The mechanisms of migraine remain incompletely understood. However, new technologies have allowed formulation of current concepts that may explain parts of the migraine syndrome.
In the 1940s and 1950s, the vascular theory was proposed to explain the pathophysiology of migraine headache. Wolff et al believed that ischemia induced by intracranial vasoconstriction is responsible for the aura of migraine and that the subsequent rebound vasodilation and activation of perivascular nociceptive nerves resulted in headache.
This theory was based on the following 3 observations:
- Extracranial vessels become distended and pulsatile during a migraine attack
- Stimulation of intracranial vessels in an awake person induces headache
- Vasoconstrictors (eg, ergots) improve the headache, whereas vasodilators (eg, nitroglycerin) provoke an attack
However, this theory did not explain the prodrome and associated features. Nor did it explain the efficacy of some drugs used to treat migraines that have no effect on blood vessels and the fact that most patients do not have an aura. Moreover, with the advent of newer imaging technologies, researchers found that intracranial blood flow patterns were inconsistent with the vascular theory. No consistent flow changes have been identified in patients suffering from migraine headache without aura. Regional cerebral blood flow (rCBF) remains normal in the majority of patients. However, bilateral decrease in rCBF, beginning at the occipital cortex and spreading anteriorly, has been reported. More recently, Perciaccante has shown that migraine is characterized by a cardiac autonomic dysfunction. As a result of these anomalous findings, the vascular theory was supplanted by the neurovascular theory.
The neurovascular theory holds that a complex series of neural and vascular events initiates migraine. According to this theory, migraine is primarily a neurogenic process with secondary changes in cerebral perfusion. At baseline, a migraineur who is not having any headache has a state of neuronal hyperexcitability in the cerebral cortex, especially in the occipital cortex. This finding has been demonstrated in studies of transcranial magnetic stimulation and with functional magnetic resonance imaging (MRI). This observation explains the special susceptibility of the migrainous brain to headaches. One can draw a parallel with the patient with epilepsy who similarly has interictal neuronal irritability.
Cortical spreading depression
In 1944, Leao proposed the theory of cortical spreading depression (CSD) to explain the mechanism of migraine with aura. CSD is a well-defined wave of neuronal excitation in the cortical gray matter that spreads from its site of origin at the rate of 2-6 mm/min. This cellular depolarization causes the primary cortical phenomenon or aura phase; in turn, it activates trigeminal fibers, causing the headache phase. The neurochemical basis of the CSD is the release of potassium or the excitatory amino acid glutamate from neural tissue. This release depolarizes the adjacent tissue, which, in turn, releases more neurotransmitters, propagating the spreading depression.
Positron emission tomography (PET) scanning demonstrates that blood flow is moderately reduced during a migrainous aura, but the spreading oligemia does not correspond to vascular territories. The oligemia itself is insufficient to impair function. Instead, the flow is reduced because the spreading depression reduces metabolism. Although CSD is the disturbance that presumably results in the clinical manifestation of migraine aura, this spreading oligemia can be clinically silent (ie, migraine without aura). Perhaps a certain threshold is required to produce symptoms in patients having aura but not in those without aura. A study of the novel agent tonabersat, which inhibits CSD, found that the agent helped to prevent migraine attacks with aura only, suggesting that CSD may but not be involved in attacks without aura.
Activation of the trigeminovascular system by CSD stimulates nociceptive neurons on dural blood vessels to release plasma proteins and pain-generating substances such as calcitonin gene-related peptide, substance P, vasoactive intestinal peptide, and neurokinin A. The resultant state of sterile inflammation is accompanied by further vasodilation, producing pain. The initial cortical hyperperfusion in CSD is partly mediated by the release of trigeminal and parasympathetic neurotransmitters from perivascular nerve fibers, whereas delayed meningeal blood flow increase is mediated by a trigeminal-parasympathetic brainstem connection. According to Moulton et al, altered descending modulation in the brainstem has been postulated to contribute to the headache phase of migraine; this leads to loss of inhibition or enhanced facilitation, resulting in trigeminovascular neuron hyperexcitability.
In addition, through a variety of molecular mechanisms, CSD upregulates genes, such as those encoding for cyclo-oxygenase 2 (COX-2), tumor necrosis factor alpha (TNF-alpha), interleukin-1beta, galanin, and metalloproteinases. The activation of metalloproteinases leads to leakage of the blood-brain barrier, allowing potassium, nitric oxide, adenosine, and other products released by CSD to reach and sensitize the dural perivascular trigeminal afferent endings. Increased net activity of matrix metalloproteinase–2 (MMP-2) has been demonstrated in migraineurs. Patients who have migraine without aura seem to have an increased ratio of matrix metalloproteinase–9 (MMP-9) to tissue inhibitors of metalloproteinase–1 (TIMP-1), in contrast to a lower MMP-9/TIMP-1 ratio in patients who have migraine with aura. Measured levels of MMP-9 alone are the same for migraine patients with or without aura.
In an experimental study, acute hypoxia was induced by a single episode of CSD. This was accompanied by dramatic failure of brain ion homeostasis and prolonged impairment of neurovascular and neurometabolic coupling.
Vasoactive substances and neurotransmitters
Perivascular nerve activity also results in release of substances such as substance P, neurokinin A, calcitonin gene-related peptide, and nitric oxide, which interact with the blood vessel wall to produce dilation, protein extravasation, and sterile inflammation. This stimulates the trigeminocervical complex, as shown by induction of c-fos antigen by PET scan. Information then is relayed to the thalamus and cortex for registering of pain. Involvement of other centers may explain the associated autonomic symptoms and affective aspects of this pain. Neurogenically induced plasma extravasation may play a role in the expression of pain in migraine, but it may not be sufficient by itself to cause pain. The presence of other stimulators may be required. Although some drugs that are effective for migraine inhibit neurogenic plasma extravasation, substance P antagonists and the endothelin antagonist bosentan inhibit neurogenic plasma extravasation but are ineffective as antimigraine drugs. Also, the pain process requires not only the activation of nociceptors of pain-producing intracranial structures but also reduction in the normal functioning of endogenous pain-control pathways that gate the pain.
A potential “migraine center” in the brainstem has been proposed, based on PET-scan results showing persistently elevated rCBF in the brainstem (ie, periaqueductal gray, midbrain reticular formation, locus ceruleus) even after sumatriptan-produced resolution of headache and related symptoms. These were the findings in 9 patients who had experienced spontaneous attack of migraine without aura. The increased rCBF was not observed outside of the attack, suggesting that this activation was not due to pain perception or increased activity of the endogenous antinociceptive system. The fact that sumatriptan reversed the concomitant increased rCBF in the cerebral cortex but not the brainstem centers suggests dysfunction in the regulation involved in antinociception and vascular control of these centers. Thalamic processing of pain is known to be gated by ascending serotonergic fibers from the dorsal raphe nucleus and from aminergic nuclei in the pontine tegmentum and locus ceruleus; the latter can alter brain flow and blood-brain barrier permeability. Because of the set periodicity of migraine, linkage to the suprachiasmatic nucleus of the hypothalamus that governs circadian rhythm has been proposed. Discovering the central trigger for migraine would help to identify better prophylactic agents.
PET scanning in patients having an acute migraine headache demonstrates activation of the contralateral pons, even after medications abort the pain. Weiler et al proposed that brainstem activation may be the initiating factor of migraine. Once the CSD occurs on the surface of the brain, H+ and K+ ions diffuse to the pia mater and activate C-fiber meningeal nociceptors, releasing a proinflammatory soup of neurochemicals (eg, calcitonin gene–related peptide) and causing plasma extravasation to occur. Therefore, a sterile, neurogenic inflammation of the trigeminovascular complex is present. Once the trigeminal system is activated, it stimulates the cranial vessels to dilate. The final common pathway to the throbbing headache is the dilatation of blood vessels.
Burstein et al described the phenomenon of cutaneous allodynia, in which secondary pain pathways of the trigeminothalamic system become sensitized during a migrainous episode. This observation demonstrates that, along with the previously described neurovascular events, sensitization of central pathways in the brain mediates the pain of migraine.
Some authors have proposed a dopaminergic basis for migraine. In 1977, Sicuteri postulated that a state of dopaminergic hypersensitivity is present in patients with migraine. Interest in this theory has recently been renewed. Some of the symptoms associated with migraine headaches, such as nausea, vomiting, yawning, irritability, hypotension, and hyperactivity, can be attributed to relative dopaminergic stimulation. Dopamine receptor hypersensitivity has been shown experimentally with dopamine agonists (eg, apomorphine). Dopamine antagonists (eg, prochlorperazine) completely relieve almost 75% of acute migraine attacks.
Another theory proposes that deficiency of magnesium in the brain triggers a chain of events, starting with platelet aggregation and glutamate release and finally resulting in the release of 5-hydroxytryptamine, which is a vasoconstrictor. In clinical studies, oral magnesium has shown benefit for preventive treatment and intravenous magnesium may be effective for acute treatment, particularly in certain subsets of migraine patients.
Vascular smooth muscle cell dysfunction may involve impaired cyclic guanosine monophosphate and hemodynamic response to nitric oxide. Nitric oxide released by microglia is a potentially cytotoxic proinflammatory mediator, initiating and maintaining brain inflammation through activation of the trigeminal neuron system. Nitric oxide levels continue to be increased even in the headache-free period in migraineurs. In premenopausal women with migraine, particularly in those with migraine aura, increased endothelial activation, which is a component of endothelial dysfunction, is evident.
Serotonin and migraine
The serotonin receptor (5-hydroxytryptamine [5-HT]) is believed to be the most important receptor in the headache pathway. Immunohistochemical studies have detected 5-hydroxytryptamine–1D (5-HT1D) receptors in trigeminal sensory neurons, including peripheral projections to the dura and within the trigeminal nucleus caudalis (TNC) and solitary tract, while 5-HT1B receptors are present on smooth muscle cells in meningeal vessels; however, both can be found in both tissues to some extent and even in coronary vessels. All the currently available triptans (see Medication) are selective 5-HT1B/D full agonists. These agents may decrease headache by abolishing neuropeptide release in the periphery and blocking neurotransmission by acting on second-order neurons in the trigeminocervical complex.
Migraine risk factors
Predisposing vascular risk factors for migraine include the following:
- Increased levels of C-reactive protein · Increased levels of interleukins
- Increased levels of TNF-alpha and adhesion molecules (systemic inflammation markers)
- Oxidative stress and thrombosis
- Increased body weight
- High blood pressure
- Impaired insulin sensitivity
- High homocysteine levels
- Coronary heart disease
Transformed migraine/medication overuse headache
In some patients, migraine progresses to chronic migraine. Acute overuse of symptomatic medication is considered one of the most important risk factors for migraine progression. Medication overuse headache can occur with any analgesic, including acetaminophen or nonsteroidal anti-inflammatory drugs (NSAIDs), such as ibuprofen, naproxen, and aspirin. In addition, Bigal and Lipton identified the following associations of medication with progression to chronic migraine:
- Opiates – Critical dose of exposure is around 8 days per month; the effect is more pronounced in men
- Barbiturates – Critical dose of exposure is around 5 days per month; the effect is more pronounced in women
- Triptans – Migraine progression is seen only in patients with high frequency of migraine at baseline (10-14 days/mo) In the study, the effect of anti-inflammatory medications varied with headache frequency. These agents were protective in patients with fewer than 10 days of headache at baseline but induced migraine progression in patients with a high frequency of headaches at baseline.