1250-1263 11 Barnstable, C. J., Hofstein, R. and Akagawa, K. (1985) Dev. Brain Res. 20, 286-290 12 Bj6rkland,A. and H6kfelt, T. (1984) in Handbook of Chemical Neuroanatomy (VoL 2: Classical Transmitters in the CNS, Part I), (Bj6rkland, A. and HOkfelt, T., eds), pp. 196-197, Elsevier 13 Russo,A. F. et a/. (1988) Neuron 1,
311-320 14 Coupland, R. E. and Tomlinson, A. (1989) Int. J. Dev. Neurosci. 7, 419--438 15 Rothman,T. P. etaL (1980) Proc. Natl Acad. Sci. USA 77, 6221-6225 16 Specht, L. A., Pickel, V. M., Joh, T. H. and Reis,D. J. (1981)J. Comp. NeuroL 199, 233-253 17 McLean, J. H. and Shipley, M. T. (1988) J. Neurosci. 8, 3658-3669
18 Trimble, W. S., Gray, T. S., Elferink, L. A., Wilson, M. C. and Scheller,R. H. (1990) J. Neurosci. 10, 1380-1387 19 Oyler, G. A. et al. (1989) J. Cell Biol. 109, 3039-3052 20 Forss-Petter, S., Danielson, P., Battenberg, E., Bloom, F. and Sutcliffe, J. G. (1989) J. Mol. Endocrinol. 1, 63-75 21 Schwanzel-Fukuda, M. and Pfaff, D. W. (1989) Nature 338, 161-164
viewpoint; Migraine: a researchfield matured for the basicneurosciences _
.II!i]i !_.I ..............................
Jl
I.
I
_
I _Jlll ...........
IIII .....
III_ H_
Jes Olesen and Lars Edvinsson Progress in migraine research has been rapid in recent years, from both the basic science and the clinical perspectives. A new internationally accepted headache classification with operational diagnostic criteria was published in 1988, eliminating much diagnostic uncertainty. More than a decade of study of regional cerebral blood flow (rCBF) has gradually shown a pathognomonic pattern of abnormalities, probably reflecting spreading cortical depression. Recently it has been shown that pain probably arises from excitation of perivascularpial arterial nociceptors. The innervation and receptor mechanismsof pial and extracranialarteries have been worked out in detail both in animal and humans. Involvement of calcitonin generelated peptide (CGRP) and 5-hydroxytryptamine (5-HT) during migraine attacks has been demonstrated. A new and specific 5-HT1D receptor agonist has proved to be highly effective in treating migraine. Therefore, major research efforts recently have been concentrated on discovering the location and function of 5-HT1o receptors, extra- and intracranJally. Thus, it is now possible to formulate useful neuroscientific researchstrategiesaimed at clarifying migraine mechanisms. Migraine affects approximately 10% of the adult population and about one third of this number have two or more attacks every month 1. Patients with migraine or other headaches constitute the largest single group of patients seen in the neurologist's office. Despite the magnitude of this clinical problem, it has been poorly described. In fact, so little has been known about its pathophysiology that basic neuroscientific studies were, until recently, almost impossible. The situation has improved gradually, and an extensive account of known basic mechanisms of migraine was given in 19882 . This article briefly describes recent scientific breakthroughs, showing that migraine research is now amenable to experimental study, both in the laboratory and in the clinic. The four areas discussed are: the classification and diagnostic criteria of migraine; the pathognomonic alterations of brain blood flow during migraine attacks; the release of perivascular peptides during migraine attacks; and the association of 5-hydroxytryptamine (5-HT) and its receptors with migraine pathogenesis.
Improved classification and diagnostic criteria for headache disorders The first attempt to classify and define headache disorders was made in 1963 by an ad hoc committee of the National Institutes of Health (USA) 3. The TINS, Vol. 14, No. 1, 1991
definition of migraine read as follows: 'attacks of headache, widely varying in intensity, frequency and duration etc.'. Such criteria cannot be used consistently, because they do not specify whether one or more of these clinical characteristics should be present, or how often they should be present. Furthermore, diagnostic terms were ambiguous at that time, and were used differently in different countries. If a disease is poorly defined, it is also difficult to study its mechanisms, since not all of the patients under investigation will actually be suffering from the disorder. However, a committee of the International Headache Society has now completed a new headache classification containing operational diagnostic criteria 4. The headache classification, originally published in English, has also been published in French, German, Italian, Slovenian and Turkish. Translations into Spanish, Portuguese and Japanese are in progress. The classification has been sponsored by the World Federation of Neurology, the Research Group on Migraine and Other Headaches and has, in simplified form, been used by the World Health Organization (WHO) in the most recent International Classification of Diseases (ICD 10). Epidemiological and genetic studies of headache will benefit immensely from the new classification, which will also facilitate pathophysiological and therapeutic studies.
JesOlesenisat the Departmentof Neurology, Gentofte Hospital, Universityof Copenhagen, Copenhagen DK-2900,Denmark, and LarsEdvinssonis at the Departmentof InternalMedicine, UniversityHospital, LundS-22221, Sweden.
Studies of regional cerebral blood flow (rCBF) during migraine attacks For more than a century, scientists have debated the importance of vascular and neurogenic (CNS) components in the pathophysiology of migraine. Despite the lack of suitable methods for studying arterial pulsations, arterial diameter or blood flow, Wolff gathered evidence in favour of the vascular hypothesis during the 1940s and 1950s 5. It was not until the advent of the intra-arterial 133Xe method that quantitative studies of rCBF could be clone during migraine attacks, and it was only with the development of computerized multi-detector systems that measurements with sufficient spatial resolution were obtained. This development has been reviewed previously 6, and therefore the rest of this article focuses on more recent observationsL Early on in an attack of migraine with aura, rCBF is reduced at the posterior pole of the brain; this reduced flow gradually involves larger areas of one
© 1991.ElsevieSci r encePublishersLtd.(UK) 0166-2236/91/$02.00
3
viewpoint 120Migraine without aura
Migraine with aura _* 1 I
100-
~ 80"./¢~. ~-~-
O
E
Q. V a
._l
60-
I
....
n-
~.
O O
_T
z~0-
in vivo.
=
I
.N ~/,
'
20
I i I I
I ...
L._
CF
EJV
EJV
C F
EJV
Y/" EJV
Fig. 1. Changes in CGRP levels during mioraine headache. Open bars represent control measurc,,nents taken either from cubital fossa (CF) or external jugular vein (EJV); hatched bars represent measurements from EJV taken during migraine attacks. An asterisk indicates that there is a significant difference between that measurement and the control Abbreviation: CGRP-LL calcitonin generelated peptide-like immunoreactivity. (Taken, with per-
mission, from Ref. 22.)
hemisphere. Patients usually report their typical aura symptoms, which last for approximately 20-60 min, in this early phase of the attack. After the aura has passed, when the headache begins, and during the early headache phase, rCBF is similarly depressed. Hypoperfusion instead of hyperperfusion occurs at the onset of headache. This contradicts Wolff's hypothesis 5, which is therefore called into question. During the headache phase, blood flow normalizes gradually and then develops into hyperemia (increased blood flow) with no apparent change in headache, nausea and vomiting. In some cases it has been noted that hyperemia outlasts headache, but in most patients measurements have not been performed at the appropriate time to answer this question. The aura symptoms, the headache and the blood-flow reduction all originate in the same hemisphere. In other words, the aura symptoms are perceived contralateral to blood-flow reduction and 4
headache when all of these factors are unilateral. Quite often, however, headache is bilateral despite unilateral aura symptoms and unilateral rCBF reduction. It has recently been found that cerebral blood vessels have bilateral innervation, particularly in the midline 8,9. This means that the headache phase of migraine with aura might be explained by simple activation of perivascular nerves by neurotransmitters, potassium and other substances released from neurones and glial cells, after the migrainous process in the cerebral cortex. In patients suffering from migraine without aura the findings are more difficult to interpret. We have been unable to demonstrate any blood-flow abnormalities in such patients, but other researchers have described hyperemia in the headache phase (for references see Refs 2 and 6). Changes of rCBF in deep cerebral structures, such as the thalamus, hypothalamus, hippocampus or brain stem,could perhaps induce activity in the trigeminovascular system, and thus cause the attack without being detectable by current blood flow techniques According to measurements of human blood flow, and recent experimental work in animals on spreading depression28 and clinical observations 29, rCBF changes in migraine with aura may be secondan/to spreading cortical depression. A recent study using magnetoencephalography demonstrated abnormal DC potentials in patients with migraine, both during an attack and when not experiencing an attack, compared with healthy controls 1°. Although these data are preliminary, they support the hypothesis of spreading depression, a model that is easy to use in animal experiments. Peptidergic involvement in migraine
Cerebral vessels are supplied with perivascular fibres, most of which are autonomic (sympathetic or parasympathetic), but there is a subpopulation of fibres from primary afferent neurones. Evidence for a sensory innervation in humans was first obtained from the responses of patients to stimulation of different cranial vessels11-13, and the distribution of the trigeminal fibres to cranial structures was outlined by Fang14. Later, the autonomic and sensory pathways were outlined and their neurotransmitter content described 15,1°. In brief, the sympathetic fibres originate in the superior cervical ganglion and contain noradrenaline and neuropeptide Y (NPY). The parasympathetic system originates from the sphenopalatine and otic ganglia, and from a number of small ganglionic formations located at the base of the brain. Thus, this system has a comparatively complex origin 17. The sensory system originates in the first division of the trigeminal ganglion and contains calcitonin gene-related peptide (CGRP), substance P and neurokinin A. A contribution from upper dorsal roots to the lower part of the circle of Willis has been described 16. Although the clinical evidence for an involvement of these systems in migraine is modest, a series of recent studies has provided exciting information. In patients with trigeminal neuralgia (tic douloreux) it was found that thermocoagulation of the trigeminal TINS, VoL 14, No. 1, 1991
.........................................
Irl
ganglion caused a significant release of CGRP and substance P into the external jugular vein 18. It is thus possible to measure in vivo the release of peptides from the trigeminal ganglion, a structure known to supply cranial vessels with sensory fibres. Furthermore, the release of these substances was seen only in association with facial flushing ipsilateral to the side of stimulation. Interestingly, such changes are not seen if CGRP and substance P levels are measured in the peripheral circulation (antecubital vein) 19. Moreover, there are no changes in the levels of either substance P or vasoactive intestinal peptide (VIP) in peripheral blood during headache in common or classical migraine 2°. Recently, the levels of several neuropeptides have been studied during headache by local cranial sampling from the external jugular vein 21,22. The study involved 22 patients (6 males and 16 females) who ranged from 22-58 years old. In each case, neuropeptides were determined from both the peripheral circulation (cubital fossa) and the cranial circulation (external jugular vein). In all patients studied, suffering from either classical (n=10) or common migraine (n=12), CGRP was elevated in the cranial but not in the peripheral blood during migraine attacks (Fig. 1). The level of CGRP in patients with classical migraine was greater than that in patients with common migraine. As a group, there were no changes in substance P, neuropeptide Y or VIP during the headache attack. However, two patients from the group with classical migraine had prominent tearing and nasal discharge, and both these had marked elevations in cranial VIP levels. Thus, taken together these findings suggest that CGRP is an important neuropeptide in migraine.
5-HT and migraine A greater understanding of molecular migraine mechanisms has come from the study of 5-HT and its receptor subtypes. This is the final step in a long development. As far back as the 1960s, Sicuteri et a/. suggested that the migraine attack is associated with increased excretion of 5-HT metabolites 23. This work was followed up by studies of the algogenic properties of 5-HT on dorsal hand veins. Lance and associates then demonstrated convincingly that there is a drop in plasma 5-HT levels during migraine attacks, as well as an increased excretion of 5hydroxyindoleacetic acid (5-HIAA) in the urine following an attack 24. The intravenous injection of 5-HT relieved the attack. In keeping with this, a 'selective' 5-HT receptor agonist, sumatriptan {3- [2-(dimethylamino)ethyl]-N-methylindole-5methanesulphonamide}, has been introduced and found to abort acute attacks of migraine headache 25,26. The introduction of this substance was based on a systematic search for migraine drugs acting on subtypes of 5-HT receptors by Humphrey and his associates. Sumatriptan is considered to be a highly specific receptor agonist, stimulating only the 5-HTI subtype (Ref. 26). Recently, this drug has been shown to possess remarkable efficiency in the treatment of migraine and, owing to its high TINS, Vol. 14, No. 1, 1991
IIII
I
Jlllll IIII]i _H .... JUIJlI!I
viewpoint
specificity, it also provides new insight into possible mechanisms of the migraine attack 27. This new development has stimulated experimental work aiming to characterize 5-HT receptor subtypes on blood vessels and neurones, and also to understand the effects of 5-HT receptor activation. Sumatriptan may be used, not only as a migraine drug, but also experimentally in human studies. What happens to CBF when patients are treated with sumatriptan? What happens to CGRP release? Why does sumatriptan work in both migraine and cluster headache, which are otherwise quite separate entities? Where does sumatriptan work? Since it is very water-soluble it almost certainly does not pass the blood-brain barrier. The answers to such human studies and to many other important animal experiments will undoubtedly greatly increase our understanding of the disease.
Selected references 1 Ziegler,D. K. (1986)in Headache Handbook of Neuro/ogy4 (Vol. 48) (Rose,F. C., ed.), pp. 13-22, Elsevier 20lesen, J. and Edvinsson,L., eds (1988) BasicMechanisms of Headache, Elsevier 3 Ad hoc Committee on Classification of Headache (1962) J. Am. Med. Assoc. 179, 717-718 4 Headache Classification Committee of the International Headache Society(1988) Cephalalgia 8 (Suppl. 7), 1-96 5 Wolff, H. G. (1963) Headache and Other Head Pain, pp. 148-153, Oxford UniversityPress 60lesen, J. (1985) Trends Neurosci. 8, 318-321 70lesen, J. et aL Ann. NeuroL (in press) 8 Uddman, R., Edvinsson, L. and Hara, H. (1989) in New Advances in Headache Research (Rose, F. C., ed.), pp. 121-125, Smith-Bond 9 Moskowitz,M. A., Buzzig, M. G., Saskas,O. E. and Linnik, M. D. (1989) Rev. NeuroL 145, 181-193 10 Barkley, G. L., Tepley, N., Simkins, R., Moran, J. E. and Welch, K. M. A. (1989) Cephalal&ia9 (Suppl. 10), 127-128 11 Fay,T. (1932)Ann. OtoL RhinoL LarynEoL 41, 1030-1032 12 Penfield, W. and McNaughton, F. L. (1949) Arch. NeuroL Physiatr. 44, 43-75 13 Ray,B. S. and Wolff, H. G. (1949) Arch. 5urg. 41,813-856 14 Fang, H. C. H. (1961) Arch. NeuroL 4, 651-656 15 Moskowitz,M. A. (1984) Ann. Neurol. 16, 157-168 16 Uddman, R. and Edvinsson, L. (1989) Cerebrovasc. Brain Metab. Rev. 1,230-252 17 Suzuki, N., Hardebo, J. E. and Owman, C. (1988) J. Cereb. Blood Flow/Vletab. 8, 697-712 18 Goadsby, P. J., Edvinsson, L. and Ekman, R. (1988) Ann. NeuroL 23, 193-196 19 Schon, F. et al. (1987)J. Neurol. Neurosurg. Psychiatr. 50, 642-643 20 Blegvad, N., Jensen, K., Fahrenkrug, J., Schaffalitzky de Muckadell, O. B. and Olesen,J. (1986) Cephalalgia5 (Suppl. 2), 352-353 21 Goadsby, P. J., Edvinsson, L. and Ekman, R. (1989) Cephalalgia 9 (Suppl. I0), 292-293 22 Goadsby, P. J., Edvinsson, L. and Ekman, R. (1990) Ann. Neurol. 28, 183-187 23 Sicuteri,F., Testi,A. and Anselmi, B. (1961) Int. Arch. Allergy Appl. Immunol. 19, 55-58 24 Anthony, M., Hinterberger, H. and Lance,J.W. (1969) Res. Clin. Stud. Headache 2, 29-59 25 Doenicke, A., Brand, J. and Perrin, M. J. (1988) Lancet i, 1309-1311 26 Humphrey, P. P. A. et a/. (1988) Br. J. Pharmacol. 94, 1123-1132 27 Lance,J. W., ed. (1989) Cephalalgia 9 (Suppl. 10), 1-464 28 Lauritzen, M. (1987)Acta. Neurol. 5cand. 76 (Suppl. 113), 1-40 29 Olesen,J. (1988) in BasicMechanisms of Headache (Oleson, J. and Edvinsson,L., eds), pp. 353-363, Elsevier 5
311-320 14 Coupland, R. E. and Tomlinson, A. (1989) Int. J. Dev. Neurosci. 7, 419--438 15 Rothman,T. P. etaL (1980) Proc. Natl Acad. Sci. USA 77, 6221-6225 16 Specht, L. A., Pickel, V. M., Joh, T. H. and Reis,D. J. (1981)J. Comp. NeuroL 199, 233-253 17 McLean, J. H. and Shipley, M. T. (1988) J. Neurosci. 8, 3658-3669
18 Trimble, W. S., Gray, T. S., Elferink, L. A., Wilson, M. C. and Scheller,R. H. (1990) J. Neurosci. 10, 1380-1387 19 Oyler, G. A. et al. (1989) J. Cell Biol. 109, 3039-3052 20 Forss-Petter, S., Danielson, P., Battenberg, E., Bloom, F. and Sutcliffe, J. G. (1989) J. Mol. Endocrinol. 1, 63-75 21 Schwanzel-Fukuda, M. and Pfaff, D. W. (1989) Nature 338, 161-164
viewpoint; Migraine: a researchfield matured for the basicneurosciences _
.II!i]i !_.I ..............................
Jl
I.
I
_
I _Jlll ...........
IIII .....
III_ H_
Jes Olesen and Lars Edvinsson Progress in migraine research has been rapid in recent years, from both the basic science and the clinical perspectives. A new internationally accepted headache classification with operational diagnostic criteria was published in 1988, eliminating much diagnostic uncertainty. More than a decade of study of regional cerebral blood flow (rCBF) has gradually shown a pathognomonic pattern of abnormalities, probably reflecting spreading cortical depression. Recently it has been shown that pain probably arises from excitation of perivascularpial arterial nociceptors. The innervation and receptor mechanismsof pial and extracranialarteries have been worked out in detail both in animal and humans. Involvement of calcitonin generelated peptide (CGRP) and 5-hydroxytryptamine (5-HT) during migraine attacks has been demonstrated. A new and specific 5-HT1D receptor agonist has proved to be highly effective in treating migraine. Therefore, major research efforts recently have been concentrated on discovering the location and function of 5-HT1o receptors, extra- and intracranJally. Thus, it is now possible to formulate useful neuroscientific researchstrategiesaimed at clarifying migraine mechanisms. Migraine affects approximately 10% of the adult population and about one third of this number have two or more attacks every month 1. Patients with migraine or other headaches constitute the largest single group of patients seen in the neurologist's office. Despite the magnitude of this clinical problem, it has been poorly described. In fact, so little has been known about its pathophysiology that basic neuroscientific studies were, until recently, almost impossible. The situation has improved gradually, and an extensive account of known basic mechanisms of migraine was given in 19882 . This article briefly describes recent scientific breakthroughs, showing that migraine research is now amenable to experimental study, both in the laboratory and in the clinic. The four areas discussed are: the classification and diagnostic criteria of migraine; the pathognomonic alterations of brain blood flow during migraine attacks; the release of perivascular peptides during migraine attacks; and the association of 5-hydroxytryptamine (5-HT) and its receptors with migraine pathogenesis.
Improved classification and diagnostic criteria for headache disorders The first attempt to classify and define headache disorders was made in 1963 by an ad hoc committee of the National Institutes of Health (USA) 3. The TINS, Vol. 14, No. 1, 1991
definition of migraine read as follows: 'attacks of headache, widely varying in intensity, frequency and duration etc.'. Such criteria cannot be used consistently, because they do not specify whether one or more of these clinical characteristics should be present, or how often they should be present. Furthermore, diagnostic terms were ambiguous at that time, and were used differently in different countries. If a disease is poorly defined, it is also difficult to study its mechanisms, since not all of the patients under investigation will actually be suffering from the disorder. However, a committee of the International Headache Society has now completed a new headache classification containing operational diagnostic criteria 4. The headache classification, originally published in English, has also been published in French, German, Italian, Slovenian and Turkish. Translations into Spanish, Portuguese and Japanese are in progress. The classification has been sponsored by the World Federation of Neurology, the Research Group on Migraine and Other Headaches and has, in simplified form, been used by the World Health Organization (WHO) in the most recent International Classification of Diseases (ICD 10). Epidemiological and genetic studies of headache will benefit immensely from the new classification, which will also facilitate pathophysiological and therapeutic studies.
JesOlesenisat the Departmentof Neurology, Gentofte Hospital, Universityof Copenhagen, Copenhagen DK-2900,Denmark, and LarsEdvinssonis at the Departmentof InternalMedicine, UniversityHospital, LundS-22221, Sweden.
Studies of regional cerebral blood flow (rCBF) during migraine attacks For more than a century, scientists have debated the importance of vascular and neurogenic (CNS) components in the pathophysiology of migraine. Despite the lack of suitable methods for studying arterial pulsations, arterial diameter or blood flow, Wolff gathered evidence in favour of the vascular hypothesis during the 1940s and 1950s 5. It was not until the advent of the intra-arterial 133Xe method that quantitative studies of rCBF could be clone during migraine attacks, and it was only with the development of computerized multi-detector systems that measurements with sufficient spatial resolution were obtained. This development has been reviewed previously 6, and therefore the rest of this article focuses on more recent observationsL Early on in an attack of migraine with aura, rCBF is reduced at the posterior pole of the brain; this reduced flow gradually involves larger areas of one
© 1991.ElsevieSci r encePublishersLtd.(UK) 0166-2236/91/$02.00
3
viewpoint 120Migraine without aura
Migraine with aura _* 1 I
100-
~ 80"./¢~. ~-~-
O
E
Q. V a
._l
60-
I
....
n-
~.
O O
_T
z~0-
in vivo.
=
I
.N ~/,
'
20
I i I I
I ...
L._
CF
EJV
EJV
C F
EJV
Y/" EJV
Fig. 1. Changes in CGRP levels during mioraine headache. Open bars represent control measurc,,nents taken either from cubital fossa (CF) or external jugular vein (EJV); hatched bars represent measurements from EJV taken during migraine attacks. An asterisk indicates that there is a significant difference between that measurement and the control Abbreviation: CGRP-LL calcitonin generelated peptide-like immunoreactivity. (Taken, with per-
mission, from Ref. 22.)
hemisphere. Patients usually report their typical aura symptoms, which last for approximately 20-60 min, in this early phase of the attack. After the aura has passed, when the headache begins, and during the early headache phase, rCBF is similarly depressed. Hypoperfusion instead of hyperperfusion occurs at the onset of headache. This contradicts Wolff's hypothesis 5, which is therefore called into question. During the headache phase, blood flow normalizes gradually and then develops into hyperemia (increased blood flow) with no apparent change in headache, nausea and vomiting. In some cases it has been noted that hyperemia outlasts headache, but in most patients measurements have not been performed at the appropriate time to answer this question. The aura symptoms, the headache and the blood-flow reduction all originate in the same hemisphere. In other words, the aura symptoms are perceived contralateral to blood-flow reduction and 4
headache when all of these factors are unilateral. Quite often, however, headache is bilateral despite unilateral aura symptoms and unilateral rCBF reduction. It has recently been found that cerebral blood vessels have bilateral innervation, particularly in the midline 8,9. This means that the headache phase of migraine with aura might be explained by simple activation of perivascular nerves by neurotransmitters, potassium and other substances released from neurones and glial cells, after the migrainous process in the cerebral cortex. In patients suffering from migraine without aura the findings are more difficult to interpret. We have been unable to demonstrate any blood-flow abnormalities in such patients, but other researchers have described hyperemia in the headache phase (for references see Refs 2 and 6). Changes of rCBF in deep cerebral structures, such as the thalamus, hypothalamus, hippocampus or brain stem,could perhaps induce activity in the trigeminovascular system, and thus cause the attack without being detectable by current blood flow techniques According to measurements of human blood flow, and recent experimental work in animals on spreading depression28 and clinical observations 29, rCBF changes in migraine with aura may be secondan/to spreading cortical depression. A recent study using magnetoencephalography demonstrated abnormal DC potentials in patients with migraine, both during an attack and when not experiencing an attack, compared with healthy controls 1°. Although these data are preliminary, they support the hypothesis of spreading depression, a model that is easy to use in animal experiments. Peptidergic involvement in migraine
Cerebral vessels are supplied with perivascular fibres, most of which are autonomic (sympathetic or parasympathetic), but there is a subpopulation of fibres from primary afferent neurones. Evidence for a sensory innervation in humans was first obtained from the responses of patients to stimulation of different cranial vessels11-13, and the distribution of the trigeminal fibres to cranial structures was outlined by Fang14. Later, the autonomic and sensory pathways were outlined and their neurotransmitter content described 15,1°. In brief, the sympathetic fibres originate in the superior cervical ganglion and contain noradrenaline and neuropeptide Y (NPY). The parasympathetic system originates from the sphenopalatine and otic ganglia, and from a number of small ganglionic formations located at the base of the brain. Thus, this system has a comparatively complex origin 17. The sensory system originates in the first division of the trigeminal ganglion and contains calcitonin gene-related peptide (CGRP), substance P and neurokinin A. A contribution from upper dorsal roots to the lower part of the circle of Willis has been described 16. Although the clinical evidence for an involvement of these systems in migraine is modest, a series of recent studies has provided exciting information. In patients with trigeminal neuralgia (tic douloreux) it was found that thermocoagulation of the trigeminal TINS, VoL 14, No. 1, 1991
.........................................
Irl
ganglion caused a significant release of CGRP and substance P into the external jugular vein 18. It is thus possible to measure in vivo the release of peptides from the trigeminal ganglion, a structure known to supply cranial vessels with sensory fibres. Furthermore, the release of these substances was seen only in association with facial flushing ipsilateral to the side of stimulation. Interestingly, such changes are not seen if CGRP and substance P levels are measured in the peripheral circulation (antecubital vein) 19. Moreover, there are no changes in the levels of either substance P or vasoactive intestinal peptide (VIP) in peripheral blood during headache in common or classical migraine 2°. Recently, the levels of several neuropeptides have been studied during headache by local cranial sampling from the external jugular vein 21,22. The study involved 22 patients (6 males and 16 females) who ranged from 22-58 years old. In each case, neuropeptides were determined from both the peripheral circulation (cubital fossa) and the cranial circulation (external jugular vein). In all patients studied, suffering from either classical (n=10) or common migraine (n=12), CGRP was elevated in the cranial but not in the peripheral blood during migraine attacks (Fig. 1). The level of CGRP in patients with classical migraine was greater than that in patients with common migraine. As a group, there were no changes in substance P, neuropeptide Y or VIP during the headache attack. However, two patients from the group with classical migraine had prominent tearing and nasal discharge, and both these had marked elevations in cranial VIP levels. Thus, taken together these findings suggest that CGRP is an important neuropeptide in migraine.
5-HT and migraine A greater understanding of molecular migraine mechanisms has come from the study of 5-HT and its receptor subtypes. This is the final step in a long development. As far back as the 1960s, Sicuteri et a/. suggested that the migraine attack is associated with increased excretion of 5-HT metabolites 23. This work was followed up by studies of the algogenic properties of 5-HT on dorsal hand veins. Lance and associates then demonstrated convincingly that there is a drop in plasma 5-HT levels during migraine attacks, as well as an increased excretion of 5hydroxyindoleacetic acid (5-HIAA) in the urine following an attack 24. The intravenous injection of 5-HT relieved the attack. In keeping with this, a 'selective' 5-HT receptor agonist, sumatriptan {3- [2-(dimethylamino)ethyl]-N-methylindole-5methanesulphonamide}, has been introduced and found to abort acute attacks of migraine headache 25,26. The introduction of this substance was based on a systematic search for migraine drugs acting on subtypes of 5-HT receptors by Humphrey and his associates. Sumatriptan is considered to be a highly specific receptor agonist, stimulating only the 5-HTI subtype (Ref. 26). Recently, this drug has been shown to possess remarkable efficiency in the treatment of migraine and, owing to its high TINS, Vol. 14, No. 1, 1991
IIII
I
Jlllll IIII]i _H .... JUIJlI!I
viewpoint
specificity, it also provides new insight into possible mechanisms of the migraine attack 27. This new development has stimulated experimental work aiming to characterize 5-HT receptor subtypes on blood vessels and neurones, and also to understand the effects of 5-HT receptor activation. Sumatriptan may be used, not only as a migraine drug, but also experimentally in human studies. What happens to CBF when patients are treated with sumatriptan? What happens to CGRP release? Why does sumatriptan work in both migraine and cluster headache, which are otherwise quite separate entities? Where does sumatriptan work? Since it is very water-soluble it almost certainly does not pass the blood-brain barrier. The answers to such human studies and to many other important animal experiments will undoubtedly greatly increase our understanding of the disease.
Selected references 1 Ziegler,D. K. (1986)in Headache Handbook of Neuro/ogy4 (Vol. 48) (Rose,F. C., ed.), pp. 13-22, Elsevier 20lesen, J. and Edvinsson,L., eds (1988) BasicMechanisms of Headache, Elsevier 3 Ad hoc Committee on Classification of Headache (1962) J. Am. Med. Assoc. 179, 717-718 4 Headache Classification Committee of the International Headache Society(1988) Cephalalgia 8 (Suppl. 7), 1-96 5 Wolff, H. G. (1963) Headache and Other Head Pain, pp. 148-153, Oxford UniversityPress 60lesen, J. (1985) Trends Neurosci. 8, 318-321 70lesen, J. et aL Ann. NeuroL (in press) 8 Uddman, R., Edvinsson, L. and Hara, H. (1989) in New Advances in Headache Research (Rose, F. C., ed.), pp. 121-125, Smith-Bond 9 Moskowitz,M. A., Buzzig, M. G., Saskas,O. E. and Linnik, M. D. (1989) Rev. NeuroL 145, 181-193 10 Barkley, G. L., Tepley, N., Simkins, R., Moran, J. E. and Welch, K. M. A. (1989) Cephalal&ia9 (Suppl. 10), 127-128 11 Fay,T. (1932)Ann. OtoL RhinoL LarynEoL 41, 1030-1032 12 Penfield, W. and McNaughton, F. L. (1949) Arch. NeuroL Physiatr. 44, 43-75 13 Ray,B. S. and Wolff, H. G. (1949) Arch. 5urg. 41,813-856 14 Fang, H. C. H. (1961) Arch. NeuroL 4, 651-656 15 Moskowitz,M. A. (1984) Ann. Neurol. 16, 157-168 16 Uddman, R. and Edvinsson, L. (1989) Cerebrovasc. Brain Metab. Rev. 1,230-252 17 Suzuki, N., Hardebo, J. E. and Owman, C. (1988) J. Cereb. Blood Flow/Vletab. 8, 697-712 18 Goadsby, P. J., Edvinsson, L. and Ekman, R. (1988) Ann. NeuroL 23, 193-196 19 Schon, F. et al. (1987)J. Neurol. Neurosurg. Psychiatr. 50, 642-643 20 Blegvad, N., Jensen, K., Fahrenkrug, J., Schaffalitzky de Muckadell, O. B. and Olesen,J. (1986) Cephalalgia5 (Suppl. 2), 352-353 21 Goadsby, P. J., Edvinsson, L. and Ekman, R. (1989) Cephalalgia 9 (Suppl. I0), 292-293 22 Goadsby, P. J., Edvinsson, L. and Ekman, R. (1990) Ann. Neurol. 28, 183-187 23 Sicuteri,F., Testi,A. and Anselmi, B. (1961) Int. Arch. Allergy Appl. Immunol. 19, 55-58 24 Anthony, M., Hinterberger, H. and Lance,J.W. (1969) Res. Clin. Stud. Headache 2, 29-59 25 Doenicke, A., Brand, J. and Perrin, M. J. (1988) Lancet i, 1309-1311 26 Humphrey, P. P. A. et a/. (1988) Br. J. Pharmacol. 94, 1123-1132 27 Lance,J. W., ed. (1989) Cephalalgia 9 (Suppl. 10), 1-464 28 Lauritzen, M. (1987)Acta. Neurol. 5cand. 76 (Suppl. 113), 1-40 29 Olesen,J. (1988) in BasicMechanisms of Headache (Oleson, J. and Edvinsson,L., eds), pp. 353-363, Elsevier 5
