New insights into headache: an update on functional and structural imaging findings

nature reviews | neurology volume 5 | aPril 2009 | 199

reviews

Department of systems Neuroscience, Universitäts-Krankenhaus eppendorf (UKe), Hamburg, Germany.

Correspondence: Department of systems Neuroscience, Universitäts-Krankenhaus eppendorf (UKe), Martinistrasse 52, D-20246 Hamburg, Germany a.may@ uke.uni-hamburg.de

new insights into headache: an update on functional and structural imaging findingsArne May

Abstract | One of the most exciting developments in modern neuroscience has been the application of imaging techniques to provide new insights into the organization of the human brain in vivo. Functional imaging methods, such as PeT and functional Mri, have become the preferred techniques for detection of the structure–function relationships within the brain that are characteristic of headache. This review focuses on neuroimaging as a diagnostic tool for headache and highlights the advances made with functional and structural neuroimaging techniques in the study of primary headache syndromes such as migraine and trigeminal autonomic headaches. several independent functional studies have reinforced the crucial role of the brainstem in acute and chronic migraine and of the hypothalamic area in trigeminal autonomic headaches. structural abnormalities that have been identified in the visual network of motion-processing areas could account for, or result from, the cortical hyperexcitability observed in patients with migraine. several morphometric studies suggest that gray matter volume and/or concentration is decreased in pain-transmitting areas in patients with migraine or tension-type headache. Given the rapid advances in functional neuroimaging, this technique will continue to be of paramount importance in patients with headache and might ultimately serve as the bridge between molecular and clinical domains in headache research.

May, A. Nat. Rev. Neurol. 5, 199–209 (2009); doi:10.1038/nrneurol.2009.28

IntroductionFunctional neuroimaging has provided unique insights into some of the most common maladies experienced by man and has revolutionized our concept of the patho­physiology of primary headache syndromes such as migraine and cluster headache. while all the early studies in headache were done exclusively with Pet, the develop­ment of novel computational techniques during the past few years and increases in image resolu tion have led to an expansion in the applications of mri­based imaging methods in headache. Both Pet and mri­based methods have now been widely used in headache and migraine research. these techniques have provided invaluable information on brain per fusion and metabolism during and outside of headache attacks and contribute to an improved understanding of the pathophysiology of these disorders. next to these imaging methods that enable us to investigate the function of the brain during headache attacks, several structural and morphometric methods have also become available. in contrast to traditional anato mical and pathological methods, magnetic reso­nance (mr) morphometry of the brain permits in vivo study of temporal changes in brain morphology and the correlation of brain morphology with brain function. mr morphometry has, thereby, recently emerged as one of the most promising fields in clinical neuroscience. the weak signal­to­noise ratio of both functional and morpho metric imaging means that the data provided by these techniques

are of interest for the study of groups but not of indivi­duals. nevertheless, human models of headache attacks are indispensable tools in patho physiologic and thera­peutic research, and new findings and methodo logical develop ments give hope of scientific breakthroughs that will change how we perceive the brain.

this review covers three features of neuroimaging in patients with headache. the first part concerns one of the most important questions for clinicians: does this patient require neuroimaging? recent meta­analyses have iden­tified several individual clinical features that are associ­ated with an intracranial abnormality, and individuals with these features should undergo neuroimaging. the second part of this review concentrates on the knowledge that has been acquired on the patho physiology of specific headache syndromes from functional imaging studies. our understanding of these syndromes has profited immensely from such studies, a point best illustrated by the successful translation of the findings into the perfor­mance of deep brain stimulation in cluster headache and sunCt (short­lasting, unilateral, neuralgiform head­ache attacks with conjunctival injection and tearing). the third part of the article focuses on the relatively new field of morphometric imaging. Further to the detec­tion of structural differences in gray and white matter in several headache and pain syndromes, mr­based morphometry has recently generated the very important finding that the brain can alter its shape within a matter of weeks, which indicates that this organ can structurally adapt, within a short time period, to physical and mental

Competing interestsThe author declared no competing interests.

© 2009 Macmillan Publishers Limited. All rights reserved

200 | APRIL 2009 | voLume 5 www.nature.com/nrneurol

reviews

activity. Consequently, mr morphometry promises to be a power ful method with which to study headache and to track the effects of novel therapies.

Diagnostic imagingno single instrumental examination has yet been able to define or ensure a correct diagnosis of headache, or differentiate between idiopathic headache syndromes. However, clinicians recognize that patients with migraine display irregularities in cranial vessels on Doppler ultra­sonography and in eeG readings between and during attacks, as well as, occasionally, nonspecific white matter changes on mri scans. these white matter changes have been linked with an increased risk of brain lesions in patients with migraine.1–3 although the question of whether these white matter changes are true vascular infarcts is an interesting one,4,5 the changes have been shown to be independent of right­to­left shunts6 and cannot, therefore, simply be attributed to the occur­rence of a patent foramen ovale.7,8 the interpretation of occasional findings of such white matter changes in patients with migraine is a clinical challenge, but, given that functional correlates are completely lacking, these changes might be a marker of future stroke or of future develop ment of chronic daily headache.3 longitudinal studies are warranted to assess whether these lesions are progressive and need the attention of clinicians.

the use of neuroimaging in patients with headache varies widely within the clinical setting. in 2004, a european Federation of neurological societies (eFns) task force used evidence from the literature to evalu­ate (among other instrumental examination tools) the usefulness of various imaging methods in patients with nonacute headache.9 according to the resulting recommen dations, the routine use of neuroimaging is not warranted in adult and pediatric patients with migraine who have experienced no recent change in attack pattern, have no history of seizures, and have no other focal neurological signs or symptoms.10 Pet and

Key points

routine use of neuroimaging is not warranted in adults or children with a primary ■headache syndrome without recent change in attack pattern, history of seizures, other focal neurological signs or symptoms

An exception to this rule should be made in the diagnosis of trigeminal ■autonomic headaches and headaches that are aggravated by exertion or a valsalva-like maneuver

The dilatation of vessels that has been observed in trigeminal pain is not, ■as has previously been implied, inherent to a specific headache syndrome

several independent studies have reinforced the crucial role of the brainstem in ■acute and chronic migraine

The activation of the hypothalamus is highly specific for trigeminal autonomic ■headaches, and imaging data from patients with cluster headache prompted successful deep brain stimulation of this area

Data from morphometric studies suggest that patients with migraine and those ■with tension-type headache have a decrease in gray matter volume in pain-transmitting areas, as a consequence of frequent pain

functional mri are rated by the task force to be of little or no value in the clinical setting but are believed to have vast potential to aid exploration of the patho physiology of headaches and the effects of pharmacological treat­ment.9 the guidelines state, however, that mri may be warranted in patients with atypical headache patterns, a history of seizures, and/or focal neurological signs or symptoms. exceptions to these rules should be made in the diag nosis of trigeminal autonomic headaches and headaches that are aggravated by exertion or a valsalva­like mane uver.11 a number of indications exist for the undertaking of neuroimaging in patients with headache. a 2007 case series demonstrated that even cluster head­ache with a typical time pattern and an excellent response to typical treatment can still be caused by underlying structural pathology such as a pituitary tumor and that such patients should, therefore, undergo neuro imaging.12 Further ‘red flags’ are patients who present with head­ache and new abnormal findings (e.g. focal deficit, altered mental status, or altered cognitive function) in a standard neuro logical examination. these patients, as well as individuals with a new, sudden­onset, severe headache, and Hiv­positive patients with a new type of headache, should be considered for neuro imaging.13 no evidence exists that elderly patients who experi­ence headache but have normal findings in a neuro­logical examination should undergo neuro imaging, but considera tion of at least a Ct scan should be given when patients who are older than 50 years present with a first or a new type of headache.10,13

Functional imaging—experimental headacheBefore the potential influence of functional studies in primary headache syndromes can be fully understood, the pattern of activation in these disorders needs to be established in experimental models of headache. Consequently, several neuroimaging investigations have been conducted in such models. in a Pet study in experimental head pain, seven healthy male volun­teers without a history of headache were investigated during an acute pain state evoked by injection of a small amount of capsaicin subcutaneously into the forehead.14 During the acute pain state, as compared with the rest state, increases in regional cerebral blood flow (rCBF) as an index for synaptic activation were found in the anterior insula bilaterally, in the contra lateral thalamus, in the ipsilateral anterior cingulate cortex, and in the cerebel lum bilaterally. these areas are all well known from several studies in experimental pain in humans and are described further in other reviews.15,16 when functional mri was used to measure Cns activation in experimental trigeminal pain, the trigeminal ganglion and spinal trigeminal nucleus were shown to be acti­vated.17,18 Figure 1 outlines the regions that are generally shown to be activated in functional imaging studies of pain and headache—the so­called pain matrix.

in addition to the activations observed in nonspecific structures associated with pain transmission following

© 2009 Macmillan Publishers Limited. All rights reserved

nature reviews | neurology volume 5 | aPril 2009 | 201

reviews

capsaicin injection,14 a bilateral pattern of activation in midline structures over several planes, slightly lateralized to the left, anterior to the brainstem and posterior to the chiasmatic region, has been shown after injection of this compound.14,19 superimposed on an mri template, the location of these areas of activation covers intra cranial arteries as well as the region of the cavernous sinus bi laterally. a strong activation was observed in the same region (the cavernous sinus) in a study in patients with cluster headache,20 which suggests that a vaso dilatation occurs with cluster headache that is mediated by the ophthalmic division of the trigeminovascular system.

in another experimental study in healthy volunteers that used mr angiographic techniques, injection of capsai cin into the skin that is innervated by the oph­thal mic (first) division of the trigeminal nerve elicited an increase in vascular diameter of the internal carotid artery when compared with the mean baseline value.21 injec tion of capsaicin into the leg, and into the skin of the chin to stimulate the mandibular (third) division of the trigeminal nerve, led to a similar pain perception but failed to produce any significant change in vessel caliber. these data suggest that a highly functionally organized, somatotopically congruent trigeminal innervation of the cranial vessels occurs, with a potent vaso dilator effect of the ophthalmic division on the large intracranial vessels.

taken together, the data from these studies suggest that neurovascular activation in the trigeminal system is a function of its afferent role in any form of pain, and that this activation is highly potent and somatotopically organized. Pain signals in the ophthalmic division can generate vascular change de novo without the occurrence of a superimposed primary headache.22 these findings are consistent with the notion that pain triggers changes in vessel caliber in migraine and cluster headache, and not vice versa; these conditions should, therefore, be regarded as primary neurovascular headaches and not as vascular headaches.

Functional imaging of headacheinsights into the fundamental physiology of primary head ache syndromes such as migraine and cluster head ache have been limited by the lack of methods with which to visualize the pathophysiological background of the trigeminovascular system and to examine its source. over the past few years, however, remarkable efforts have been made in the study of the trigemino vascular system with functional imaging. the findings from these investigations demand renewed consideration of the neural influences at work in many primary headache syndromes. However, the studies have been performed only for episodic headache types; no such investigations have so far been conducted in chronic headache—such as tension­type headache or chronic migraine—because of methodological issues. the following sections review the findings from functional imaging studies in migraine and in several types of trigeminal autonomic headaches and examine the current knowledge in these areas. the

functional data cited within the text are summarized in Figure 2.

MigraineThe clinical picturemigraine is an idiopathic headache disorder that is charac terized by moderate to severe, often unilateral and pulsating, headache attacks that are typically aggravated by physical activity. the individual attacks are accompa­nied by a loss of appetite (almost always), nausea (80% of cases), vomiting (40–50%), light sensi tivity (photo phobia; 60%), noise sensitivity (phonophobia; 50%) and odor hypersensitivity (10%). if the headaches are unilateral, they can change sides within an attack, or from attack to attack. at least five attacks must have occurred before a diagnosis of migraine can be established.23 the frequency and duration of migraine attacks varies widely between individuals; the median frequency is around one attack per month, and the median duration is roughly 24 h, with a range of 4–72 h. in 15–30% of cases, the migraine head­ache is preceded by a recurrent dis order that manifests as attacks of reversible focal neuro logical symptoms, which usually develop gradually over a period of 5–20 min and last for less than 60 min. this neurological sign, known as

PAG

Amygdala

PFC

SMA

S1

Thalamus

ACCPPC

Insula

S2

Figure 1 | The central network involved in the transmission of nociceptive input. This so-called pain matrix mainly involves the thalamus, the amygdala, the insula cortex, the supplementary motor area, the posterior parietal cortex, the prefrontal cortex, the cingulate cortex, the periaqueductal gray, the basal ganglia and cerebellar cortex (not shown), and the primary and secondary sensory cortex. Abbreviations: ACC, cingulate cortex; insula, insula cortex; PAG, periaqueductal gray; PFC, prefrontal cortex; PPC, posterior parietal cortex; s1, primary sensory cortex; s2, secondary sensory cortex; sMA, supplementary motor area. with kind permission from springer science + Business Media: Neurological sciences, Neuroimaging: visualising the brain in pain, volume 28, 2007, May.

© 2009 Macmillan Publishers Limited. All rights reserved

202 | APRIL 2009 | voLume 5 www.nature.com/nrneurol

reviews

an aura, consists mostly of visual phenomena—typically, jagged zig­zag lines that move slowly across the visual field, often followed by visual loss (scotoma). Cortical spreading depression has been suggested to underlie the migraine visual aura, on the basis of a slow spread of clinical and electrophysiological events observed in animal experiments.

The headache in migraineseveral independent functional imaging studies have reinforced the fact that the brainstem has a crucial role in acute, and probably also in chronic, migraine. the

common ground between these studies is a consis­tently observed increase in rCBF in the rostral brain­stem during migraine attacks, that persisted even after sumatriptan had induced complete relief from headache, nausea, phono phobia and photophobia.24 the increase was not seen outside the attacks and has been confirmed in a single case study, which further refined the observed activation to the dorsal rostral pons.25 Dysfunction of the regulation of brainstem nuclei involved in anti­nociception, extracerebral and intra cerebral vascular control and sensory gating provides an explanation for many of the facets of migraine. the importance of the

Mid cingulate cortex (6, 36, 11)68,70,75

Anterior cingulate cortex (6, 22, 23)57,71

Anterior cingulate cortex (6, 29, 18)34

Anterior cingulate cortex (6, –5, 29)20,57,70

Thalamus (6, –13, 5)20,57,68

Thalamus (6, –9, –2)20

Hypothalamus (6, –14, –8)57,71

Midbrain (6, –19, –27)35

Vessels (6, 10, –20)19,22

Lower pons (6, –30, –45)68,75

a

bLeft mid cingulate cortex (–3, 0, 48)82,85

Mid frontal gyrus (–3, 38, 39)84

Anterior cingulate cortex (–3, 35, 21)80,82,85

Anterior cingulate cortex (–3, 29, 14)81

Anterior cingulate cortex (–3, 40, 0)85

Anterior cingulate cortex, rostral part (–3, 22, –3)80

Hypothalamus (–3, –17, –12)78

Dorsal rostral pons (–3, –27, –25)85

Ventral pons (–3, –21, –42)85

Figure 2 | summary of the functional and structural data from all studies cited in the text. a | Functional (PeT and functional Mri) data. b | structural (voxel-based morphometry) data. For each paper, all available stereotactic coordinates were included in a meta-analysis (http://www.brainmap.org/index.html), using the GingerALe application, via the ALe method. we used a 10 mm full width at half-maximum filter and 5,000 permutations to determine the null distribution of the ALe statistic at each voxel. we used a normalized template as the anatomical underlay and the thresholded ALe results as the overlay. The coordinates (brackets) show the finding for each headache type in Talairach space. The papers that report significant activations, and form the basis for this analysis, are cited in superscript. Blue indicates trigeminal autonomic headaches; red indicates migraine; and yellow indicates tension-type headache. Abbreviation: ALe, activation likelihood estimation.

© 2009 Macmillan Publishers Limited. All rights reserved

nature reviews | neurology volume 5 | aPril 2009 | 203

reviews

brainstem in the genesis of migraine is further under­lined by reports of patients without headache who devel­oped migraine­like episodes after stereo tactic placement of electrodes in the periaqueductal gray for treatment of chronic pain.26,27 Certainly, brainstem activation per se has been reported in many pain conditions other than migraine, including tonic cold stimulation,28 laser­induced pain,29 painful touch produced by a stylus,30 and even the experience of empathy when someone else is suffering pain.31 However, the brainstem activation seen in almost all of these studies seems to be a caudal extension of periaqueductal gray activation rather than a discrete area of pontine activation as seen in migraine.

Goadsby and co­workers reinforced the view that migraine is a subcortical disorder with substantial brain stem involvement, by their study of five patients who underwent H2

15o­labeled Pet during a sponta­neous migraine attack, which provided evidence of dorsal pontine activation in migraine.32 Furthermore, this same group defined the laterality of brainstem acti­vation by use of H2

15o­labeled Pet in 24 patients during glyceryl­trinitrate ­induced migraine attacks. the group demonstrated the occurrence of ipsilateral activation in the dorsal pons during strictly unilateral migraine attacks, whereas they observed a bilateral activation of this area in patients who had bilateral headaches.33 in addition, eight patients with chronic migraine (>15 days per month of attacks of migraine without aura) who had shown a marked positive response to implanted bi lateral sub occipital stimulators were studied with Pet.34 Comparison of stimulation (improved headache) with no stimulation (headache) demonstrated signifi­cant changes in rCBF in the dorsal rostral pons, ante­rior cingulate cortex and cuneus, which correlated with pain scores.34 the localization and persistence of activity during stimula tion was exactly consistent with the activ­ity in the dorsal pontine region previously observed in episodic migraine and suggests a crucial role for this structure in the pathophysiology of chronic migraine.

a further paper reported the findings from a Pet study in seven individuals with migraine.35 the authors reported significant activations during migraine attacks not only in the midbrain and pons but also in the hypo­thalamus, which activations—just like the brainstem activation observed in the first study24—persisted after headache relief with sumatriptan. specific hypothalamic activation had been reported in the trigeminal autonomic cephalgias (taCs; see below)20 but had not previously been observed in migraine. a major limitation of the Pet study, however, is that it did not have a control group and the results are, therefore, potentially confounded by order and session effects. this issue is crucial, because in a sub­sequent study in which an appropriate control group was included, brain activation in other pain states—such as hypothalamic activation with cardiac pain—was shown to be the result of an order effect.36

only a few studies have been published on the use of ligand Pet in migraine. two investigations have

explored the brain distribution of 5­hydroxytryptamine (5­Ht) receptors in patients with migraine. the findings suggested that these patients have increased 5­Ht1a­receptor density or decreased endogenous 5­Ht levels, which could explain the much discussed altered corti­cal excitability of individuals with migraine.37 with regard to 5­Ht2, one study found that this receptor was upregulated in patients with migraine compared with controls,38 whereas another study found no difference between patients with migraine and controls.39 with use of Pet, and α­[11C]methyl­ l­tryptophan as a surrogate marker of brain 5­Ht synthetic rate, 5­Ht synthesis in the brain of patients who experience migraine has been demonstrated to be highest during migraine attacks, lowest immediately after administration of sumatrip­tan, and intermediate when patients are migraine free.40 However, rates of 5­Ht synthesis did not differ between patients during migraine attacks and age­matched and sex­matched controls.40 unlike in cluster headache (as discussed later41), no opioid ligand studies have been published in migraine. the challenge now is to reveal the functional consequences of such findings as discussed above, to understand their implications, and to assess their therapeutic potential.42

Aura and the motion-processing visual networkthe pioneering work of olesen and collegues with single­photon emission Ct (sPeCt) revealed a focal reduction in rCBF during migraine attacks with aura, usually in the posterior parts of one hemisphere.43,44 these data have been reproduced and are robust. no blood flow changes were noticed by olesen’s group with sPeCt in migraine without aura.45 since this pioneering work, several studies have used different techniques to demonstrate gross changes in cortical perfusion during migraine, in an attempt to explain either the aura or the headache in migraine.46–48 However, the nature of the link between aura and headache remains controversial, despite some valuable theoretical considerations.49 the origin of the aura and general susceptibility to migraine attacks seems to lie in the cerebral cortex, but it must be remembered that only 15–30% of individuals with migraine have aura,50 and the rest do not show any consistent demonstrable changes in blood flow.

a study by Granziera et al. used surface­based func­tional imaging analysis methods, which included measure ment of cortical thickness and diffusion tensor imaging, to examine the motion­processing visual network in 24 patients with migraine and 15 age­matched individuals without migraine.51 the study found that patients with migraine had an increase in cortical thickness in motion­processing visual areas. the thick­ness abnormalities identified were in area v3a, an area that was previously described to be a source of spread­ing changes involved in visual aura.48 in addition, in the Granziera study, diffusion tensor imaging revealed that patients with migraine have alterations in the superior colliculus and the lateral geniculate nucleus, which areas

© 2009 Macmillan Publishers Limited. All rights reserved

204 | APRIL 2009 | voLume 5 www.nature.com/nrneurol

reviews

are also involved in visual processing.51 the authors con­cluded that a structural abnormality in the network of motion­processing areas could account for, or be the result of, the cortical hyperexcitability that is observed in individuals with migraine. in summary, an inherited suspectibility for migraine seems to be responsible for a developmental change that leads to the structural dif­ferences in these areas, irrespective of the occurrence of aura symptoms.52 Further studies to investigate migraine attacks are clearly needed as soon as possible to confirm these results.

Medication overuse headachePatients with migraine are physiologically and perhaps psychologically hyperresponsive to a variety of internal and external stimuli.23 one such trigger is the regular intake of pain medication, which can lead to a daily or near­daily headache (>15 days per month). the mecha­nism by which medication overuse transforms episodic migraine into chronic daily headache is unknown. a recent Pet study measured glucose metabolism before and 3 weeks after medication withdrawal in 16 patients with chronic migraine who had analgesic overuse and compared these findings to those in healthy volun­teers.53 Before medication withdrawal, the bilateral thalamus, orbitofrontal cortex, anterior cingulate gyrus, insula–ventral striatum and right inferior parietal lobule were hypometabolic in these patients. all of these areas recovered to almost normal levels of glucose uptake after withdrawal of analgesics, with the exception of the orbito frontal cortex. the authors concluded that medica­tion overuse headache might be associated with rever­sible metabolic changes in pain­processing structures, but also with persistent orbitofrontal hypofunction. the latter process is known to occur in drug dependence and could predispose subgroups of individuals with migraine to recurrent analgesic overuse. whether these findings are indeed specific to medication overuse headache is a matter of debate. another study that used the same method but investigated cluster headache without medi­cation overuse found that ‘in bout’ compared with ‘out of bout’ scans showed evidence of increases in metabolism in the perigenual anterior cingulate cortex, posterior cingulate cortex, prefrontal cortex, insula, thalamus and temporal cortex. Compared with the brain metabolism in healthy volunteers, hypometabolism was found in the patients (irrespective of their bout status) in the peri­genual anterior cingulate cortex, prefrontal cortex and orbitofrontal cortex.54 Further investigations are required before the functional consequences of such findings can be discussed.

Trigeminal autonomic headachesThe clinical picturethe taCs are a group of primary headache disorders characterized by unilateral trigeminal distribution of pain that occurs in association with ipsilateral cranial autonomic features. the taCs include cluster headache,

paroxysmal hemicrania, and sunCt.23 Cluster head­ache is defined as a paroxysmal, strictly unilateral, very severe headache, in which most of the pain is typically focused in the retro­orbital area. the unilateral auto­nomic symptoms such as ptosis, miosis, lacrimation, conjunctival injection, rhinorrhea and nasal congestion that are present only during pain attacks are ipsilateral to the pain, indicative of parasympathetic hyperactivity and sympathetic impairment. the most remarkable clinical feature of cluster headache is the striking rhythmicity or cycling of the attacks and bouts. in paroxysmal hemi­crania, the headache attacks, character and localization of the pain, and autonomic symptoms are very similar to those observed in cluster headache; however, the attacks are shorter (2–30 min) and more frequent (>5 attacks per day). the autonomic symptoms are often less severe in paroxysmal hemicrania than in cluster headache. the full name of the third taC, short­lasting unilateral neural­giform headache attacks with conjunctival injection and tearing (sunCt), describes the typical clinical features of this syndrome. sunCt is characterized by very short (5–240 s) attacks with neuralgiform pain quality and severe intensity. the attacks occur at a frequency of 60 per day on average (range 3–200 per day), are strictly uni­lateral (periorbital), and are often triggered by touching, speaking, or chewing. taCs are relatively rare in com­parison with migraine or tension­type headache, which is probably why taCs are poorly recognized in primary care. neuroimaging has made substantial contributions in recent times to understanding of this relatively rare but important syndrome.55

Hypothalamic involvement in cluster headachestudies of cerebral blood flow in cluster headache are scarce. most such studies have been done with sPeCt, and the results of this semi quantitative method have been quite heterogeneous; some data indicate an increase, some a decrease, and some no differences in cortical blood flow. this incongruency probably results from methodo logical differences.56 in a large sample of patients, signifi cant activations in the ipsilateral hypo thalamic gray matter were observed with Pet during cluster headache attacks when compared with the headache­free state.57 later observations questioned whether the area of activation seen is indeed within the hypothalamus; stimulation studies (see below) indicate that the position of the tip of the electrodes might be within the triangle of sano rather than in the hypothalamus. in any case, this highly significant activation was not seen in patients with clus ter headache out of the bout when compared with patients who were experiencing an acute cluster headache attack.20 in contrast to patients with migraine, no brainstem acti­vation was found during the acute attack compared to the resting state (Figure 2). this finding is remarkable, as migraine and cluster headache are often discussed as related disorders, and specific compounds, such as ergo­ta mine and sumatriptan, are currently used in the acute treatment of both types of headache.

© 2009 Macmillan Publishers Limited. All rights reserved

nature reviews | neurology volume 5 | aPril 2009 | 205

reviews

these findings prompted for the first time the use of deep brain stimulation in the posterior hypothalamic gray matter in a patient with intractable cluster headache; this procedure led to a complete relief of attacks.58 to date, nearly 50 patients with intractable cluster headache have been reported as having undergone deep brain stimu­lation,59 some of whom have been followed up for more than 7 years.59,60 this procedure seems to be success­ful in 50–60% of patients. in order to unravel the brain circuitry that mediates stimulation­induced effects in cluster headache, Pet was applied in patients who had undergone hypothalamic deep brain stimulation.61 this investigation demonstrated that stimula tion induced activation in the ipsilateral hypothalamic gray matter (the site of the stimu lator tip), the ipsilateral thalamus, the somato sensory cortex and precuneus, the anterior cingu­late cortex, and the ipsilateral trigeminal nucleus and ganglion. the authors additionally observed deactiva­tions in the middle temporal gyrus, posterior cingulate cortex, and contralateral anterior insula. the observed activations and deactivations are both situated in cerebral structures that belong to neuronal circuits that are usually activated in pain transmission and, notably, in acute cluster headache attacks. these data are evidence against a nonspecific, antinociceptive effect or pure inhibition of hypothalamic activity in cluster headache. instead, the findings suggest a hitherto unrecognized functional modulation of the pain­processing network as the mode of action of hypothalamic deep brain stimulation in this headache type.59,62

two proton mr spectroscopy (1H­mrs) studies, pub­lished in 2006, provide further evidence that a hypo­thalamic derangement occurs in cluster headache.63,64 1H­mrs permits a noninvasive biochemical analysis of a defined volume of matter within the brain. the metabo­lites visible in vivo with 1H­mrs include creatine phos­phocreatine (Cr), choline­containing compounds (Cho) and N­acetyl aspartate (naa). the naa/Cr ratio is a marker of neuronal function and is reduced in condi­tions of neuronal loss or dysfunction. the Cho/Cr ratio is an indirect marker of myelination and cell­membrane metabolism. wang and colleagues reported that the hypo­thalamic naa/Cr and Cho/Cr ratios were significantly lower in 47 patients with episodic cluster headache than they were in either headache­free controls or patients with migraine.63 no differences were seen between metabolite ratios when patients with episodic cluster headache were studied during the active bout and remission phases. similarly, lodi and colleagues reported that the hypo­thalamic naa/Cr ratio was reduced in 26 patients with episodic or chronic cluster headache when compared with headache­free controls.64 these data suggest that persistent hypothalamic neuronal dysfunction occurs in patients with cluster headache.

sprenger and colleagues used Pet with the opio idergic ligand [11C]diprenorphine to demonstrate decreased tracer binding in the pineal gland, but not in any other brain structures, of patients with cluster headache

compared with healthy volunteers.41 Furthermore, the data indicated an inverse linear relationship between the duration of cluster headaches and opioid receptor availability in the ipsilateral hypothalamus and bilateral cingu late cortices, a finding that suggests that opioidergic mechanisms in the pineal gland and hypothalamus might be involved in the generation of cluster headache attacks. However, opioids are generally considered to be ineffec­tive abortive and preventive agents in cluster headache and the observed results might be secondary to pain effects in general. to determine whether this reduced opioid­receptor availability persists when the patients are in remission and not taking any drugs would certainly be of interest.

TACs: a shared pathophysiology? if the taCs do indeed share a common patho physio­logical background,65 functional imaging should be able to deline ate similar structures in the dif ferent conditions in this group. the clinical similarities between these syndromes, such as strict half­sidedness and marked autonomic features,55 prompted the suggestion to unify them on clinical grounds as the taCs.66 little is known about the pathophysiology of this group of short­lived circadian headaches, but marked differences occur in the clinical picture of the different taCs in, for instance, the frequency and duration of attacks and the different approaches to treatment.

when functional mri was performed during sponta­neous pain attacks in patients with sunCt, activation was seen in the ipsilateral inferior posterior hypo thalamic gray matter as compared with the resting state.67–69 the activation in the hypothalamus was seen solely in the pain state and was in the same area that has been demon­strated to be activated in patients with cluster headache57 and in patients with paroxysmal hemicrania,70 an obser­vation that suggests that considerable commonali ties exist between paroxysmal hemicrania and cluster headache. indeed, the data might explain the episodic nature of the pain in these headache types. Furthermore, in a case study, functional mri was used to assess a 68­year­old patient who was experiencing excruciating trigeminal autonomic headache attacks, in whom the frequency and duration of attacks, and therapeutic response, enabled no clear­cut classification as one of the subtypes of taC.71 However, the cerebral activation pattern was similar—although not identical—to those previously observed in cluster headache72 and sunCt,67 with a prominent activation in the hypothalamic gray matter.71 this case study underlines the conceptual value of the term ‘taC’ for the group of headaches that are focused around the trigeminal autonomic reflex and, moreover, emphasizes the importance of the hypothalamus as a key region in the patho physiological process of this entity.

Further to these studies, two different groups have reported on two cases of sunCt that were successfully treated with hypothalamic deep brain stimulation.73,74 these reports certainly strengthen the hypothesis for a role

© 2009 Macmillan Publishers Limited. All rights reserved

206 | APRIL 2009 | voLume 5 www.nature.com/nrneurol

reviews

of the hypothalamus in the pathophysiology of the taCs, but considering that only three patients with sunCt have been reported as having changes on functional imaging in the hypothalamus and only two patients with sunCt have been reported who had hypothalamic deep brain stimula­tion, a discussion of this invasive opera tion as a standard treatment for this headache type is at present unjustified.

Hemicrania continua is a strictly unilateral, continu­ous headache of moderate intensity, with superimposed exacerbations of severe intensity that are accompanied by trigeminal autonomic features and migrainous symp­toms. this headache type is currently not listed in the group of taCs because of missing or subtle autonomic symptoms, although several features, such as half­ sidedness and circadian behavior of the pain, suggests at least a similarity to the taCs. moreover, the syndrome is exquisitely responsive to indometacin.23 in seven patients with hemicrania continua, a significant activation of the contra lateral posterior hypothalamus and ipsilateral dorsal rostral pons was described in association with the headache.75 in addition, activation of the ipsilateral ventro lateral midbrain occurred, which extended over the red nucleus and the substantia nigra, and over the bi lateral pontomedullary junction. this study demonstrated effectively that the neuro imaging markers of trigeminal autonomic headaches and migraine syndromes are also apparent in hemicrania continua. this finding mirrors the clinical phenotype of this disorder, which also exhibits a certain overlap with trigeminal autonomic headaches and migraine.75 taken together, just as in an atypical tri­geminal autonomic headache,71 the functional imaging data in hemicrania continua75 impressively emphasize that primary headache syndromes can be distinguished on a functional neuroanatomic basis by areas of activation specific to the clinical presentation.

Structural imaging—morphometric changesstudies of brain morphology used to be performed com­pletely in autopsy material. this situation changed with the advent of modern in vivo imaging methods, and in particular mri. while early neuroimaging studies pro­vided a qualitative description of normal brain mor­phology and its deviations in disease, more recently developed mr­based methods permit a semi quantitative evaluation of these features. the number of methods based on measurements of surface, shape or volume, which capture different and complementary morpho­logical characteristics of the brain, is steadily growing.76 one of the widely used and validated morphometric techniques used to capture structural alterations in the brain is voxel­based morphometry (vBm), a whole­brain method for analysis of automatically preprocessed structural high­ resolution mri data that treats images as continuous scalar measurements.77

one of the first studies that used vBm to investigate possible brain differences between patients with head­ache and individuals without headache was conducted in cluster headache.78 the researchers found a significant

structural difference in gray matter density (a ‘lesion’ that was positioned in the inferior posterior hypo thalamus) in the patients with cluster headache, indicative of a co­ localization of morphometric alterations and functional activation in these patients.78 although this co­localization strongly points away from a false­positive finding, further and independent studies are clearly required to confirm the finding. in relation to migraine, a pioneering study by matharu et al. did not find any significant morphometric changes in gray or white matter in patients with episodic migraine.79 However, five more­recent studies have ques­tioned matharu et al.’s findings. the first of these studies was published by rocca et al. and reported increased den­sities of the periaqueductal gray and dorsolateral pons in patients with migraine.80 the authors also identified a decrease in gray matter volume in the anterior cingu­late cortex and both insulae in patients with migraine. to date, the finding of decreased gray matter volume has been replicated in four independent studies,81–84 whereas none of these investigations replicated the increase in gray matter in brainstem structures. one possible explanation for this discrepancy is that the latter studies used a 1.5 t scanner, whereas rocca et al. used a higher field strength (3 t), which might enable the detection of more­subtle differences. see Figure 2 for a summary of the structural data from the studies cited within the text.

a 2005 study used mri and vBm to demonstrate that patients with chronic tension­type headache have a signifi cant gray matter decrease compared with healthy controls in regions known to be involved in pain pro­cessing, namely the cingulate cortex, insula, and the orbito frontal cortex and parahippocampus bilaterally.85 the same study compared patients with medication overuse headache against healthy controls and showed that patients with medication overuse headache had a non significant decrease in gray matter in the left orbito­ frontal cortex and the right midbrain. the fact that the change in gray matter in patients with chronic tension­type headache was restricted to structures involved in pain processing led the authors to conclude that the change might be interpreted as the consequence of central sensitization, generated by prolonged nociceptive input from the pericranial myofascial tissues.15

the structural studies in migraine, as well as the study in chronic tension­type headache, have to be viewed, however, in the light of recent morphometric studies in several chronic pain syndromes.86 all of these investi­gations (including ones in fibromylagia,87 irritable bowel syndrome,88 phantom pain,89 chronic back pain90,91 and thoracic spinal cord injury92) showed a decrease in gray matter volume in some of the above­mentioned areas. a striking feature of all of these studies is the fact that the gray matter changes were not randomly distributed but involved defined and functionally highly specific brain areas—namely, those involved in supraspinal nociceptive processing. this common ‘brain signature’ of patients with chronic pain raises the question of whether the changes observed in patients with migraine and those

© 2009 Macmillan Publishers Limited. All rights reserved

nature reviews | neurology volume 5 | aPril 2009 | 207

reviews

with chronic tension­type headache are the cause or the consequence of chronic pain.86 most of the changes observed in the studies of pain syndromes correlated with pain duration, so it seems plausible to argue that the alteration of these regions is a consequence, rather than a cause, of frequent nociceptive input. the nature of chronic pain makes it difficult to prove this point. migraine usually remits with age, for reasons that are, however, unknown. a very interesting question for future studies is whether morphological changes reverse when migraine, and hence the disproportionate amount of nociceptive stimulation, stops.

ConclusionsGiven the rapid advances in functional neuroimaging methods, these techniques continue to have an impor­tant role and open new avenues in the targeting of neural substrates in individual primary headache syn­dromes. the use of functional imaging for a syndrome that changes within a time frame of hours rather than seconds or minutes remains a challenge and explains why thorough literature exists for functional imaging with mri in experimental pain, whereas Pet data have so far yielded the best results for headache. many excellent studies have confirmed that the aura phase of migraine is associated with a reduction in rCBF, mainly in the occipital cortex. However, the classic vasoconstriction– vasodilation hypothesis for aura and headache seems now to be largely disproved. the cortical blood flow changes could possibly be interpreted in relation to brain stem mechanisms, as several groups have demonstrated a con­sistent increase in rCBF in the rostral brainstem in acute migraine attacks and have thereby reinforced the view that migraine is a subcortical dis order with substantial brainstem involvement.

in trigeminal autonomic headache syndromes, areas of brain activation have been divided into two broad

groups: areas that are known to be involved in tri geminal pain processing or responses to pain, and a region that is probably involved in the initiation of the attack. the latter activation has not been observed in migraine or experi­mental head pain. Furthermore, a significant structural difference in gray matter density of the hypo thalamus has been shown in patients with cluster headache com­pared with healthy volunteers. the co­localization of morphometric and functional changes demonstrates the precise anatomical location for the Cns lesion of cluster headache and suggests an involvement of this area as a primum movens in acute attack (a discovery that prompted leone et al. to use deep brain stimulation of this area in an attempt to prevent cluster headache attacks in selected severe therapy­refractory cases58). taken together, functional imaging data impressively emphasize that primary headache syndromes can be distinguished on a functional neuroanatomical basis by areas of activation that are specific to the clinical presen­tation. these advances have a direct clinical application for assessment of prognosis, and for allocation of patients to different treatment strategies (for example, deep brain stimulation in cluster headache and sunCt).

Review criteria

A literature search was performed for papers published up until November 2008, using the reference database MeDLiNe. The key words used were as follows: “migraine”, “headache”, “cluster headache”, “trigemino-autonomic headache”, “imaging”, “functional imaging”, “PeT”, “f-Mri”, “morphometry” and “vBM”. All papers published in english or German were considered. Papers discovered by this search were reviewed, as were references cited therein. in addition, review books and the european and German diagnostic recommendations for headache were considered.

1. Kruit, M. C. et al. Migraine as a risk factor for subclinical brain lesions. JAMA 291, 427–434 (2004).

2. Tietjen, G. e. stroke and migraine linked by silent lesions. Lancet Neurol. 3, 267 (2004).

3. swartz, r. H. & Kern, r. Z. Migraine is associated with magnetic resonance imaging white matter abnormalities: a meta-analysis. Arch. Neurol. 61, 1366–1368 (2004).

4. rozen, T. D. vanishing cerebellar infarcts in a migraine patient. Cephalalgia 27, 557–560 (2007).

5. Agarwal, s., Magu, s. & Kamal, K. reversible white matter abnormalities in a patient with migraine. Neurol. India 56, 182–185 (2008).

6. Adami, A. et al. right-to-left shunt does not increase white matter lesion load in migraine with aura patients. Neurology 71, 101–107 (2008).

7. Dowson, A. et al. Migraine intervention with sTArFlex Technology (MisT) trial: a prospective, multicenter, double-blind, sham-controlled trial to evaluate the effectiveness of patent foramen ovale closure with sTArFlex septal repair implant to resolve refractory migraine headache. Circulation 117, 1397–1404 (2008).

8. wilmshurst, P. T., Nightingale, s., walsh, K. P. & Morrison, w. L. effect on migraine of closure of cardiac right-to-left shunts to prevent recurrence of decompression illness or stroke or for haemodynamic reasons. Lancet 356, 1648–1651 (2000).

9. sandrini, G. et al. Neurophysiological tests and neuroimaging procedures in non-acute headache: guidelines and recommendations. Eur. J. Neurol. 11, 217–224 (2004).

10. May, A. et al. Diagnostik und apparative Zusatzuntersuchungen bei Kopfschmerzen. (Thieme, stuttgart, New York, 2008).

11. Detsky, M. e. et al. Does this patient with headache have a migraine or need neuroimaging? JAMA 296, 1274–1283 (2006).

12. Favier, i. et al. Trigeminal autonomic cephalgias due to structural lesions: a review of 31 cases. Arch. Neurol. 64, 25–31 (2007).

13. edlow, J. A., Panagos, P. D., Godwin, s. A., Thomas, T. L. & Decker, w. w. Clinical policy: critical issues in the evaluation and management of adult patients presenting to the emergency department with acute headache. Ann. Emerg. Med. 52, 407–436 (2008).

14. May, A. et al. experimental cranial pain elicited by capsaicin: a PeT-study. Pain 74, 61–66 (1998).

15. Peyron, r., Laurent, B. & Garcia-Larrea, L. Functional imaging of brain responses to pain. A review and meta-analysis. Neurophysiol. Clin. 30, 263–288 (2000).

16. Tracey, i. Neuroimaging of pain mechanisms. Curr. Opin. Support Palliat. Care 1, 109–116 (2007).

17. Borsook, D., Burstein, r. & Becerra, L. Functional imaging of the human trigeminal system: opportunities for new insights into pain processing in health and disease. J. Neurobiol. 61, 107–125 (2004).

18. Upadhyay, J., Knudsen, J., Anderson, J., Becerra, L. & Borsook, D. Noninvasive mapping of human trigeminal brainstem pathways. Magn. Reson. Med. 60, 1037–1046 (2008).

19. May, A., Büchel, C., Bahra, A., Goadsby, P. J. & Frackowiak, r. s. J. intra-cranial vessels in trigeminal transmitted pain: a PeT study. Neuroimage 9, 453–460 (1999).

20. May, A., Bahra, A., Büchel, C., Frackowiak, r. s. & Goadsby, P. J. PeT and MrA findings in cluster headache and MrA in experimental pain. Neurology 55, 1328–1335 (2000).

© 2009 Macmillan Publishers Limited. All rights reserved

208 | APRIL 2009 | voLume 5 www.nature.com/nrneurol

reviews

21. May, A., Buchel, C., Turner, r. & Goadsby, P. J. Magnetic resonance angiography in facial and other pain: neurovascular mechanisms of trigeminal sensation. J. Cereb. Blood Flow Metab. 21, 1171–1176 (2001).

22. Bednarczyk, e. M. et al. Brain blood flow in the nitroglycerin (GTN) model of migraine: measurement using positron emission tomography and transcranial Doppler. Cephalalgia 22, 749–757 (2002).

23. Headache Classification Committee of the international Headache society. The international Classification of Headache Disorders, 2nd edition. Cephalalgia 24 (Suppl. 1), 1–160 (2004).

24. weiller, C. et al. Brain stem activation in spontaneous human migraine attacks. Nat. Med. 1, 658–660 (1995).

25. Bahra, A., Matharu, M. s., Buchel, C., Frackowiak, r. s. & Goadsby, P. J. Brainstem activation specific to migraine headache. Lancet 357, 1016–1017 (2001).

26. raskin, N. H., Hosobuchi, Y. & Lamb, s. Headache may arise from perturbation of brain. Headache 27, 416–420 (1987).

27. veloso, F., Kumar, K. & Toth, C. Headache secondary to deep brain implantation. Headache 38, 507–515 (1998).

28. Petrovic, P., Petersson, K. M., Hansson, P. & ingvar, M. Brainstem involvement in the initial response to pain. Neuroimage 22, 995–1005 (2004).

29. Bingel, U. et al. subcortical structures involved in pain processing: evidence from single-trial fMri. Pain 99, 313–321 (2002).

30. rolls, e. T. et al. representations of pleasant and painful touch in the human orbitofrontal and cingulate cortices. Cereb. Cortex 13, 308–317 (2003).

31. singer, T. et al. empathy for pain involves the affective but not sensory components of pain. Science 303, 1157–1162 (2004).

32. Afridi, s. et al. A positron emission tomographic study in spontaneous migraine. Arch. Neurol. 62, 1270–1275 (2005).

33. Afridi, s. K. et al. A PeT study exploring the laterality of brainstem activation in migraine using glyceryl trinitrate. Brain 128, 932–939 (2005).

34. Matharu, M. s. et al. Central neuromodulation in chronic migraine patients with suboccipital stimulators: a PeT study. Brain 127, 220–230 (2004).

35. Denuelle, M., Fabre, N., Payoux, P., Chollet, F. & Geraud, G. Hypothalamic activation in spontaneous migraine attacks. Headache 47, 1418–1426 (2007).

36. rosen, s. D. et al. silent ischemia as a central problem: regional brain activation compared in silent and painful myocardial ischemia. Ann. Intern. Med. 124, 939–949 (1996).

37. Lothe, A. et al. interictal brain 5-HT1A receptors binding in migraine without aura: a 18F-MPPF-PeT study. Cephalalgia 28, 1282–1291 (2008).

38. srikiatkhachorn, A., Govitrapong, P. & Limthavon, C. Up-regulation of 5-HT2 serotonin receptor: a possible mechanism of transformed migraine. Headache 34, 8–11 (1994).

39. Chabriat, H. et al. 5HT2 receptors in cerebral cortex of migraineurs studied using PeT and 18F-fluorosetoperone. Cephalalgia 15, 104–108 (1995).

40. sakai, Y., Dobson, C., Diksic, M., Aube, M. & Hamel, e. sumatriptan normalizes the migraine attack-related increase in brain serotonin synthesis. Neurology 70, 431–439 (2008).

41. sprenger, T. et al. Opioidergic changes in the pineal gland and hypothalamus in cluster headache: a ligand PeT study. Neurology 66, 1108–1110 (2006).

42. Borsook, D. et al. A ‘BOLD’ experiment in defining the utility of fMri in drug development. Neuroimage 42, 461–466 (2008).

43. Friberg, L., Olesen, J., Lassen, N. A., Olsen, T. s. & Karle, A. Cerebral oxygen extraction, oxygen consumption, and regional cerebral blood flow during the aura phase of migraine. Stroke 25, 974–979 (1994).

44. Olesen, J. et al. Timing and topography of cerebral blood flow, aura, and headache during migraine attacks. Ann. Neurol. 28, 791–798 (1990).

45. Olesen, J. & Friberg, L. Xenon-133 sPeCT studies in migraine without aura in Migraine and Other Headaches: The Vascular Mechanisms (ed Olesen, J) 237–243 (raven Press, London, 1991).

46. Olesen, J., Tfelt-Hansen, v. & welch, K. M. A. The Headaches 2nd edn (Lippincott, williams & wilkins, Philadelphia, 1999).

47. Cutrer, F. M. et al. Perfusion-weighted imaging defects during spontaneous migrainous aura. Ann. Neurol. 43, 25–31 (1998).

48. Hadjikhani, N. et al. Mechanisms of migraine aura revealed by functional Mri in human visual cortex. Proc. Natl Acad. Sci. USA 98, 4687–4692 (2001).

49. Bolay, H. et al. intrinsic brain activity triggers trigeminal meningeal afferents in a migraine model. Nat. Med. 8, 136–142 (2002).

50. Kelman, L. The aura: a tertiary care study of 952 migraine patients. Cephalalgia 24, 728–734 (2004).

51. Granziera, C., Dasilva, A. F., snyder, J., Tuch, D. s. & Hadjikhani, N. Anatomical alterations of the visual motion processing network in migraine with and without aura. PLoS Med. 3 e402 (2006).

52. Goadsby, P. J. The migrainous brain: what you see is not all you get? PLoS Med. 3 e404 (2006).

53. Fumal, A. et al. Orbitofrontal cortex involvement in chronic analgesic-overuse headache evolving from episodic migraine. Brain 129, 543–550 (2006).

54. sprenger, T. et al. Altered metabolism in frontal brain circuits in cluster headache. Cephalalgia 27, 1033–1042 (2007).

55. May, A. Cluster headache: pathogenesis, diagnosis, and management. Lancet 366, 843–855 (2005).

56. May, A. A review of diagnostic and functional imaging in headache. J. Headache Pain 7, 174–184 (2006).

57. May, A., Bahra, A., Büchel, C., Frackowiak, r. s. J. & Goadsby, P. J. Hypothalamic activation in cluster headache attacks. Lancet 352, 275–278 (1998).

58. Leone, M., Franzini, A. & Bussone, G. stereotactic stimulation of posterior hypothalamic gray matter in a patient with intractable cluster headache. N. Engl. J. Med. 345, 1428–1429 (2001).

59. Leone, M. et al. Lessons from 8 years’ experience of hypothalamic stimulation in cluster headache. Cephalalgia 28, 787–797 (2008).

60. schoenen, J. et al. Hypothalamic stimulation in chronic cluster headache: a pilot study of efficacy and mode of action. Brain 128, 940–947 (2005).

61. May, A. et al. Hypothalamic deep brain stimulation in positron emission tomography. J. Neurosci. 26, 3589–3593 (2006).

62. May, A. Hypothalamic deep-brain stimulation: target and potential mechanism for the treatment of cluster headache. Cephalalgia 28, 799–803 (2008).

63. wang, s. J., Lirng, J. F., Fuh, J. L. & Chen, J. J. reduction in hypothalamic 1H-Mrs metabolite ratios in patients with cluster headache. J. Neurol. Neurosurg. Psychiatry 77, 622–625 (2006).

64. Lodi, r. et al. study of hypothalamic metabolism in cluster headache by proton Mr spectroscopy. Neurology 66, 1264–1266 (2006).

65. Goadsby, P. J. Pathophysiology of cluster headache: a trigeminal autonomic cephalgia. Lancet Neurol. 1, 251–257 (2002).

66. Goadsby, P. J. & Lipton, r. B. A review of paroxysmal hemicranias, sUNCT syndrome and other short-lasting headaches with autonomic feature, including new cases. Brain 120, 193–209 (1997).

67. May, A., Bahra, A., Buchel, C., Turner, r. & Goadsby, P. J. Functional magnetic resonance imaging in spontaneous attacks of sUNCT: short-lasting neuralgiform headache with conjunctival injection and tearing. Ann. Neurol. 46, 791–794 (1999).

68. sprenger, T. et al. sUNCT: bilateral hypothalamic activation during headache attacks and resolving of symptoms after trigeminal decompression. Pain 113, 422–426 (2005).

69. Cohen, A. s., Matharu, M. s., Kalisch, r., Friston, K. & Goadsby, P. J. Functional Mri in sUNCT shows differential hypothalamic activation with increasing pain. Cephalalgia 24, 1098–1099 (2004).

70. Matharu, M. s., Cohen, A. s., Frackowiak, r. s. & Goadsby, P. J. Posterior hypothalamic activation in paroxysmal hemicrania. Ann. Neurol. 59, 535–545 (2006).

71. sprenger, T. et al. Hypothalamic activation in trigeminal autonomic cephalgia: functional imaging of an atypical case. Cephalalgia 24, 753–757 (2004).

72. May, A., Bahra, A., Büchel, C., Frackowiak, r. & Goadsby, P. Cerebral activation in cluster headache in and outside the bout: a PeT study. Eur. J. Neurol. 5, 7–8 (1998).

73. Leone, M. et al. Deep brain stimulation to relieve drug-resistant sUNCT. Ann. Neurol. 57, 924–927 (2005).

74. Lyons, M. K., Dodick, D. w. & evidente, v. G. responsiveness of short-lasting unilateral neuralgiform headache with conjunctival injection and tearing to hypothalamic deep brain stimulation. J. Neurosurg. 110, 279–281 (2009).

75. Matharu, M. s. et al. Posterior hypothalamic and brainstem activation in hemicrania continua. Headache 44, 747–761 (2004).

76. May, A. & Gaser, C. Magnetic resonance-based morphometry: a window into structural plasticity of the brain. Curr. Opin. Neurol. 19, 407–411 (2006).

77. Draganski, B. & May, A. Training-induced structural changes in the adult human brain. Behav. Brain Res. 192, 137–142 (2008).

78. May, A. et al. Correlation between structural and functional changes in brain in an idiopathic headache syndrome. Nat. Med. 5, 836–838 (1999).

© 2009 Macmillan Publishers Limited. All rights reserved

nature reviews | neurology volume 5 | aPril 2009 | 209

reviews

79. Matharu, M. s., Good, C. D., May, A., Bahra, A. & Goadsby, P. J. No change in the structure of the brain in migraine: a voxel-based morphometric study. Eur. J. Neurol. 10, 53–57 (2003).

80. rocca, M. A. et al. Brain gray matter changes in migraine patients with T2-visible lesions: a 3-T Mri study. Stroke 37, 1765–1770 (2006).

81. valfre, w., rainero, i., Bergui, M. & Pinessi, L. voxel-based morphometry reveals gray matter abnormalities in migraine. Headache 48, 109–117 (2008).

82. schmidt-wilcke, T., Ganssbauer, s., Neuner, T., Bogdahn, U. & May, A. subtle grey matter changes between migraine patients and healthy controls. Cephalalgia 28, 1–4 (2008).

83. schmitz, N. et al. Attack frequency and disease duration as indicators for brain damage in migraine. Headache 48, 1044–1055 (2008).

84. Kim, J. H. et al. regional grey matter changes in patients with migraine: a voxel-based morphometry study. Cephalalgia 28, 598–604 (2008).

85. schmidt-wilcke, T. et al. Gray matter decrease in patients with chronic tension type headache. Neurology 65, 1483–1486 (2005).

86. May, A. Chronic pain may change the structure of the brain. Pain 137, 7–15 (2008).

87. Kuchinad, A. et al. Accelerated brain gray matter loss in fibromyalgia patients: premature aging of the brain? J. Neurosci. 27, 4004–4007 (2007).

88. Davis, K. D. et al. Cortical thinning in iBs: implications for homeostatic, attention, and pain processing. Neurology 70, 153–154 (2008).

89. Draganski, B. et al. Decrease of thalamic gray matter following limb amputation. Neuroimage 31, 951–957 (2006).

90. Apkarian, A. v. et al. Chronic back pain is associated with decreased prefrontal and

thalamic gray matter density. J. Neurosci. 24, 10410–10415 (2004).

91. schmidt-wilcke, T. et al. Affective components and intensity of pain correlate with structural differences in gray matter in chronic back pain patients. Pain 125, 89–97 (2006).

92. wrigley, P. J. et al. Anatomical changes in human motor cortex and motor pathways following complete thoracic spinal cord injury. Cereb. Cortex 19, 224–232 (2009).

AcknowledgmentsThe author acknowledges the help of Kristin ihle in performing the meta-analysis of functional and structural imaging data in headache and for her help in preparing Figure 2. A May is supported by a grant from the Deutsche Forschungsgemeinschaft (MA 1862/2). This study is supported by a grant from the Federal Ministry of education and research (BMBF project no. 371 57 01).

© 2009 Macmillan Publishers Limited. All rights reserved