/1D agonist properties, has been found to share similar effects (5). Sumatriptan is also effective in reducing the increase of CGRP in the jugular vein during migraine attack (6), however, this data should be interpreted cautiously, as CGRP plasma levels increase at the very beginning of electrical stimulation of the trigeminal ganglion and start decreasing when the stimulation is still running (4). Therefore, there is probably a massive release from nerve endings with a rapid fall of the peptide contained, and a consequent lack of further release. During a migraine attack there is probably time for the peptide to increase again in the vescicles, but this takes time and would only be evident some hours from first increase of the peptide in the blood. Administration of sumatriptan may or may not inhibit such further release and this mechanism might explain both the effectiveness of the drug and the recurrence of headache, as seen in a certain number of patients treated with triptans.
In the NI rat model, mast cells contribute to further sensitization of sensory fibres by releasing substances locally in the tissue (7). The dense sensory innervation of dura mater puts this membrane at high risk of being continuously activated and thereby sustaining a migraine attack for many hours, via a trigeminovascular mechanism. Following the above observation, several drugs have been developed to ameliorate the pharmacological performances of triptans. Discovery of a central site of action, due to penetration in the central nervous system (CNS), has been added to the peripheral site, and several compounds are now on the market or at the very advanced preclinical stage (for a recent review on triptan key sites of action see reference 8). The central mechanism has been advocated because migraine attack is considered a discharge from a central ‘generator’, probably located in the brainstem (9). However, the central action of the ‘second-generation’ triptans does not seem to have added any further efficacy against pain, while headache recurrence seems more influenced by the longer duration of action of a triptan than by its central activity (10).
NI has been proposed as the peripheral component of a migraine attack (11) following other primary steps occurring in the CNS. Most of the drugs that have been tried in this model are highly effective at inhibiting or decreasing the response, although not all of them have been shown to be effective in migraine pain. The lack of efficacy on pain might well be due to dose adjustment. In addition, it is possible that the efficacy in experimental NI does not need recruitment of vascular receptors as ‘coworkers’ in the inhibition of oedema, which might be a necessary step in clinical practice. This could be the case for the conformationally sumatriptan restricted analogue CP122,288, a compound that shows high potency in NI models (12) but not on pain (13), and is devoid of 5-HT1B-mediated vascular action.
The usefulness of the NI model is also evident when testing prophylactic drugs such as valproate, a gamma-amino butyric acid (GABA)-agonist, which has been shown to be effective in blocking dural plasma extravasation following UETGS and SP administration, via a bicuculline-reversible mechanism (14) and c-Fos expression following intracisternal capsaicin, suggesting a role for GABAA receptor in migraine attack pathophysiology (15).
The role of NO donors in headache mechanisms
The capability of nitrovasodilators to act as pro-drugs that release nitric oxide (NO) in several body tissues (vessel wall, lungs and brain) has favoured a resurgence of scientific interest for this group of substances, which were originally used for their vasodilatory effect in the treatment of ischaemic cardiac disease (16, 17). The demonstration that NO plays a pivotal role in the control of several functions in the CNS (nociception, toxicity, degeneration, memory) has prompted the use of NO donors as probes for the study of the role of NO in a variety of neurological diseases (18–21).
Among the various nitrovasodilators commercially available, nitroglycerin (NTG) has undergone extensive experimental investigation in the neurological field, because typical headaches develop in migraineurs (but not normals) with a 4–6 h latency after its administration (22–25). NTG is highly lypophilic and easily crosses the blood–brain barrier (17). Experimental evidence for its accumulation in the brain tissue has been provided (26), and it has been demonstrated that systemic administration of this organic nitrate induces neuronal activation in several brain nuclei belonging to the neurovegetative, neuroendocrine, behavioural and nociceptive systems (27, 28). This activation, in some areas, develops with a latency of hours, which contrasts with the very short plasma half-life of NTG. Co-localization studies have shown that NTG-induced neuronal activation takes place in adrenergic, nitrergic and neuropeptidergic structures (29, 30), thus suggesting some of the possible signalling pathways involved in the phenomenon. Neuropharmacological manipulation of nitroglycerin-induced neuronal activation showed that a dual mechanism, acting on a double target, is probably involved: (i) exogenous (NTG-derived) NO might directly act at both the vascular and neuronal levels; and (ii) the gaseous mediator might activate indirectly neurovascular responses via multiple pathways that include the synthesis of its endogenous homologous, the activation of cyclo-oxygenase-mediated mechanisms, and the induction of a trigeminovascular-mediated biochemical response (31).
The temporal profile of neuronal activation following NTG shows that neuronal activation begins as early as 60 min postinjection in brain areas that control the cardiovascular function, and it reaches maximum expression 3 h later in nociceptive and integrative structures (27). This modulated temporal course again suggests a dual mechanism of action for NTG on the brain; an initial effect on the vascular compartment followed by the involvement of integrative–nociceptive structures.
Data from our group have shown that NTG administration induces changes in the noradrenergic asset at both the vascular and neuronal levels, and in the serotonergic asset in specific brain areas (32). The observed changes in central and/or peripheral neurotransmission are likely to induce a hyperalgesic state, which translates in the sustained activation of nociceptive nuclei in the rat (27). This explanation could also account for the initiation of a spontaneous migraine-like attack in predisposed subjects following NTG administration (22–25). This hypothesis has also been supported by the recent findings obtained by other groups, which strongly suggest a definite role for NTG in pain mediation. Pardutz et al. (34) showed that NTG administration increases the number of NOS-immunoreactive cells in the rat spinal trigeminal nucleus, which points to the activation of second-order neurones via a presynaptic excitatory mechanism. Lambert et al. (35) demonstrated that systemic NTG increases the firing rate of second-order trigeminal neurones, which transport inputs from cranial structures via a serotonin-mediated mechanism that is prevented by the administration of selective 5-HT1B/1D agonists. Finally, Reuter et al. (36) demonstrated that NTG administration induces an up-regulation of pro-inflammatory genes, with a subsequent, delayed inflammatory reaction in the dura mater of the rat.
Taken together, these findings strongly suggest that the study of neuronal and vascular effects of nitric oxide donors, and more specifically of NTG, might yield an increasing body of evidence for a better understanding of the pathophysiology of migraine attacks and of the role played by NO.
Conclusions
The experimental NI model represents a simple procedure to obtain preliminary information on well-characterized receptor drugs. The apparent paradox observed with certain drugs that are shown to be effective in this model but not in clinical trials offers the opportunity of better manipulating compounds and delineating the best pharmacological profile. In addition, the model offers the opportunity to observe other consequences of peripheral trigeminal activation that might be useful for understanding central components of the complex trigeminovascular–CNS pain pathway. Headache induced by NO donors in migraine patients relates to observations in experimental animal models and provides evidences for the involvement of NO on neuronal and vascular components in migraine pathophysiology.