Abstract
In the last 30 years dopamine has been considered as playing a role in the pathogenesis of migraine. The literature indicates that migraineurs are hypersensitive to dopamine agonists with respect to some of the premonitory symptoms of migraine such as nausea and yawning. There are various nonspecific dopamine D2 receptor antagonists that show good clinical efficacy in migraine, and also a number of polymorphisms of dopaminergic genes related to migraine. Animal studies have also shown that dopamine receptors are present in the trigeminovascular system, the area believed to be involved in headache pain, and neuronal firing here is reduced by dopamine agonists. There appears to be little effect of dopamine on peripheral trigeminal afferents. We assess some of the limitations of the clinical studies with regard to the therapeutics, and those found in the studies that discovered differences in genetic polymorphisms in migraine, and consider the implications of this on a dopaminergic hypothesis of migraine.
Migraine is a common neurological problem whose pathophysiology is still to be completely understood (1). Migraine is a multifactorial disorder affecting both peripheral and central neurotransmitter systems. A great many hypotheses exist and the main focus has concentrated on the headache phase of migraine. A substantial body of work points towards the involvement of monoamine serotonin 5-HT receptors and agonists (1), and the classic antimigraine therapy now is the serotonin, 5-HT1B/1D, receptor agonists, ‘the triptans’ (2). One of the main neuropeptides believed to be involved in migraine is calcitonin gene-related peptide (CGRP) (3), and antagonists to the CGRP receptor show good efficacy in the clinic (4, 5). Although triptans and CGRP receptor antagonists represent very effective acute therapies, many other approaches are used as preventive and acute treatments, and rely on other receptor systems and endogenous agonists. Looking at other pharmacological approaches, indeed at migraine as a whole, may be important for understanding the disorder and improving its therapy.
Other investigations have looked at some of the autonomic ‘non-headache’ related symptoms that are often associated with migraine. Sicuteri has proposed a role for dopamine in the nausea, vomiting and blood pressure changes that take place during the attack (6). Coupled with the data that migraine patients are hypersensitive to dopamine agonists such apomorphine, a dopaminergic theory has been developed. Although the impetus for studying dopamine as a major neurotransmitter in the pathogenesis of migraine has slowed somewhat in recent years, the ability to screen patients genetically has revived interest. Genetic studies have demonstrated that certain alleles of dopamine D2 receptors (DRD2) may be implicated in migraine with aura (7, 8). Moreover, there is recent evidence that dopamine may, in some part, be involved in the modulation of trigeminovascular transmission (9), at least in animal models; therefore dopamine once again is being considered an important monoaminergic transmitter in migraine.
THE DOPAMINERGIC SYSTEM
Dopamine (DA) belongs to the group of neurotransmitters called catecholamines that were first identified in the brain by Carlsson and colleagues (10). They are derived from tyrosine and also include noradrenaline and adrenaline. Their synthesis starts with the conversion of phenylalanine to tyrosine via phenylalanine hydroxylase. Tyrosine hydroxylase oxidases then converts tyrosine to
Dopaminergic cell groups within the brainstem. (A) A summary of the major dopaminergic pathways in the midbrain, including the mesostriatal and mesocorticolimbic systems located in the A8, 9 and 10. (B) The other dopaminergic pathways. The A11-14 diencephalic cell groups include tuberoinfundibular and tuberohypophysial pathways and hypothalamic projections to the spinal cord, A15 includes cells in hypothalamus, and dorsal and ventrolateral preoptic area, A16 is the olfactory bulb and A17 (not shown) is the retinal dopaminergic neurons. Adapted from (11, 12, 16, 96). EC, Entorhinal cortex; LC, locus coeruleus; LPN, lateral parabrachial nucleus; LS, lateral septum; NA, nucleus accumbens; olfT, olfactory tubercle; PAG, periaqueductal grey; PC, pyriform cortex; RMN, raphe magnus nucleus; STN, spinal trigeminal nucleus.
DOPAMINE AND MIGRAINE
Premonitory symptoms
Migraine is often characterized by yawning, drowsiness, mood changes, irritability and hyperactivity in the days and hours preceding an attack (23–26). Administration of apomorphine, a dopamine agonist, at low doses (0.25 mg/kg sublingual, 0.02–0.1 mg/kg subcutaneous) caused an increase in the cumulative number of yawns assessed as a function of time in migraineurs compared with age-matched controls (27–30). Some of these studies have also measured nausea, vomiting, dizziness and sweating and also found an increase in their occurrence compared with controls (27). This suggests that migraineurs are hypersensitive to dopamine agonists. In rats, yawning can be induced by agonists that are specific to the family of D2 receptors, not D1, but the response is antagonized by both D1 and D2 receptor antagonists (31, 32). The proposed mode of action is central, as domperidone, the peripheral D2 antagonist, was unable to inhibit the yawning seen in rats. Increasing doses of dopamine agonists also cause hyperactivity and irritability in animals (32). Nausea and vomiting are other symptoms associated with migraine, according to the International Headache Society (IHS) criteria (33), occurring during both premonitory and headache phases. Dopamine D2 antagonists are the classic antiemetics, which implicates them in migraine, and apomorphine and piribedil provoke increased emetic responses in migraineurs (34). There may also be gastrokinetic dysfunction during migraine, with reduced absorption of drugs during the attack that is reversed by metoclopramide, a dopamine antagonist (35). These examples suggest a hypersensitivity of migraineurs to dopamine, particularly with regard to the premonitory symptoms. Weaknesses of these studies are that they have not been replicated since the introduction of either of the IHS classifications for headache (33), and in the main represent small and uncontrolled samples, which limits their impact (36).
Hypotension is another consideration in migraine, with bromocriptine (2.5 mg) inducing a hypotensive reaction in normotensive migraineurs, but not in a control group (37). Domperidone was able to reverse these hypotensive and nausea effects, indicating a peripheral locus. This study again suffers from the limitations of a sample size of 18, but is consistent with a role for dopamine in the biology of migraine.
Endocrinological changes
The tuberoinfundibular dopaminergic system that controls the secretion of prolactin from the pituitary gland (see Fig. 1B) may be altered in migraine. A monoamine oxidase-B inhibitor was more effective at reducing prolactin levels in migrainous individuals than controls (38). Monoamine depletors, which decrease peripheral dopamine, cause a more prolonged release of prolactin in migraineurs than in controls (39). In migrainous women, prolactin levels in response to dopamine agonists and antagonists during the follicular phase of the menstrual cycle vary significantly from controls (40). Taken together, these data might imply an increased lactotroph prolactin reserve in migraineurs (41). Although this highlights differences between migraineurs and controls, with regard dopaminergic pathways and their related systems, it does not provide a direct link between dopamine and the cause or even induction of migraine.
Headache phase and migraine pain
The data that report dopamine itself inducing migraine, or head pain, are conflicting. If one considers Parkinson's disease, many patients receive dopamine receptor agonists or L-DOPA, thus representing a unique patient group (42). In general there are more studies showing (28) migraine does not occur with dopamine receptor agonists than those that do; thus studies suggesting dopamine receptor agonists cause migraine pain should be treated with caution (27, 28, 43).
It is likely that the headache phase of migraine involves activation of the trigeminal afferents that innervate the intra- and extracranial vasculature, including dural and cerebral blood vessels (1). It is known that dopamine receptors do exist in the trigeminal ganglion and spinal trigeminal nucleus, in the mouse and rat (44–46), and more recently D1 and D2 receptors have been found on cell bodies of the rat trigeminocervical complex using antibodies specific to the individual receptors (9). D1 receptors were most densely populated in the deeper laminae, whereas D2 receptors were much less densely populated in deeper laminae and showed fibre staining in laminae I/IIo.
Microiontophoretic application of dopamine is able to inhibit reversibly the rate of neuronal firing of durally activated neurons in the trigeminocervical complex (9). Similarly, dopamine is able to inhibit trigeminal firing in response to iontophoresed
Studies on the cerebral blood flow have shown an increase in flow after apomorphine or piribedil (34, 50), showing alteration with the cerebral vasculature. Animal studies have also shown the pial arteries and the middle cerebral arteries in cats respond to apomorphine and dopamine, with vasodilation at low doses, and vasoconstriction at higher doses, in an in vitro model (51, 52). There were decreases in canine external carotid artery conductance (an indication of vasodilation) after dopamine and apomorphine injection into the internal carotid (53). These changes were reversed by a D2 receptor antagonist. Meningeal dural blood vessel vasoconstriction occurred with both dopamine hydrochloride (20 and 40 µg kg−1 min−1) and a D1 receptor agonist, A68930 hydrochloride (10 and 50 µg kg−1 min−1), at high doses (54). A α2-adrenoceptor antagonist was able to reverse the effects of dopamine hydrochloride on the vessel diameter, and the blood pressure changes of the D1 receptor agonist. Neurogenic dural vasodilation, which uses electrical stimulation of dura mater to activate trigeminal afferents, allows one to observe this response with changes in vessel diameter. It was found that dopamine hydrochloride and a D1 receptor agonist were able to inhibit neurogenic dural vasodilation, but this response was partially inhibited by both a α2-adrenoceptor and a D1 receptor antagonist (54). This seems to indicate that the major blood vessel changes that take place within the cranium may be α2-adrenoceptor mediated—related to blood pressure changes, with a small component of D1 receptor activation.
Dopamine genetics
Molecular genetics offer a new approach to studying migraine, since the discovery that migraine may have a genetic component to it with regard to channelopathies (55–58). These channelopathies have been implicated thus far in a rare form of migraine, familial hemiplegic migraine, but open up the possibility that more common forms of migraine have a genetic component, indeed involve ionopathies. Peroutka and colleagues (1997) were the first to suggest a migraine gene that implicates dopamine biology. A polymorphism in the NcoI gene that encodes DRD2 was increased in migraine with aura (0.84) compared with migraine without aura (0.71) or controls (0.70) (7). From these data, the authors have suggested that modification of the gene that encodes the DRD2 may be involved in migraine pathophysiology. Another study has found no link with DRD3 and DRD4 polymorphisms in migraine without aura, but a subgroup of patients that presented with yawning and nausea prior to or during the headache phase did have a DRD2 polymorphism; this could not be replicated throughout the entire sample (8). Transmission distortion with a marker for DRD4 polymorphism has been found in a migraine without aura group compared with controls (59). Studies using larger sample sizes, looking at polymorphisms in genes that encode for DRD1, DRD2, DRD3 and DRD5, have not shown any significance in polymorphisms with either migraine with aura or migraine without aura compared with controls (60, 61). Other studies have similarly not found a relationship with polymorphisms on the DRD2 gene and migraine (62–64).
Investigations of polymorphisms in other dopamine genes have shown some relationships in migraine. There is a reduced allelic distribution for dopamine-β-hydroxylase (d.b.h.) polymorphisms compared with control (60), suggesting increased levels of dopamine in migraineurs. The possible role of d.b.h. polymorphism distribution has also been shown to be marked in migraine with aura compared with controls, particularly in males (65). Other studies have highlighted different possible variations in dopamine genetics, with under-represented dopamine transporter (DAT) polymorphisms in chronic headache compared with migraine without aura (66), but no difference in DAT allelic distribution in migraine with or without aura compared with controls (66, 67). A separate study has examined three functional polymorphisms in dopamine genes, on variable number of tandem repeats in the DRD4, DAT and d.b.h. The study concentrated on child migraine and adhered strictly to the IHS classification criteria (68). The authors found no significant transmission distortion in any of the markers studied. Although there are accumulating data on alterations in dopamine genetics in migraine, the picture is still very cloudy. There are as many, if not more, studies showing no relationship between differing dopamine genetics and migraine than there are those that do. Certainly some of the early studies suffer from very small samples sizes; therefore, it is clear that further work needs to be done in this area before any definite conclusions can be drawn.
Dopaminergic therapeutics
Dopamine DRD2 receptor antagonists have been used frequently in the clinic, mainly as antiemetics, and some have broader efficacy against other symptoms in migraine. One theory is that dopamine antagonists prevent the nausea accompanying the attack, and then correct the possible hypersensitivity to dopamine in the brain. Domperidone (20/30 mg), the peripheral DRD2 antagonist, decreased the duration of an attack by 30%, only when combined with paracetamol (69). It is possible that domperidone is improving the absorption of paracetamol (acetaminophen) to relieve the headache rather than having any effect on migraine per se. Domperidone/paracetamol has also been shown to be similarly efficacious as sumatriptan (50 mg) over a 2- and 4-h postdose period (70). Chlorpromazine (CPZ, another DRD2 antagonist) is commonly used in the emergency room, and intramuscular injection has been more effective than placebo (71). It is also effective at relieving many other symptoms, such as nausea, phonophobia and photophobia, in patients compared with placebo, with and without aura (72). In comparative studies, with dihydroergotamine (1 mg, i.v.) and lidocaine (50–150 mg, i.v.), CPZ (12.5–37.5 mg, i.v.) was deemed to provide most headache relief (73), although sample size was very small (24–26 per group). CPZ (0.1 mg, i.v., up to three doses) in comparison with metoclopramide (same dose) (74), or CPZ (12.5–37.5 mg, i.v.) compared with sumatriptan (6 mg, i.m.) proved equally efficacious (75). Prochlorperazine (PCZ) is a potent DRD2 antagonist that has been used in migraine clinics. In the emergency room PCZ (10 mg) was more effective against pain and nausea than both metaclopramide (10 mg) and placebo (76). PCZ (3 mg) was also more effective than combined ergotamine tartrate (1 mg) and caffeine (100 mg) (77). In a trial using only a small sample size, comparing naratriptan/PCZ with naratriptan/placebo, naratriptan/placebo presented with more pain-free subjects at 4 h (78).
Flunarizine is a calcium channel blocker often used as both an acute and prophylactic migraine treatment (79–81) and also appears to have action as a dopamine antagonist, specifically DRD2, although it is not clear whether its antimigraine action is actually via dopaminergic mechanisms (81, 82). Other dopamine antagonists studied in migraine prophylaxis include lisuride, bromocriptine and ergocryptine, none of which is widely regarded as more helpful than flunarizine. Lisuride and ergocryptine are both ergot alkaloids with effects on DRD2 receptors. Lisuride significantly reduced migraine frequency and severity (83, 84). α-Dihydroergocryptine (10 mg twice daily) has proved comparable to both flunarizine (5 mg daily) and propranolol (40 mg twice daily) at reducing headache frequency (85, 86), and also to dihydroergotamine with regard to reducing pain severity (87). Continuous bromocriptine reduced menstrual migraine frequency when combined with the patient group's current regimens (88). These studies imply a possible role for dopaminergic mechanisms as a therapeutic target in migraine. The compounds used vary greatly in their selectivity and potency to the DRD2 receptor, therefore it is not clear how they exert their action in migraine. They could be as easily acting via another mechanism than the dopaminergic system, such as serotoninergic, cholinergic, adrenergic, histaminergic or even calcium channels (89–93), and until their action is clarified one has therefore to be aware of their dopaminergic potential, but acknowledge their abilities at other receptor systems.
Possible mechanism of dopaminergic pathway in migraine
Proposing a dopaminergic mechanism of migraine is difficult with the level of conflicting data present in the current literature, and any theories are clouded as a consequence. The dopamine theory originally proposed by Sicuteri in 1977 implied that dopamine agonists cause yawning, nausea and blood pressure changes that are similar to some of the premonitory symptoms found in migraine, and therefore drive symptoms in migraine itself. This would seem to indicate that the migrainous brain is hypersensitive to dopamine. It is believed that migraineurs have a lower threshold for migraine triggers that then drive the migraine attack. The dopaminergic theory would indicate that dopamine is such a trigger, although this is not borne out in the literature. Genetic studies imply that polymorphisms in dopamine genes may create this hypersensitivity. Polymorphisms in DRD2 receptors may make them over-responsive to dopamine, or dysfunction of dopamine transporters may prevent the removal of dopamine from the synaptic cleft, or in d.b.h. may suggest increased basal levels of dopamine in the brain, therefore less dopamine is needed to drive central changes. Genetic factors would certainly contribute to the idea of a migrainous brain. The recent evidence of the presence of dopamine receptors in the trigeminocervical complex, and dopaminergics contributing to the modulation of neuronal firing in this area, would indicate a direct link to migraine pathophysiology. Furthermore, the evidence of a direct descending link from A11 hypothalamic neurons to trigeminal neurons, which serve as the sole source of dopaminergic neurons to the spinal cord, may explain a pathway of action and site for the dopaminergic polymorphisms to exert their effect. If we then consider therapeutics, dopamine antagonists are known to reduce both migraine severity and occurrence, particularly DRD2 antagonists. DRD2 receptors are known to predominate over DRD1 in the trigeminocervical complex, which may explain a site of action of therapeutics. A mechanism of dopamine pathophysiology begins to take shape, where dopamine can induce premonitory symptoms, by an action at dopaminergic proteins that are in some way hypersensitive to dopamine, and this is carried forward into the migraine pain generated in the trigeminovascular systems, mediated predominantly by DRD2 receptors.
The data and ideas presented above have lent strong support implicating dopamine biology in migraine pathophysiology at the turn of the 21st century. Unfortunately, nearly 10 years on, and with an updated review of the literature, the role of dopamine in migraine pathogenesis may be less clear. The genetic studies are less than convincing, particularly for DRD2 gene polymorphisms, where sample size has often been too small to be reliable. Polymorphisms in DRD1, 3 and 4 or dopamine transporter have also been discounted. In fact, more reliable data seem to implicate dopamine-β-hydroxylase, where two studies have shown a link with migraine, particularly with aura in males (60, 65). Dopaminergic activity in the trigeminocervical complex is restricted to dopamine agonists inhibiting neuronal firing, whereas in the clinic it is dopamine antagonists that are effective. It is undeniable that these dopamine antagonists are effective in migraine; however, it is still not clear whether their action is via dopamine receptors or some other mechanism. The majority of the compounds used do have a lack of selectivity at dopamine receptors, which always makes it possible or even likely that they are acting via another system. There has never been a rigid testing of these compounds, few have used double-blind, placebo-controlled trials, and sample size has often been limited. Furthermore, the success of these trials and the hypersensitivity to dopamine have often been found in migraine without aura, and yet the genetic polymorphisms indicate predominance in migraine with aura.
Dopamine may be involved in some of the premonitory symptoms such as nausea and yawning, and may also contribute to the hypotensive changes. The role of dopamine antagonists as treatment for migraine may help with some of the premonitory symptoms, and also combination treatments may work by improving gastric absorption that DRD2 antagonists provide, but it is still certainly not clear whether dopamine is a major transmitter involved in the pathogenesis of migraine in the same way that serotonin, CGRP or nitric oxide are implicated (1, 94, 95). It is clear that more sustained and controlled clinical studies with dopamine antagonists are necessary, combined with animal experimental data that describe the potential pathway of action of dopamine in migraine, before drawing any conclusion about a dopamine theory of migraine.