Migraine is a complex disorder of the brain whose mechanisms are only now being unravelled.1 It is common,2 disabling,3 and economically costly.4 Elsewhere in this supplement, the specific burden and cost implications for frequent, medically intractable migraine are covered. Here, we shall attempt to address some possible mechanisms for treatment refractory migraine. Some definitions will need to be arbitrarily adopted to move forward in what is an area that has not been formally addressed in the past. First, we recognized the definitions of the International Headache Society for migraine5 and for chronic migraine.6 However, for this exercise, the rigid use of the 15-day cut-off probably makes no sense, and we will try to address whether frequency of headache is an issue for tractability for treatment. Second, the definition of refractory to treatment is here taken to mean medical treatment. Definitions for this have been suggested,7 although for our purpose the precise definition of, for example, the number of failed preventives is less important than the general principle that some patients with clinically indistinguishable phenotypes seem refractory to medical management. Third, we assume for this discussion that the most common reasons for apparent intractability, diagnostic errors, and inappropriate use of normally effective medicines,8 are not at issue. Last, we will set aside a natural history argument that many patients would improve or become responsive to therapy if given time. This may be true but is not helpful as a blanket explanation.
Despite increased understanding of primary headaches and their treatment, the underlying causes of refractory migraine remain unknown. This note considers potential genetic, structural, functional and pharmacological factors that could contribute to this relatively intractable condition. Further understanding of refractory migraine will require the use of medical imaging technologies, clinical experimental medicine studies on novel pharmacological agents and astute observations in clinical practice to direct potential novel therapeutic approaches.
IS THERE A GENETIC BASIS FOR MEDICALLY INTRACTABLE MIGRAINE?
Migraine is a substantially inherited condition. It is clear from clinical practice that many patients have first degree relatives who also suffer from migraine,9,10 and numerous published studies have reported a positive family history.11 An explained but epidemiologically well-established predisposition relates to methyltetrahydrofolate reductase gene mutation C677T, which is certainly overrepresented in migraine with aura.12 The specific genetic cause of familial hemiplegic migraine (FHM) has been identified in a number of cases. Mutations involving the Cav2.1 (P/Q) type voltage-gated calcium channel CACNA1A gene are causative in FHM-I.13 Mutations in the ATP1A2 gene14,15 have been identified to be responsible for FHM-II. The gene codes for a Na+/K+ ATPase and the mutation results in a smaller electrochemical gradient for Na+. One effect of this change is to reduce or inactivate astrocytic glutamate transporters leading to a build-up of synaptic glutamate. Dichgans et al16 reported a missense mutation (Q1489K) in SCN1A in 3 German families, thus characterizing the genetic defect of what is now known as FHM-III. This mutation affects a highly conserved amino acid in a part of the channel that contributes to its rapid closure after opening in response to membrane depolarization (fast inactivation). This represents a gain of function: instead of the channel rapidly closing, allowing the membrane to repolarize fully after an action potential, the mutated channel allows a persistent sodium influx. While devastating, FHM is not more troublesome to treat than any other form of migraine, often responding to the calcium channel/dopamine blocker flunarizine. It is very attractive to consider a genetic basis for treatment refractory migraine, but one has yet to emerge. The physiological consequences of the FHM mutations are only now being explored. Certainly, there is reduced threshold to initiation of cortical spreading depression (CSD) in the mouse;17 however, since most antimigraine preventives have actions to reduce susceptibility to CSD,18 a CSD-based mechanism for intractability is unlikely. Changes in neurotransmitter expression, specifically calcitonin gene-related peptide in trigeminal ganglion of animals with an FHM-I mutation, have also been reported.19 Perhaps there is a paradigm here for starting to understand how genetic change may alter treatment responses especially as CGRP antagonists are now being developed as potential anti-migraine therapies20,21,22 Population-based family studies in migraine probands that had either disabling or very painful migraine showed higher levels of family aggregation of migraine.23 Interestingly, in the more recent study, the early age of onset of migraine in the proband was also associated with increased family aggregation and increased severity.24 Since migraine is a complex multi-factorial genetic disorder25 it could be speculated that probands with more severe pain, more disability or an earlier age of onset may have a greater genetic load and perhaps greater risk of being refractory but this has yet to be proven. Treatment refractory migraine is certainly a subset of all migraine suggesting that there must be an intersecting problem of migraine genetics plus something else than confers refractoriness.
IS THERE A STRUCTURAL EXPLANATION?
One possible explanation for treatment refractory migraine would be structural changes in the brain. Raskin26 initially reported, and Veloso confirmed,27 that stimulation of the periaqueductal gray region can produce migraine in otherwise nonheadache people. Similarly, Welch et al28 noted excess iron in the periaqueductal gray (PAG) of patients with episodic and chronic migraine. Indeed, chronic migraine can develop after a bleed into a cavernoma in the region of the PAG,29 or with a lesion of the pons.30,31 Functional brain imaging with positron emission tomography (PET) similarly points to the brainstem by the demonstration of activation of the dorsal midbrain, including the PAG, and the dorsal pons, near the locus coeruleus, in studies during migraine without aura.32 Dorsolateral pontine activation is seen with PET in spontaneous episodic33-35 and chronic migraine,36 and with nitrogylcerin-triggered attacks.37,38 In patients with chronic migraine successfully treated with occipital nerve stimulation, who had been previously refractory to medical treatment, these brain areas were still active.36 It has been speculated, with support from preclinical studies,39,40 that extra-cranial disease at the back of the head or upper neck may make headaches refractory through sensitizing neurons in the trigeminocervical complex that receive convergent input from both the trigeminal and occipital nerves. Persistent input from extra-cranial sources nerves enhances the sensitivity to input from intracranial pain producing structures thereby potentially exacerbating headache. Anecdotal reports have suggested that addressing these extra-cranial structural problems may, for some people, be the key to effective treatment.
Is one plausible explanation for medically refractory migraine structural changes in the brain stem? If structural changes were an explanation, one might expect to see it on brain imaging. Careful comparisons of migraine and nonmigraine brains using voxel-based morphometry initially failed to demonstrate changes.41 Subtle changes in the visual processing network have been now shown in migraine with and without aura.42 However, there is no suggestion that these change with time or are related strictly to attack load or any other measure of disease burden,43 so they may be an anatomical trait. This is likely to be the case with the relationship between migraine with aura and patent foramen ovale.44 The only population-based data to suggest a change in brain structure show posterior fossa changes, pons, and cerebellum only in migraine with aura.45 It is not yet clear what the nature of these changes is, or indeed whether they are transient.46 From a clinical viewpoint, these changes are unlikely to explain medical intractability, which is much more commonly an issue in migraine without aura than migraine with aura.
IS THERE A SYSTEMS FUNCTIONAL EXPLANATION?
Central nervous system (CNS) plasticity changes are an attractive way to conceptualize medically intractable migraine. Clearly, changes in brain function, without overt lesions, can be seen on PET studies of chronic migraine. There is persistence of pontine activation without any change on magnetic resonance imaging.36 One common theme in medically intractable migraine is the coexistence of medication overuse.47
Medication overuse is to be found in about 1% of Western populations, so it is very common as a problem.48 Medication overuse headache can be seen in patients using analgesics for other reasons, such as in rheumatology49 or gastrointestinal indications,50 or indeed in patients with cluster headache.51 There are 2 common themes here. First, not all patients overusing analgesics will develop medication overuse; indeed, perhaps only about one-third do so. Second, almost all patients who develop the problem in these settings have a personal or family history, or both, of migraine or disabling headache. Some patients do extremely well when medication overuse is controlled, although many do not.52,53 Furthermore, it has generally been held that medication overuse plays a substantial role in medical intractability such that medication withdrawal is recommended as part of the management of frequent migraine. However, it is clear from randomized placebo-controlled trials that this is not necessarily always correct, since topiramate is effective in a proportion of patients with chronic migraine and medication overuse.54,55
There is certainly a trait in migraine that can be demonstrated interparoxysmally using electrophysiological methods. Visual,56 auditory,57 and event-related58 potentials are each exaggerated between attacks, suggesting a lack of habituation that normalizes during attacks59 and can be altered by standard migraine preventives.60 Remarkably, patients without migraine but with a family history can also be shown to have impaired habituation of the nociception-specific blink reflex, behaving like migraineurs rather than controls.61 A further marker of importance may be found in changes in orbitofrontal cortex 18flurodeoxyglucose PET activity in patients with medication overuse headache.62
Taken together, there is considerable emerging evidence for CNS dysfunction in migraine; however, none specifically has been implicated in medical intractability. The tools are available and electrophysiological63 and functional imaging64 studies in cohorts of relatively tractable and intractable patients may yield insights into potential brain regions that influence tractability.
IS THERE A PHARMACOLOGICAL EXPLANATION?
A seemingly obvious possibility for medically intractable migraine would be a pharmacological explanation. This must exclude a simply pharmacokinetic argument, as one can always augment absorption with antiemetics and pro-kinetics, or indeed turn to nonoral routes. Few migraineurs do not respond to sumatriptan 6 mg subcutaneous,65 and in clinical practice they are neither phenotypically obvious, nor necessarily those that do not respond to preventive medicines. Is there any rational pharmacological explanation?
One might have hoped that the era of molecular genetics and molecular pharmacology would have yielded insights into medically intractable migraine, but results are thus far disappointing. Receptor polymorphisms for the serotonin (5-HT) receptors that mediate the actions of triptans, as best that is understood,66 such as the 5-HT1B receptor,67,68 are unrelated to clinical responses to sumatriptan. 5-HT1F receptor genetics69 also seem unrelated to antimigraine activity. Perhaps genes that mark a good response, for example it has been suggested that particular dopamine receptor 2 forms may be important for rizatriptan,70 will when absent, indicate a poor response; but such work is yet to be done. Perhaps understanding how genes such as those in dopamine pathways that have been associated with increased severity71 or protection from migraine72 intersect with those increasing risk of migraine will hold the key to understanding the basis of refractoriness.
From a treatment perspective it will be interesting to see whether the CGRP receptor antagonists, a new class of anti-migraine drugs currently in clinical investigations21,22 have any impact on refractory migraine and if so this might help define its likely pathophysiology if not underlying cause. From the preventive literature, there is little as yet to be gleaned. Chronic migraine has barely been studied, with the notable exception of topiramate.54,55,73 While these studies suggest that patients previously considered difficult to treat are tractable, they were not designed to give insights into nonresponsive subjects.
Basic science studies of potential mechanisms in migraine have shown a high potential for sensitizing changes within trigeminal sensory pathways74,75 that could influence headache severity and associated sensory symptoms. Cutaneous allodynia has been proposed as a clinical marker of central sensitization and its presence appears associated with frequency, severity and disability of migraine in over 60% of migraineurs76 and also with therapeutic response.77 It therefore seems remarkable that the treatment of migraine is not more intractable and that most patients seen in practice remain, to some degree, responsive to therapy. However, despite increased understanding of the mechanisms involved in idiopathic primary headache and its therapy, the condition of refractory migraine remains somewhat of an enigma and there are few clinical treatment options. the potential pathophysiological mechanisms in migraine have become interwoven into a sophisticated integrated hypothesis that encompasses initiating primary CNS dysfunction, trigeminovascular activation, central sensitization and heightened pain sensation. This framework provided the platform for the discovery of novel anti-migraine medications and non-pharmacological therapeutic strategies and may help unravel the functional basis of refractory migraine. There are also many exciting new technological advances that may enable us to understand refractory migraine better and place it within or outside these current theories of migraine pathogenesis. Neuroimaging studies, for example, may point to the CNS structures, transmitter systems and circuits involved in refractory migraine to provide a rational basis for strategies to improve clinical care.
Indeed, it has recently been suggested that chronic pain may change the gray matter structure of the brain and that patients may have a common brain “signature” in regions known to be involved in pain regulation.78
Further understanding will also certainly require a combination of astute clinical observations and clinical experimental medicine studies to direct potential novel therapeutic approaches. Refractory migraine is another challenge for the science and treatment of headache that warrants increased research.