From the Department of Neurology, Chicago Medical School at Rosalind Franklin University of Medicine and Science, North Chicago, IL, USA.
Address all correspondence to Dr. Nabih M. Ramadan, Department of Neurology, Chicago Medical School at Rosalind Franklin University of Medicine and Science, 3333 Green Bay Road, North Chicago, IL 60064, USA.
A variety of drugs from diverse pharmacological classes are in use for migraine prevention. Traditionally, they have been discovered by serendipity. Examples include β-adrenergic blockers, anticonvulsants, tricyclic antidepressants, and serotonin receptor antagonists. The mechanisms of action of migraine preventive drugs are multiple but it is postulated that they converge on two targets: (1) inhibition of cortical excitation; (2) restoring nociceptive dysmodulation. The antiepileptic drugs (e.g., topiramate, valproate, gabapentin), calcium channel blockers such as verapamil, and inhibitors of cortical spreading depression are some examples of drugs that reduce neuronal hyperexcitability. On the other hand, modulators of the serotonergic and adrenergic systems and cholinergic enhancing drugs may restore descending nociceptive inhibition and play a role in migraine prevention. To date, Level 1 evidence and clinical experience favor the use of the antidepressant amitriptyline, the anticonvulsants divalproex and topiramate, and the β-adrenergic blockers propranolol, timolol and metoprolol as first line migraine preventive drugs. The evidence for others (e.g., verapamil) is not as strong. Migraine preventive drugs have varying degrees of adverse effects, some of which could be limiting, and their efficacy should balanced with their risks of adverse effects, patients' expectations and desires, and compliance. It is hoped that future migraine preventive drugs target migraine mechanisms more specifically, which could well enhance the therapeutic index.
In assessing migraine preventive mechanisms, we propose that there are 2 primary factors involved—neuronal hyperexcitability and nociceptive dysmodulation. In theory, preventive medications may act on either of these sites by inhibiting neuronal hyperexcitation or restoring nociceptive dysmodulation.1,2 Cortical hyperexcitation with subsequent cortical spreading depression (CSD) is believed to underlie aura.
Cortical excitability has been recognized in migraine for many years, as evidenced in electrophysiological studies, carbon dioxide hyperreactivity of the blood vessels, generation of phosphene studies, or assessment of markers of oxidative stress. These model systems suggest that selected neurons in the migraine brain are more excitable, either by some ubiquitous sensory stimuli or some recognized trigger such as changes in hormone levels or flickering lights.3 The cause of neuronal hyperexcitability of the migraine brain is not well understood, but perhaps glutamate plays a significant role. Glutamate dysfunction may be caused by a genetic defect in the FHM1 gene, which results in excessive gain of function of the calcium channel of the Cav2.1 with calcium entry into the cell leading to release of glutamate into the postsynaptic cleft. Alternatively, dysfunction of glutamate uptake or lack of glutamate clearance may cause postsynaptic receptor stimulation and neuronal depolarization.4
Chronic migraine is not that much different from other chronic painful disorders, which are characterized by a triad of pain (headache in the case of migraine), depression or anxiety, and sleep disturbance.5 So, as we develop specific preventive therapies for migraine, perhaps we also should be searching for the common biological substrate or substrates of the triad (Fig. 1), and target new migraine preventive therapies accordingly.
Inhibition of Cortical Hyperexcitability.— The most recent addition to the migraine preventive medication armamentarium is topiramate. Table 1 illustrates some of the suggested mechanisms that are believed to be operant with topiramate and how it works on the mechanisms of hyperexcitability. Topiramate is believed to work on mechanisms of phosphorylation,6 with blockade of the AMPA/kainate glutamate receptor, a modulatory effect on the GABA receptor, and presynaptic sodium, and potentially calcium, channel blockade.6 The net effect of these properties is suppression of cortical neuronal excitability.
Table 1.—. Potential Mechanism for Topiramate in Blocking Neuronal Hyperexcitability
• Blocks voltage-dependent Nav
• Potentiates GABA activity
• AMPA/KA blockade
• L-, N-channel Cav blockade
Acts on mechanisms of phosphorylation
Other blockers of neuronal excitability are the gabapentinoids, which are derivatives of gabapentin. Essentially, gabapentin is believed to prevent intracellular calcium entry by interfering with the alpha-2-delta (nonpore forming) component of the calcium channel.7 In addition, the gabapentinoids are believed to modulate various glutamate transporters.
The mechanism of neuroexcitatory inhibition has been extensively studied in migraine, and there are a few new areas of development in terms of studying new calcium channel blockers for potential use in migraine prevention. New drugs have focused on the Cav2.1 and 2.2 channels that are implicated in neurotransmitter release. Also, the PQ channel is expressed throughout the brain, and is highly concentrated in the cerebellum where, at least based on the white matter lesions, it is believed to have some relevance to migraine. These PQ channels are located on presynaptic terminals and do have a regulatory role in neurotransmitter release8 and a role postsynaptically in neuronal excitability and integration. They are also believed to contribute to the mechanisms of nociceptive transmission.9 Specifically, if the PQ channel is blocked by injecting omega-agatoxin into the ventral lateral periaqueductal gray, this results in facilitation of transmission in the trigeminal nucleus caudalis. More recently, Ferrari and colleagues developed a knock-in mouse model of the FHM1 mutation on the PQ calcium channel gene.10 The animals demonstrated an increased susceptibility to CSD and also had a transient hemiparesis. This model system is exciting as it provides a way to study CSD caused by specific calcium gene mutations, which will then allow further development of new treatments specific for this mutation. It has been well documented that mutations in the calcium channel are implicated in FHM.11 The question remains as to what might be the potential role of PQ channel modulators for treatment of more common migraine?
Pietrobon and colleagues have shown us that there are slightly over 14 mutations on the calcium gene in familial hemiplegic migraine type-1.12 Some of these have been associated with familial hemiplegia only, and others have been associated with progressive cerebellar ataxia and hemiplegia. The other gene locus or FHM2 gene, the one that is associated with sodium/potassium ATPase dysfunction, initially was thought to be only associated with familial hemiplegic migraine. More recently, a mutation on the FHM2 gene has been linked to an ataxic syndrome as well.13 It is important to keep in mind that hemiplegic migraine and cerebellar disorders might have relevance when discussing Kruit's MRI study showing cerebellar lesions and cerebellar infarcts in migraineurs.14
Besides the PQ calcium channels, the L calcium channel blockers (eg, verapamil) may have some role in migraine prevention. L calcium channel blockers might work by potentiating opioid and acetaminophen analgesia.1 They also may have some role on nitric oxide inhibition. They do tend to colocalize on calcitonin gene-related peptide (CGRP) neurons, and they may have a role in blocking hyperalgesia.
Dotarizine has been around for several years, but is only recently being studied for prophylaxis in migraine.15,16 Whether its role is through an effect on the serotonin 2 receptors or a Cav blockade has not been determined. Another drug that has been studied for pain is ziconotide, but the limitation is that it is not available for oral administration (delivered intrathecally).17 It is an N channel calcium blocker; however, it is yet to be discovered if it is a molecule that can be used as a prototype for future development of more selective and more orally available medications for migraine prevention.
Another method of CSD inhibition may be through the sigma R receptor agonism. Currently available as cough syrups, dextromethorphan and carbetapentane are believed to be sigma R receptor agonists and these effects on CSD appear to be independent of N-methyl D-aspartate (NMDA) receptor activation.18,19
The other potential inhibitors of CSD are the various NR2Bs, the non-NMDA (AMPA, Kainate) ionotropic receptor blockers, and the glycine site modulators.1 These exert an effect exclusive of the NMDA blockers, like ketamine, which may help alleviate some of the side effects associated with ketamine-like and related compounds. Potassium current modulators are also CSD inhibitors and can be looked at in future studies. The chloride channel enhancers are also potential targets for the future for modulating cortical hyperexcitability.
The metabotropic glutamate receptor modulators are also of interest. Group 1 antagonists, the group 2-3 agonists, or positive modulators could be looked at for migraine prevention.1 This approach to treatment is based on the aim of reducing the load of glutamate and perhaps suppressing the cortical hyperexcitability.
Lastly, 5-HT2A antagonists may prove beneficial as they suppress sensitization and prevent chronification.20 These need to be studied further in terms of therapeutic value and potential side effects.
Nociceptive Dysmodulation.— Another approach to migraine prevention is to use pharmacotherapies that target nociceptive mechanisms. Pain is not just discriminatory, but is perceived, memorized, and expressed through multiple pathways, as studied through cortical representation within the supplementary motor area and the anterior cingulate cortex, for example. Through these different pathways, there are several ways of modulating pain beyond the basic pathways or initial pathways commonly recognized. These alternate pathways could be modulated, for example, either at the level of the cortex or at the descending level from the hypothalamus down into the trigeminal nucleus caudalis.
Along these pathways, multiple biological targets might be modulated and could prove helpful (Table 2). These include cholinergic, noradrenergic, serotonergic, 5-HT, dopaminergic, glutamatergic, and GABAergic modulation. One receptor system that is of particular interest is the cannabinoid receptor system, which has not been well studied because of its association with marijuana. However, with the discovery of the endogenous ligand for this system, anandamide, we can develop anandamide receptor uptake inhibitors or transport inhibitors allowing us to potentially stay away from the problems of the cannabinoids or the marijuana-like compounds. Cholecystokinin antagonists and galanin also are likely to play a role in nociceptive modulation and they should be kept in mind as potential targets. Others include adenosine and glycine, and various kinases or neurotrophic factors that have become increasingly studied in the last decade.
Table 2.—. Potential Targets for Pain Modulation Using Preventive Pharmacotherapies
• NE, 5-HT, DA
• Selective opioids
• VR1 (TRPV1)
• Kinases (ERK, PK)
The cholinergic system contains alpha 4, 7, and 10 nicotinic receptors that are colocalized on nociceptive neurons that are VR1 receptor positive. This suggests that there may be a potential interaction between the cholinergic, vanilloid, nicotinic, and muscarinic systems—and particularly the M2 messenger RNA is expressed on trigeminal ganglion neurons, which are CGRP positive. There has been some suggestion that the cholinesterase inhibitors may have some value in migraine prevention.21–23 At this time, it is not clear how cholinergic modulation may exert its therapeutic effects for migraine prevention.
Newer serotonergic/noradrenergic reuptake inhibitors (SNRIs), particularly mirtazapine, venlafaxine, and duloxetine, may also prove helpful in modulating descending inhibition in migraine. These drugs do have some synergistic effects with opioids and they have an enhancing role on the descending noxious inhibitory control (see Buchanan et al, for review1). There is some limited evidence that these agents might be effective, but large prospective randomized studies are needed. Additionally, it is likely that industry also will be developing newer SNRIs that have a good therapeutic ratio in terms of the side effect/efficacy potential given their role in multiple different therapeutic areas such as depression, schizophrenia, and other areas.
The new 5-HT7 modulators may potentially prevent migraine attacks, along with the 5-HT1F and 5-HT4. The 5-HT4 agonist appears promising for migraine given its therapeutic role on irritable bowel syndrome and gut motility.24
Open label studies have suggested that the atypical antipsychotics quetiapine25 and olanzapine26 may prevent migraine, but randomized controlled trials are needed to further validate this mechanism. It is not clear how these aminergic agents exert a therapeutic role in migraine.
Another approach to modulating pain in migraine may be through interference with brain energy metabolism. For example, agents that alter brain magnesium may play a role in regulating normal brain function. In the brain, magnesium has a stabilizing role on the sodium potassium pump. Dysfunction of the sodium/potassium pump may have relevance in terms of increased glutamate in the synaptic cleft. Low levels of magnesium may also be responsible for release of the NMDA receptor, which may lead to spontaneous discharges and CSD. Additionally, an increase in matrix metalloproteinases may potentially lead to a leaky blood-brain barrier, iron deposition, and nitric oxide release, which may cause additional glutamate release. Once NO diffuses out of the cell, it is a presynaptic enhancer of further glutamate release. This may be the initiation of an ongoing cycle. CoQ10 exerts its effects by modulating energy metabolism. Preliminary studies report that CoQ10 is effective for migraine prevention.27
The angiotensin-converting enzyme (ACE) inhibitor, lisinopril, has a single positive randomized control trial as a preventive therapy for migraine.28 Whether newer ACE inhibitors may have further value in migraine is yet to be determined and how they exert their effect is not known.
Another potential agent is candesartan, which is an angiotensin II type-1 inhibitor that has demonstrated efficacy in migraine prophylaxis.29 The mechanism of action of candesartan in migraine prevention is poorly understood. It presynaptically inhibits GABA release, which theoretically would enhance excitation. However, the colocalization of AT1, glutamate, and GABA receptors on medullary rostral ventromedial neurons suggest a nociceptive modulatory role.
In the literature, there are conflicting reports on the efficacy of botulinum toxin for migraine prevention.30,31 It is believed that botulinum toxin type A blocks neurotransmitter-rich vesicle attachment to presynaptic membrane and thus prevents the exocytosis of acetylcholine, and perhaps as it relates to migraine, of substance P and glutamate.32 One argument for how it might work has been proposed based on suppressing sensitization.
Petadolax™ (Weber and Weber, Germany) (also known as butterbur) appears to exert its effects through leukotriene inhibition. There are 2 published reports supporting the efficacy of Petadolex™ for migraine prevention.33–35
There is a much work to be done to further understand where these different drugs exert their pain modulation or cortical hyperexcitability modulation specifically in migraine. As shown in Table 3, there are many different areas of study still warranted, which are pharmacologic and surgical (eg, closure of patent foramen ovale). For example, the possible role of dipyridamole in migraine is being revisited. Inflammatory pain modulators, COX-3 inhibitors such as dipyrone, and leukotriene receptor selective inhibitors, may be antimigrainous, particularly if migraine pain proves inflammatory in origin. Psychostimulant drugs, such as dextroamphetamine, are unexplored, and the SNRIs need further study specifically in migraine. New anticonvulsants may prove even more useful than what we see with currently available therapies such as divalproex sodium and topiramate. Lastly, the newer and more controversial surgical approaches are beginning to find their way into the migraine preventive literature. Vagal nerve stimulation and right-to-left shunt (RLS) cardiac repair are both controversial as preventive treatments for migraine.36 Small, uncontrolled studies have suggested that RLS closure prevents migraine attacks and, perhaps migraine infarction. This potential therapy needs further scientific research before advocating it. Also, it could be argued that cardiac defects causing RLS and migraine coexist on a common genetic predisposition36 instead of implying that RLS increases the risk of migrainous infarction.
Table 3.—. Areas Needed for Further Study in Migraine Prevention
Over the next decade, the results from further clinical studies in the exploratory areas described in this article will provide the foundation for new evidenced-based medicine that will guide clinical practice. As reviewed herein, future targets and areas for further study are plenty.
Conflict of Interest: Nabih M. Ramadan, M.D., Colucid Pharmaceuticals, Inc.; Eisai Inc.; Endo Pharmaceuticals Inc.; GlaxoSmithKline; MAP Pharmaceuticals, Inc.; Merck & Co., Inc.; Ortho-McNeil Neurologics, Inc.