The Pharmacology of Current Anti-Migraine Drugs
Stephen J. Peroutka, M.D., Ph.D.
Department of Neurology, Stanford University Medical Center, Stanford, CA 94305. The first well-documented pharmacological approach-to migraine was in the late nineteenth century, when aspirin was reported to be effective in the acute relief of migraine. Today, aspirin and other anti-inflammatory agents are arguably the most commonly used effective anti-migraine agents. However, the role of anti-inflammatory agents and prostaglandins will not be discussed in detail here. Instead, the focus of this presentation will be on the interactions of anti-migraine agents with the 5-hydroxytryptamine (5-HT) receptor system. Historically, the anti-migraine efficacy of ergots was documented in the 1920's, although their ability to interact with 5-HT receptors was not known until the 1950's.1 On the other hand, methysergide was clearly considered to be a 5-HT-related drug at the time of its first analysis in migraine therapy.2 More recent anti-migraine therapies have included beta-adrenergic blockers (first reported effective in 1966), amitriptyline (1973), and the calcium channel blockers (1981). The goal of this review is to demonstrate that, with the exception of aspirin and related agents, all anti-migraine drugs appear to have important direct, or indirect, effects on the 5-HT receptor system. 5-HT has been implicated also in many other diseases such as depression, anxiety, and sleep disturbances. However, migraine represents the strongest link between this neurotransmitter and a specific neurological illness. As is well known, 5-HT is a potent intracranial vasoconstrictor.3 5-HT levels are known to rise during the aura phase of the attack and decrease at the onset of the headache.4 5-HT can acutely reduce the headache if infused intravenously. While the hypothesis that 5-HT is an important component of the migraine syndrome is neither new nor novel, what is new is the fact that over the past decade very selective pharmacological agents have been developed which can be used as "tools" to pick apart this complicated system. From the receptor viewpoint, the 5-HT system is quite complex. In terms of the current thinking about mammalian 5-HT receptors, some general points should be made. There are multiple subtypes of 5-HT receptor in the brain.5 It is now clear that at least three main "families" of 5-HT receptors exist: 5-HT1, 5-HT2, and 5-HT3. Within each family, there are distinct subtypes. For instance, in the 5-HT1 "family," there are at least three subtypes: 5-HT1A, 5-HT1B, and 5-HT1D receptors. The 5-HT1B receptor exists only in the rat and the mouse,6 whereas the 5-HT1D receptor appears to be the analog of the 5-HT1B receptor in all other species, including human. The location of each of these receptor subtypes differs anatomically, as wall as functionally. The hypothesis to be tested is that acute migraine drugs derive their anti-migraine efficacy from their ability to mimic the effect of 5-HT at 5-HT1D and/or 5-HT1A receptors, while a number of prophylactic anti-migraine drugs appear to share an ability to antagonize 5-HT2 receptors. This hypothesis is summarized in Table 1. Table 1 Hypothetical Role of 5-HT Receptor Subtypes in Migraine Therapy Acute Migraine Treatment Agonist activity at 5-HT1D and/or 5-HT1A receptors Prophylactic Migraine Treatment Antagonist activity at 5-HT2 receptors ACUTE TREATMENTS
Analgesics/Anti-inflammatory agents. The vast majority of mild to moderate migraine attacks can be treated with analgesics such as aspirin or acetaminophen (Table 2). Aspirin has been used in migraine for almost 100 years and remains the drug used most frequently in abortive therapy.7 Non-steroidal anti-inflammatory agents (NSAIAs) are also effective pain relievers and differ from aspirin primarily in cost to the patient. A number of stronger but non-narcotic analgesics have been developed. In general, each preparation contains a combination of aspirin or acetaminophen with sedatives such as butalbital or mild vasoconstrictors like isometheptene. Two of these commercial products (Midrin and Migralam) have been officially indicated for use in migraine. By contrast, Axotal, Fiorinal and Phrenilin are similar compounds which, although officially indicated for muscle contraction headaches, can be
Table 2 Acute Migraine Therapies Drug Trade Name Dosage Aspirin 650 mg q 4 h prn. Acetaminophen 650 mg q 4 h prn. Acetaminophen, Midrin Two capsules at onset Isometheptene and followed by 1 capsule q 1 Dichloralphenazone h prn. (maximum 5 capsules) Acetaminophen, Migralam Two capsules at onset Caffeine and followed by 1 capsule q 1 Isometheptene h prn. (maximum 5 capsules) Aspirin and Butalbital Axotal* One tablet q 4 h prn. (maximum 6 tablets) One or two tablets q 4h prn. (maximum 6 tablets) * Acetaminophen and Phrenilin One or two tablets q 4h Butalbital prn. (6 capsules maximum) Ergotamine and Wigraine; Two tablets at on set then Caffeine Cafergot one tablet q 1/2 h prn. (maximum 6 per day, 10 per week. Ergotamine, Caffeine, Cafergot P-B; One suppository at onset, Belladonna and Wigraine P-B then one hour later if Pentobarbital necessary. (maximum 2 per day or 5 per week) Ergotamine Ergostat; One sublingual tablet at Ergomar onset and q 1/2 h. prn. (maximum 3 per day, 5 per week) Ergotamine MedihalerSingle inhalation at onset Ergotamine followed by q 5 minutes prn. (maximum 6 per day; 15 per week) Dihydroergotamine D.H.E. 45 1 ml IM or IV at onset and q 1 h prn. (maximum 3 ml per day, 6 ml per week) *Not specifically indicated by FDA for migraine. Aspirin, Caffeine and Butalbital
Fiorinal*
used to treat relatively mild migraines. It is not clear if the analgesic and/or anti-inflammatory effects of these drugs are the primary basis for their efficacy in migraine. Ergots. A large proportion of migraine patients will respond to ergot preparations. Ergots have potent vasoconstrictor effects. In the pharmacological sense, the ergots are "dirty" drugs since they interact with a large number of different neuro-transmitter receptors. If the goal is to identify a single receptor subtype which is critically involved in the pathogenesis of migraine, then the ergots are clearly not optimal agents because they interact with multiple receptors. An extensive literature has arisen in the past sixty years concerning the effectiveness of ergots in the acute treatment of migraine headache.1 Indeed, ergot compounds are the most common class of agents used in the acute treatment of severe migraine. Dihydroergotamine (DHE), for example, has been shown to relieve migraine attacks in approximately 70% of patients.8,9 However, ergot compounds also have significant side effects. For example, the drugs often induce nausea, a property which frequently exacerbates the underlying nausea and vomiting of a migraine attack. Chronic use and abuse of ergots may be associated with vasoconstrictive disorders (ergotism). Despite these side effects, ergots are the most widely prescribed acute anti-migraine agents due to their documented efficacy. Sumatriptan. Recently, sumatriptan has been reported to be extremely effective in the acute treatment of migraine. Doenicke et al.10 reported that 2 mg sumatriptan IV completely abolished headache in 71% of 24 migraine attacks and significantly reduced headache in the remaining patients. Only minor side effects (transient pressure in the head, feeling of warmth or tingling) were observed. Although the mechanism of action of sumatriptan in migraine remains unclear, it has been suggested that the drug may selectively stimulate a subpopulation of 5-HT receptors.11 Sumatriptan is now in phase III trials and it continues to show promise as an acute migraine therapy with minimal side effects. I would like to suggest that the ability of sumatriptan to stimulate 5-HT1D receptors might account for its anti-migraine efficacy. The 5-HT1D receptor was initially characterized in 198712 and has been shown to be widespread in human brain.13,14 In fact, 5-HT1D receptors are the most common of 5-HT receptor subtypes observed in the human brain.14 At the same time, vascular studies have identified a "5-HT1-like" receptor in the cranial vasculature which may be identical to the 5-HT1D receptor.15-17 Multiple vascular studies have implicated this "5-HT1-like" receptor in the constriction of cerebral blood vessels. Sumatriptan appears to be an extremely selective agonist of these vascular 5-HT receptors.15 Moreover, recent studies have indicated that sumatriptan is a potent and selective 5-HT1D receptor agent.18,19 Therefore, the "5-HT1-like" receptor in certain cerebral vessels may, in fact, be the 5-HT1D receptor. An attempt was made to determine if DHE and sumatriptan (Figure 1) share a common site of action at the molecular level. The potencies of DHE and sumatriptan were determined by drug competition studies at 13 neurotransmitter receptors using radioligand binding techniques (Table 3).19 DHE is most potent at 5-HT1A receptors (1.2 nM) and is approximately an order of magnitude less potent at 5-HT1D receptors. DHE does display moderate affinity for 5-HT1C and 5-HT2 receptors but is essentially inactive at 5-HT3 receptors. As we reported recently, a different pharmacological profile is observed with sumatriptan.18 The drug is most potent at 5-HT1D receptors and approximately 5-fold less potent at 5-HT1A receptors. Sumatriptan is essentially inactive at all other 5-HT receptor subtypes. DHE interactions
Table 3 Dihydroergotamine and Sumatriptan Interactions with Neurotransmitter Receptor Sites RECEPTOR Ki VALUES (nM) Serotonergic Dihydroergotamine Sumatriptan 5-HT1D 19 ± 3 17 ± 3 5-HT1A 1.2 ± 0.2 100 ± 20 5-HT1C 39 ± 10 > 10,000 5-HT2 78 ± 20 > 10,000 5-HT3 > 10,000 > 10,000 Adrenergic Alpha1 6.6 ± 0.9 > 10,000 AlPha2 3.4 ± 0.5 > 10,000 Beta 960 ± 30 > 10,000 Dopaminergic Dopamine1 700 ± 100 > 10,000 Dopamine2 98 ± 10 > 10,000 Other Sites Muscarinic > 10,000 > 10,000 Benzodiazepine > 10,000 > 10,000
Data are taken from previous publications.18,19 with other biogenic amine receptor sites were also assessed. DHE displays high (i.e. nanomolar) affinity for both alpha1- and alpha2-adrenergic receptors. DHE displays moderate affinity for beta-adrenergic and dopamine receptors. In marked contrast, sumatriptan is totally inactive at each of these neurotransmitter receptor binding sites. Both DHE and sumatriptan are inactive at muscarinic cholinergic and benzodiazepine receptors. DHE and sumatriptan are similar in that they both display high affinity for the 5-HT1D receptor. DHE, however, is also potent at a number of other biogenic amine receptors, such as 5-HT1c, 5-HT2, dopamine2, alpha1- and alpha2-adrenergic receptors. Therefore, sumatriptan is a much more selective agent in that it interacts most potently with 5-HT1D receptors, displays moderate affinity for 5-HT1A receptors, and is essentially inactive at all other receptor sites tested. An unresolved issue is whether the anti-migraine effects of DHE and sumatriptan relate to their vascular and/or neuronal receptor actions. The receptor(s) stimulated by both DHE and sumatriptan has been implicated in the constriction of arteriovenous anastomoses. Under the migraine model proposed by Heyck,20 as yet unknown events lead to the opening of carotid arteriovenous anastomoses in the head. Blood is diverted from the capillary beds, and ischemia and hypoxia result. Based on this hypothesis of migraine, an effective migraine agent would close the shunts and restore blood flow. Indeed, Feniuk et al.15 have shown that sumatriptan is a selective vasoconstrictor of the carotid circulation in the dog. Alternatively, the 5-HT1D receptor appears to be an "auto-receptor" which modulates neurotransmit-ter release.6 Conceivably, 5-HT1D agonists like DHE and sumatriptan may act by blocking the release, at the nerve terminal, of transmitters such as 5-HT, norepinephrine and/or acetylcholine. The ability of potent 5-HT1D agonists to antagonize endogenous transmitter release may, theoretically, account for their efficacy in the acute treatment of migraine. If the early clinical success of sumatriptan is confirmed in future clinical trials, then the 5-HT1D receptor is a likely candidate to play an important role in the acute relief of a migraine attack. As a result, drugs which selectively interact with these sites may allow for the development of effective antimigraine agents with minimal side effects. Analysis of such agents may also provide important insights into the pathophysiological basis of migraine. Other Drugs. A variety of other medications have been recommended for acute migraine attacks.7 For example, corticosteroids have been reported to be effective in some refractory migraine cases. However, the actual use of steroids in acute migraine is quite rare. By contrast, many patients are routinely
treated with potent narcotics, despite the fact that such addicting drugs have no effect on the underlying migraine process. Therefore, the use of narcotics should be limited to severe acute migraine cases in which all other measures have failed. Long-term use of narcotics to treat migraine is never indicated. PROPHYLACTIC TREATMENTS
Clinically available prophylactic anti-migraine drugs are shown in Table 4. Most of these agents also interact with 5-HT receptor subtypes. However, their pharmacological profile is quite different than either sumatriptan or DHE. With the notable exception of the beta-adrenergic antagonists, a common pharmacological property seems to be an ability to block 5-HT2 receptors. Table 4 Prophylactic Migraine Therapies DRUG TRADE NAME DOSAGE Beta-Adrenergic Agents Propranolol Inderal 80 - 320 mg qd Atenolol Tenormin* 50 - 100 mg qd Nadolol Corgard* 40 - 80 mg qd Metoprolol Lopressor* 100 -450 mg qd Timolol Blocadren* 20 -60 mg qd Antidepressants Amitriptyline Elavil* 50 - 150 mg qd Phenelzine Nardil* 15 mg TID Isocarboxazid Marplan* 10 mg QID 5-HTAntagonist Drugs Methysergide Sansert 5 - 8 mg q day Cyproheptadine Periactin* 4-16 mg q day Ergots Ergonovine Ergotrate* 0.2 mg TID Calcium Channel Blockers Diltiazem Cardizem* 180 - 240 mg qd Nifedipine Procardia* 10 - 40 mg TID Verapamil Isoptin*; Calan* 80-120 mg TlD *Not
specifically indicated by FDA for migraine.
5-HT Receptor Antagonists. 5-HT antagonists such as methysergide represent the first class of drugs that were shown effective in migraine prophylaxis.2 Methysergide is an ergot derivative that has complex effects on serotonergic and other neurotransmitter systems. Other serotonin antagonists such as cyproheptadine (Periactin) and pizotifen have also been reported effective in migraine prophylaxis. These drugs all share the ability to potently block 5-HT2 receptors. Amitriptyline. The tricyclic antidepressant, amitriptyline, is an effective prophylactic agent in migraine and this effect is independent of its antidepressant actions.7 Its clinical efficacy is particularly interesting because it is a potent blocker of 5-HT uptake and it is also an antagonist of multiple neuro-transmitter receptors. However, its mechanism of action in migraine prophylaxis remains unknown. Conceivably, either its 5-HT2 receptor antagonism or its ability to block 5-HT uptake may account for its prophylactic anti-migraine efficacy. Increasing 5-HT levels, via uptake blockade, may have a mild analgesic effect in humans. Whether or not this is an important pharmacological feature should be determined following the completion of clinical migraine trials with selective 5-HT uptake blocker drugs such as fluoxetine (Prozac). Fluoxetine (Prozac) is now available in the United States and its efficacy, or lack thereof, in migraine, should provide important insights into the pathophysiology of migraine. For example, if amitriptyline is found to be more effective than fluoxetine in migraine prophylaxis, then blockade of 5-HT2 receptors would be its most likely mechanism of action. Beta-adrenergic antagonists. A review of multiple clinical studies shows that approximately 50-70% of patients will derive some benefit from prophylactic propranolol therapy.7 A variety of other beta-adrenergic agents have been used in the treatment of migraine. Atenolol (Tenormin), metoprolol (Lopressor), nadolol (Corgard) and timolol (Blocadren) appear to be at least as effective as propranolol in migraine prophylaxis. More variable results have been obtained with pindolol (Visken).24-26 By contrast, a number of other beta-adrenergic antagonists (e.g. acebutolol, oxprenolol, alprenolol) do not appear to be effective in migraine therapy. It has been suggested that only pure beta-adrenergic antagonists (without intrinsic sympathomimetic activity) are effective agents in migraine prophylaxis. Some of these drugs do interact with 5-HT1A receptors in both animal18,19,21 and human brains.22 Perhaps more importantly, 5-HT1A selective agonists inhibit intrinsic raphe cell firing.23 The inhibitory effects of 5-HT1A agonists such as 8-OH-DPAT can be blocked by propranolol.23 However, beta-adrenergic agents vary widely in their affinity for the 5-HT1A receptor. For example, pindolol is the most potent drug studied with a Ki value = 4.5 nM. Alprenolol and propranolol are 21and 36-fold, respectively, less potent than pindolol at the 5-HT1A receptor. On the other hand, atenolol and verapamil are inactive at concentrations below 1,000 nM. Therefore, a number of beta-adrenergic anti-migraine agents such as propranolol and timolol display moderate affinity for 5-HT1A receptors. However, no obvious correlation appears to exist between drug affinity and potency in terms of migraine relief For example, pindolol is the most potent anti-migraine agent at the 5-HT1A receptor with a Ki value of 4.5 nM. Clinical studies of pindolol in the prophylactic treatment of migraine have produced conflicting results. In three published reports, the drug was found to be effective in one clinical trial,24 possibly effective in a small subgroup of patients in a second study25 and ineffective in a third study.25 In addition, atenolol is totally inactive at all 5-HT receptor subtypes yet has been reported to be effective in two independent clinical trials27,28 As a
result, the antimigraine efficacy of certain beta-adrenergic agents cannot derive solely from antagonism of the 5-HT receptors. Calcium Channel Blockers. Beginning in 1981, a variety of clinical studies have demonstrated that calcium channel blockers may be effective in migraine phrophylaxis, although more recent data suggest that these drugs may not be as efficacious as initially reported.29,30 It should also be stressed that none of these agents has been specifically approved for use in migraine in the United States. In addition, side effects can be expected to develop in up to 60% of the patients and usually consist of constipation, orthostatic hypotension and red, swollen feet. The drugs are contraindicated in pregnancy. Previously, their anti-migraine efficacy has been attributed to their protective effects during anoxia31 and/or their ability to block intracranial vasoconstriction, irrespective of the constricting agent.32 Surprisingly, certain calcium channel blockers (verapamil, nifedipine) display moderate affinity for the 5-HT2 receptor.22,33,34 Their ability to block 5-HT21 receptors in human brain must now also be considered as a possible therapeutic effect of these drugs. Other Agents. A variety of other medications have been reported to be effective in migraine prophylaxis.7 Chlorpromazine, a phenothiazine with antiemetic properties, has been recommended in headache prophylaxis.35 NSAIAs such as naproxen (Naprosyn) have been reported to be effective migraine prophylactic agents. Monoamine oxidase inhibitors (MACIs) such as phenelzine (Nardil) and isocarboxazid (Marplan) may also be effective in migraine prophylaxis. There is evidence that MACIs' efficacy may relate to their ability to increase levels of endogenous 5-HT. Frequent side effects include orthostatic hypotension, insomnia and nausea. SUMMARY
The major finding of this analysis is that acute antimigraine agents (e.g., ergots, sumatriptan) share high affinity for 5-HT1D receptors. This receptor appears to be present in certain intracranial blood vessels. It is also found on nerve terminals where it inhibits the release of 5-HT and other neurotransmitters. Theoretically, 5-HT1D receptor agonists may acutely inhibit the release of vasoactive and/or pain-inducing substances in the perivascular space. Conceivably, drugs acting at this receptor would stop the progression of this perivascular process. In contrast, a number of prophylactic anti-migraine drugs share a relatively high affinity for 5-HT2 receptors in human brain. Although this receptor is also found in certain blood vessels, it is present throughout the nervous system. The receptor appears to mediate neuronal depolarizations at the cellular level. No hypothesis, at present, readily explains the effectiveness of prophylactic anti-migraine drugs based on this receptor. These data offer a novel approach to the analysis of anti-migraine agents. Drugs could be selected for use in clinical migraine studies based on their selectivity for a specific 5-HT receptor subtype. "Pure" drugs could be chosen which would essentially limit the number of possible sites of action for the drugs. For example, an agent which displays both high affinity and selectivity for 5-HT1D receptors could be clinically evaluated. Its effectiveness, or lack thereof, would indicate the importance of the specific 5-HT receptor site in the pathogenesis of migraine. Further attempts to determine a common mechanism of action for effective anti-migraine agents should facilitate the elucidation of the pathogenesis of this neurological syndrome. In addition, the site of action of these drugs should also be decided in the near future. The raphe nuclei, the trigeminal nuclei, pain pathways and the intracranial circulation are all likely to play important roles in this disorder. The selective pharmacological "tools" which will be developed in the near future should allow for a much clearer understanding of the pathophysiology of migraine. DISCUSSION
Dr. Raskin: We have a few minutes and if there are any questions directed at Dr. Peroutka, we might take one or two right now. Dr. David Codden? Dr. Codden: Thank you for a very comprehensive presentation, Steve. I have a question that has occurred to me recently about the absence, in my experience, of headaches in patients with carcinoid syndrome. I wonder which serotonin receptors are involved because I would think these people, at least some of them, would be very headache prone, but in my 25 years of experience I have not really seen this occur. I would appreciate your comments. Dr. Peroutka: That is a good question. First of all, carcinoid causes a diffuse increase in serotonin levels. So you could argue that migraine derives not from general increase in blood 5-HT or brain 5-HT but from an increased level of 5-HT perhaps in the "wrong places:" maybe perivascularly, maybe in the trigeminal system. So I think that a general increase in serotonin levels is probably not painful. For example, CSF levels show a wide variation between individuals in the levels of 5-HT in the nervous system. The standard deviation of CSF levels of 5-HT and 5-HIAA is huge. The levels do not correlate, to my knowledge, with pain sensitivity. Just as taking nitro-paste doesn't give you a headache despite a sense of pounding blood vessels, I would say that increasing serotonin, per se, is not painful. There must be something else wrong with the system. Dr. Raskin: The diarrhea of carcinoid syndrome is eminently treatable with the migraine drugs, methysergide and propanolol. Dr. Moskowitz: Steve, that was a very important presentation. Two questions: First, is there any information about the use of sumatriptan as a prophylaxis in migraine, because if your scheme is correct, then one would predict it perhaps is not, but if it is,
how would you fit that into the framework? Dr. Peroutka: I don't know of any data, but I'm sure there are people in this room who would know if there are any prophylactic studies. Can anybody answer that? Dr. Moskowitz: We talked about that at breakfast. My understanding is that there are no data regarding prophylaxis. Is that correct? Dr. Humphrey: There are no data on prophylaxis. Dr. Peroutka: On the other hand, it is interesting that some physicians use chronic ergots for migraine prophylaxis. Unlike sumatriptan, ergots are potent 5-HT2 blockers. Ergots are very interesting drugs, since they appear to be agonsits at 5-HT1D receptors and antagonists at 5-HT2receptors. So you can fit the ergots into the hypothesis since they have a multiple pharmacological effects, For example, ergonovine is probably a 5-HT2 blocker but also a 5-HT1D agonist. Dr. Moskowitz: The second question relates to the 3 types and 9 subtypes of 5-HT receptor as defined by binding studies and I am wondering, as a point of information, could you tell us how many of those 9 subtypes have physiological correlates? Dr. Peroutka: Pretty much all of the ones I discussed. They all have functional correlates right now. And three of them have been cloned: 5-HT1A, 5-HT1C, and 5-HT2. Dr. Couch: You mentioned two of the antidepressants, or actually three. You did not mention some of the other ones which have been touted in the past including Trazadone, Ludiomil, Asendin, and Vivactil. Could you comment on those compounds in relation to your scheme? Dr. Peroutka: Doxepin (Sinequan) is a very potent 5-HT2 blocker. In general, antidepressants modulate 5-HT2 receptor density. If you give antidepressants chronically, you see a down regulation of the 5-HT2 receptor. All antidepressants, regardless of class, lower the density and the functional activity of 5-HT2 receptors. We haven't done similar studies with the anti-migraine drugs. It would be interesting to see whether they chronically affect the 5-HT2 system. Whether the migraine drugs do the same thing, I don't know. But that is actually an interesting question and the data may even be available. I just cannot quote it directly. Acknowledgements: I thank Mary Keller for editorial assistance. This work was supported in part by the Stanley Foundation and NIH Grants NS 12151-14 and NS 23560-03. REFERENCES
1.
Rothlin E: Historical development of the ergot therapy of migraine. Int Arch Allergy 7:205-209, 1955.
2.
Sicuteri F: Prophylactic and therapeutic properties of 1-methyl-lysergic acid butanolamide in migraine. Int Arch Allergy 15:300-307, 1959.
3.
Peroutka S J: Vascular serotonin receptors: Correlation with 5-HT1 and 5-HT2 binding sites. Biochem Pharmacol 33:2349-2353, 1984.
4.
Lance JW: Fifty years of migraine research. Aust NZJ Med 18:311-317, 1988.
5.
Peroutka S J: 5-Hydroxytryptamine receptor subtypes. Ann Rev Neurosci 11:45-60, 1988.
6.
Hoyer D, Middlemiss DN: Species differences in the pharmacology of terminal 5-HT autoreceptors in mammalian brain. TIPS 10:130-132, 1989.
7.
Peatfield RC, Fozard JR, Rose FC: Drug treatment of migraine, in Rose FC (ed): Handbook of Clinical Neurology. New York, Elsevier, 1986, Vol 4, pp 173-216.
8.
Callaham M, Raskin N: A controlled study of dihydroergotamine in the treatment of acute migraine headache. Headache 26: 168-171, 1986.
9.
Raskin NH: Repetitive intravenous dihydroergotamine as therapy for intractable migraine. Neurology 36:995-997, 1986.
10.
Doenicke A, Brand J, Perrin VL: Possible benefit of GR43175, a novel 5-HTl-like receptor agonist, for the acute treatment of severe migraine. Lancet ii:1309-1311, 1988.
11.
Humphrey PPA, Feniuk W, Perren MJ, et al: GR43175, a selective agonist for the 5-HT1-like receptor in dog isolated saphenous vein. Br J Pharmacol 94:1123-1132, 1988.
12.
Heuring RE, Peroutka SJ: Characterization of a novel 3H-5-hydroxytryptamine binding site subtype in bovine brain membranes. J Neurosci 7:894-903, 1987.
13.
Waeber C, Dietl MM, Hoyer D, et al: Visualization of a novel serotonin recognition site (5-HT1D) in the human brain by autoradiography. Neurosci Letters 88:11-16, 1988.
14.
Peroutka SJ, Switzer JA, Hamik A: Identification of 5-hydroxytryptamine1D binding sites in human brain membranes. Synapse 3:61-66, 1989.
15.
Feniuk W, Humphrey PPA, Perren MJ: The selective carotid arterial vasoconstrictor action of GR43175 in anaesthetized dogs. Br J Pharmacol 96:83-90, 1989.
16.
Parsons AA, Whalley ET, Feniuk W, et al: 5-HTl-like receptors mediate 5-hydroxytryptamine-induced contraction of human isolated basilar artery. Br J Pharmacol 96:434-449, 1989.
17.
Connor HE, Feniuk W, Humphrey PPA: Characterization of 5-HT receptors mediating contraction of canine and primate basilar artery by use of GR43175, a selective 5-HTl-like receptor agonist. Br J Pharmacol 96:379-387, 1989.
18.
Peroutka S J, McCarthy BG: Sumatriptan (GR 43175) interacts selectively with 5-HT1B and 5-HT1D binding sites. Eur J Pharmacol 163:133-136, 1989.
19.
McCarthy BG, Peroutka S J: Comparative neuropharmacology of dihydroergotamine and sumatriptan (GR 43175). Headache 29:420-422, 1989.
20.
Heyck H: Pathogenesis of migraine. Res Clin Stud Headache 2: 1-28, 1989.
21.
Hiner BC, Roth HL, Peroutka SJ: Antimigraine drug interactions with 5-hydroxytryptamine1A receptors. Ann Neurol 19:511-513, 1986.
22.
Peroutka S J: Antimigraine drug interactions with serotonin receptor subtypes in human brain. Ann Neurol 23;500-504, 1988.
23.
Aghajanian GK, Sprouse JS, Rasmussen K: Physiology of the midbrain serotonin system, in Meltzer HY (ed): Psychopharmacology: The third generation of progress. New York, Raven Press, 1987, pp 141-150.
24.
Anthony M, Lance JW, Somerville B: A comparative trial of pindolol, clonidine and carbamazepine in the interval therapy of migraine. Med J Australia 1: 1343-1346, 1972.
25.
Sjaastad O, Stensrud P: Clinical trial of a beta-receptor blocking agent (LB 46) in migraine prophylaxis. Acta Neurol, Scandinav 48:124-128, 1972.
26.
Ekbom K, Lundberg PO: Clinical trial of LB-46 (d, 1-4-[2-hydroxy-3-isopropylaminopropoxy]) indol, an adrenergic beta-receptor blocking agent, in migraine prophylaxis. Headache 12:15-17, 1972.
27.
Stensrud P, Sjaastad O: Comparative trial of tenormin (atenolol) and inderal (propranolol) in migraine. Headache 20;204-207, 1980.
28.
Forssman B, Lindblad C J, Zbornikova V: Atenolol for migraine prophylaxis. Headache 23: 188-190, 1983.
29.
McArthur JC, Marek K, Pestronk A et al: Nifedipine in the prophylaxis of classic migraine: A crossover, double-masked, placebo-controlled study of headache frequency and side effects. Neurology 39:284-286, 1989.
30.
Albers GW, Simon LT, Hamik A et al: Nifedipine versus propranolol for the initial prophylaxis of migraine. Headache 29:215-218, 1989.
31.
Amery WK: Brain hypoxia: The turning-point in the genesis of the migraine attack? Cephalalgia 2:83-109, 1982.
32.
Peroutka S J: The pharmacology of calcium channel antagonists: A novel class of anti-migraine agents? Headache 23:278-283, 1983.
33.
Affolter H, Burkard WP, Pletscher A: Verapamil, an antagonist at 5-hydroxytryptamine receptors of human blood platelets. Eur J Pharmacol 108:157-162, 1985.
34.
Taylor JE, Defeudis FV: Inhibition of [3H]spiperone binding to 5-HT2 receptors of rat cerebral cortex by the calcium antagonists verapamil and D600. Eur J Pharmacol 106:215-216, 1985.
35.
Caviness VS, Jr, O'Brien P: Headache. New England J Med 446-450, 1980.
Stephen J. Peroutka, M.D., Ph.D.
Department of Neurology, Stanford University Medical Center, Stanford, CA 94305. The first well-documented pharmacological approach-to migraine was in the late nineteenth century, when aspirin was reported to be effective in the acute relief of migraine. Today, aspirin and other anti-inflammatory agents are arguably the most commonly used effective anti-migraine agents. However, the role of anti-inflammatory agents and prostaglandins will not be discussed in detail here. Instead, the focus of this presentation will be on the interactions of anti-migraine agents with the 5-hydroxytryptamine (5-HT) receptor system. Historically, the anti-migraine efficacy of ergots was documented in the 1920's, although their ability to interact with 5-HT receptors was not known until the 1950's.1 On the other hand, methysergide was clearly considered to be a 5-HT-related drug at the time of its first analysis in migraine therapy.2 More recent anti-migraine therapies have included beta-adrenergic blockers (first reported effective in 1966), amitriptyline (1973), and the calcium channel blockers (1981). The goal of this review is to demonstrate that, with the exception of aspirin and related agents, all anti-migraine drugs appear to have important direct, or indirect, effects on the 5-HT receptor system. 5-HT has been implicated also in many other diseases such as depression, anxiety, and sleep disturbances. However, migraine represents the strongest link between this neurotransmitter and a specific neurological illness. As is well known, 5-HT is a potent intracranial vasoconstrictor.3 5-HT levels are known to rise during the aura phase of the attack and decrease at the onset of the headache.4 5-HT can acutely reduce the headache if infused intravenously. While the hypothesis that 5-HT is an important component of the migraine syndrome is neither new nor novel, what is new is the fact that over the past decade very selective pharmacological agents have been developed which can be used as "tools" to pick apart this complicated system. From the receptor viewpoint, the 5-HT system is quite complex. In terms of the current thinking about mammalian 5-HT receptors, some general points should be made. There are multiple subtypes of 5-HT receptor in the brain.5 It is now clear that at least three main "families" of 5-HT receptors exist: 5-HT1, 5-HT2, and 5-HT3. Within each family, there are distinct subtypes. For instance, in the 5-HT1 "family," there are at least three subtypes: 5-HT1A, 5-HT1B, and 5-HT1D receptors. The 5-HT1B receptor exists only in the rat and the mouse,6 whereas the 5-HT1D receptor appears to be the analog of the 5-HT1B receptor in all other species, including human. The location of each of these receptor subtypes differs anatomically, as wall as functionally. The hypothesis to be tested is that acute migraine drugs derive their anti-migraine efficacy from their ability to mimic the effect of 5-HT at 5-HT1D and/or 5-HT1A receptors, while a number of prophylactic anti-migraine drugs appear to share an ability to antagonize 5-HT2 receptors. This hypothesis is summarized in Table 1. Table 1 Hypothetical Role of 5-HT Receptor Subtypes in Migraine Therapy Acute Migraine Treatment Agonist activity at 5-HT1D and/or 5-HT1A receptors Prophylactic Migraine Treatment Antagonist activity at 5-HT2 receptors ACUTE TREATMENTS
Analgesics/Anti-inflammatory agents. The vast majority of mild to moderate migraine attacks can be treated with analgesics such as aspirin or acetaminophen (Table 2). Aspirin has been used in migraine for almost 100 years and remains the drug used most frequently in abortive therapy.7 Non-steroidal anti-inflammatory agents (NSAIAs) are also effective pain relievers and differ from aspirin primarily in cost to the patient. A number of stronger but non-narcotic analgesics have been developed. In general, each preparation contains a combination of aspirin or acetaminophen with sedatives such as butalbital or mild vasoconstrictors like isometheptene. Two of these commercial products (Midrin and Migralam) have been officially indicated for use in migraine. By contrast, Axotal, Fiorinal and Phrenilin are similar compounds which, although officially indicated for muscle contraction headaches, can be
Table 2 Acute Migraine Therapies Drug Trade Name Dosage Aspirin 650 mg q 4 h prn. Acetaminophen 650 mg q 4 h prn. Acetaminophen, Midrin Two capsules at onset Isometheptene and followed by 1 capsule q 1 Dichloralphenazone h prn. (maximum 5 capsules) Acetaminophen, Migralam Two capsules at onset Caffeine and followed by 1 capsule q 1 Isometheptene h prn. (maximum 5 capsules) Aspirin and Butalbital Axotal* One tablet q 4 h prn. (maximum 6 tablets) One or two tablets q 4h prn. (maximum 6 tablets) * Acetaminophen and Phrenilin One or two tablets q 4h Butalbital prn. (6 capsules maximum) Ergotamine and Wigraine; Two tablets at on set then Caffeine Cafergot one tablet q 1/2 h prn. (maximum 6 per day, 10 per week. Ergotamine, Caffeine, Cafergot P-B; One suppository at onset, Belladonna and Wigraine P-B then one hour later if Pentobarbital necessary. (maximum 2 per day or 5 per week) Ergotamine Ergostat; One sublingual tablet at Ergomar onset and q 1/2 h. prn. (maximum 3 per day, 5 per week) Ergotamine MedihalerSingle inhalation at onset Ergotamine followed by q 5 minutes prn. (maximum 6 per day; 15 per week) Dihydroergotamine D.H.E. 45 1 ml IM or IV at onset and q 1 h prn. (maximum 3 ml per day, 6 ml per week) *Not specifically indicated by FDA for migraine. Aspirin, Caffeine and Butalbital
Fiorinal*
used to treat relatively mild migraines. It is not clear if the analgesic and/or anti-inflammatory effects of these drugs are the primary basis for their efficacy in migraine. Ergots. A large proportion of migraine patients will respond to ergot preparations. Ergots have potent vasoconstrictor effects. In the pharmacological sense, the ergots are "dirty" drugs since they interact with a large number of different neuro-transmitter receptors. If the goal is to identify a single receptor subtype which is critically involved in the pathogenesis of migraine, then the ergots are clearly not optimal agents because they interact with multiple receptors. An extensive literature has arisen in the past sixty years concerning the effectiveness of ergots in the acute treatment of migraine headache.1 Indeed, ergot compounds are the most common class of agents used in the acute treatment of severe migraine. Dihydroergotamine (DHE), for example, has been shown to relieve migraine attacks in approximately 70% of patients.8,9 However, ergot compounds also have significant side effects. For example, the drugs often induce nausea, a property which frequently exacerbates the underlying nausea and vomiting of a migraine attack. Chronic use and abuse of ergots may be associated with vasoconstrictive disorders (ergotism). Despite these side effects, ergots are the most widely prescribed acute anti-migraine agents due to their documented efficacy. Sumatriptan. Recently, sumatriptan has been reported to be extremely effective in the acute treatment of migraine. Doenicke et al.10 reported that 2 mg sumatriptan IV completely abolished headache in 71% of 24 migraine attacks and significantly reduced headache in the remaining patients. Only minor side effects (transient pressure in the head, feeling of warmth or tingling) were observed. Although the mechanism of action of sumatriptan in migraine remains unclear, it has been suggested that the drug may selectively stimulate a subpopulation of 5-HT receptors.11 Sumatriptan is now in phase III trials and it continues to show promise as an acute migraine therapy with minimal side effects. I would like to suggest that the ability of sumatriptan to stimulate 5-HT1D receptors might account for its anti-migraine efficacy. The 5-HT1D receptor was initially characterized in 198712 and has been shown to be widespread in human brain.13,14 In fact, 5-HT1D receptors are the most common of 5-HT receptor subtypes observed in the human brain.14 At the same time, vascular studies have identified a "5-HT1-like" receptor in the cranial vasculature which may be identical to the 5-HT1D receptor.15-17 Multiple vascular studies have implicated this "5-HT1-like" receptor in the constriction of cerebral blood vessels. Sumatriptan appears to be an extremely selective agonist of these vascular 5-HT receptors.15 Moreover, recent studies have indicated that sumatriptan is a potent and selective 5-HT1D receptor agent.18,19 Therefore, the "5-HT1-like" receptor in certain cerebral vessels may, in fact, be the 5-HT1D receptor. An attempt was made to determine if DHE and sumatriptan (Figure 1) share a common site of action at the molecular level. The potencies of DHE and sumatriptan were determined by drug competition studies at 13 neurotransmitter receptors using radioligand binding techniques (Table 3).19 DHE is most potent at 5-HT1A receptors (1.2 nM) and is approximately an order of magnitude less potent at 5-HT1D receptors. DHE does display moderate affinity for 5-HT1C and 5-HT2 receptors but is essentially inactive at 5-HT3 receptors. As we reported recently, a different pharmacological profile is observed with sumatriptan.18 The drug is most potent at 5-HT1D receptors and approximately 5-fold less potent at 5-HT1A receptors. Sumatriptan is essentially inactive at all other 5-HT receptor subtypes. DHE interactions
Table 3 Dihydroergotamine and Sumatriptan Interactions with Neurotransmitter Receptor Sites RECEPTOR Ki VALUES (nM) Serotonergic Dihydroergotamine Sumatriptan 5-HT1D 19 ± 3 17 ± 3 5-HT1A 1.2 ± 0.2 100 ± 20 5-HT1C 39 ± 10 > 10,000 5-HT2 78 ± 20 > 10,000 5-HT3 > 10,000 > 10,000 Adrenergic Alpha1 6.6 ± 0.9 > 10,000 AlPha2 3.4 ± 0.5 > 10,000 Beta 960 ± 30 > 10,000 Dopaminergic Dopamine1 700 ± 100 > 10,000 Dopamine2 98 ± 10 > 10,000 Other Sites Muscarinic > 10,000 > 10,000 Benzodiazepine > 10,000 > 10,000
Data are taken from previous publications.18,19 with other biogenic amine receptor sites were also assessed. DHE displays high (i.e. nanomolar) affinity for both alpha1- and alpha2-adrenergic receptors. DHE displays moderate affinity for beta-adrenergic and dopamine receptors. In marked contrast, sumatriptan is totally inactive at each of these neurotransmitter receptor binding sites. Both DHE and sumatriptan are inactive at muscarinic cholinergic and benzodiazepine receptors. DHE and sumatriptan are similar in that they both display high affinity for the 5-HT1D receptor. DHE, however, is also potent at a number of other biogenic amine receptors, such as 5-HT1c, 5-HT2, dopamine2, alpha1- and alpha2-adrenergic receptors. Therefore, sumatriptan is a much more selective agent in that it interacts most potently with 5-HT1D receptors, displays moderate affinity for 5-HT1A receptors, and is essentially inactive at all other receptor sites tested. An unresolved issue is whether the anti-migraine effects of DHE and sumatriptan relate to their vascular and/or neuronal receptor actions. The receptor(s) stimulated by both DHE and sumatriptan has been implicated in the constriction of arteriovenous anastomoses. Under the migraine model proposed by Heyck,20 as yet unknown events lead to the opening of carotid arteriovenous anastomoses in the head. Blood is diverted from the capillary beds, and ischemia and hypoxia result. Based on this hypothesis of migraine, an effective migraine agent would close the shunts and restore blood flow. Indeed, Feniuk et al.15 have shown that sumatriptan is a selective vasoconstrictor of the carotid circulation in the dog. Alternatively, the 5-HT1D receptor appears to be an "auto-receptor" which modulates neurotransmit-ter release.6 Conceivably, 5-HT1D agonists like DHE and sumatriptan may act by blocking the release, at the nerve terminal, of transmitters such as 5-HT, norepinephrine and/or acetylcholine. The ability of potent 5-HT1D agonists to antagonize endogenous transmitter release may, theoretically, account for their efficacy in the acute treatment of migraine. If the early clinical success of sumatriptan is confirmed in future clinical trials, then the 5-HT1D receptor is a likely candidate to play an important role in the acute relief of a migraine attack. As a result, drugs which selectively interact with these sites may allow for the development of effective antimigraine agents with minimal side effects. Analysis of such agents may also provide important insights into the pathophysiological basis of migraine. Other Drugs. A variety of other medications have been recommended for acute migraine attacks.7 For example, corticosteroids have been reported to be effective in some refractory migraine cases. However, the actual use of steroids in acute migraine is quite rare. By contrast, many patients are routinely
treated with potent narcotics, despite the fact that such addicting drugs have no effect on the underlying migraine process. Therefore, the use of narcotics should be limited to severe acute migraine cases in which all other measures have failed. Long-term use of narcotics to treat migraine is never indicated. PROPHYLACTIC TREATMENTS
Clinically available prophylactic anti-migraine drugs are shown in Table 4. Most of these agents also interact with 5-HT receptor subtypes. However, their pharmacological profile is quite different than either sumatriptan or DHE. With the notable exception of the beta-adrenergic antagonists, a common pharmacological property seems to be an ability to block 5-HT2 receptors. Table 4 Prophylactic Migraine Therapies DRUG TRADE NAME DOSAGE Beta-Adrenergic Agents Propranolol Inderal 80 - 320 mg qd Atenolol Tenormin* 50 - 100 mg qd Nadolol Corgard* 40 - 80 mg qd Metoprolol Lopressor* 100 -450 mg qd Timolol Blocadren* 20 -60 mg qd Antidepressants Amitriptyline Elavil* 50 - 150 mg qd Phenelzine Nardil* 15 mg TID Isocarboxazid Marplan* 10 mg QID 5-HTAntagonist Drugs Methysergide Sansert 5 - 8 mg q day Cyproheptadine Periactin* 4-16 mg q day Ergots Ergonovine Ergotrate* 0.2 mg TID Calcium Channel Blockers Diltiazem Cardizem* 180 - 240 mg qd Nifedipine Procardia* 10 - 40 mg TID Verapamil Isoptin*; Calan* 80-120 mg TlD *Not
specifically indicated by FDA for migraine.
5-HT Receptor Antagonists. 5-HT antagonists such as methysergide represent the first class of drugs that were shown effective in migraine prophylaxis.2 Methysergide is an ergot derivative that has complex effects on serotonergic and other neurotransmitter systems. Other serotonin antagonists such as cyproheptadine (Periactin) and pizotifen have also been reported effective in migraine prophylaxis. These drugs all share the ability to potently block 5-HT2 receptors. Amitriptyline. The tricyclic antidepressant, amitriptyline, is an effective prophylactic agent in migraine and this effect is independent of its antidepressant actions.7 Its clinical efficacy is particularly interesting because it is a potent blocker of 5-HT uptake and it is also an antagonist of multiple neuro-transmitter receptors. However, its mechanism of action in migraine prophylaxis remains unknown. Conceivably, either its 5-HT2 receptor antagonism or its ability to block 5-HT uptake may account for its prophylactic anti-migraine efficacy. Increasing 5-HT levels, via uptake blockade, may have a mild analgesic effect in humans. Whether or not this is an important pharmacological feature should be determined following the completion of clinical migraine trials with selective 5-HT uptake blocker drugs such as fluoxetine (Prozac). Fluoxetine (Prozac) is now available in the United States and its efficacy, or lack thereof, in migraine, should provide important insights into the pathophysiology of migraine. For example, if amitriptyline is found to be more effective than fluoxetine in migraine prophylaxis, then blockade of 5-HT2 receptors would be its most likely mechanism of action. Beta-adrenergic antagonists. A review of multiple clinical studies shows that approximately 50-70% of patients will derive some benefit from prophylactic propranolol therapy.7 A variety of other beta-adrenergic agents have been used in the treatment of migraine. Atenolol (Tenormin), metoprolol (Lopressor), nadolol (Corgard) and timolol (Blocadren) appear to be at least as effective as propranolol in migraine prophylaxis. More variable results have been obtained with pindolol (Visken).24-26 By contrast, a number of other beta-adrenergic antagonists (e.g. acebutolol, oxprenolol, alprenolol) do not appear to be effective in migraine therapy. It has been suggested that only pure beta-adrenergic antagonists (without intrinsic sympathomimetic activity) are effective agents in migraine prophylaxis. Some of these drugs do interact with 5-HT1A receptors in both animal18,19,21 and human brains.22 Perhaps more importantly, 5-HT1A selective agonists inhibit intrinsic raphe cell firing.23 The inhibitory effects of 5-HT1A agonists such as 8-OH-DPAT can be blocked by propranolol.23 However, beta-adrenergic agents vary widely in their affinity for the 5-HT1A receptor. For example, pindolol is the most potent drug studied with a Ki value = 4.5 nM. Alprenolol and propranolol are 21and 36-fold, respectively, less potent than pindolol at the 5-HT1A receptor. On the other hand, atenolol and verapamil are inactive at concentrations below 1,000 nM. Therefore, a number of beta-adrenergic anti-migraine agents such as propranolol and timolol display moderate affinity for 5-HT1A receptors. However, no obvious correlation appears to exist between drug affinity and potency in terms of migraine relief For example, pindolol is the most potent anti-migraine agent at the 5-HT1A receptor with a Ki value of 4.5 nM. Clinical studies of pindolol in the prophylactic treatment of migraine have produced conflicting results. In three published reports, the drug was found to be effective in one clinical trial,24 possibly effective in a small subgroup of patients in a second study25 and ineffective in a third study.25 In addition, atenolol is totally inactive at all 5-HT receptor subtypes yet has been reported to be effective in two independent clinical trials27,28 As a
result, the antimigraine efficacy of certain beta-adrenergic agents cannot derive solely from antagonism of the 5-HT receptors. Calcium Channel Blockers. Beginning in 1981, a variety of clinical studies have demonstrated that calcium channel blockers may be effective in migraine phrophylaxis, although more recent data suggest that these drugs may not be as efficacious as initially reported.29,30 It should also be stressed that none of these agents has been specifically approved for use in migraine in the United States. In addition, side effects can be expected to develop in up to 60% of the patients and usually consist of constipation, orthostatic hypotension and red, swollen feet. The drugs are contraindicated in pregnancy. Previously, their anti-migraine efficacy has been attributed to their protective effects during anoxia31 and/or their ability to block intracranial vasoconstriction, irrespective of the constricting agent.32 Surprisingly, certain calcium channel blockers (verapamil, nifedipine) display moderate affinity for the 5-HT2 receptor.22,33,34 Their ability to block 5-HT21 receptors in human brain must now also be considered as a possible therapeutic effect of these drugs. Other Agents. A variety of other medications have been reported to be effective in migraine prophylaxis.7 Chlorpromazine, a phenothiazine with antiemetic properties, has been recommended in headache prophylaxis.35 NSAIAs such as naproxen (Naprosyn) have been reported to be effective migraine prophylactic agents. Monoamine oxidase inhibitors (MACIs) such as phenelzine (Nardil) and isocarboxazid (Marplan) may also be effective in migraine prophylaxis. There is evidence that MACIs' efficacy may relate to their ability to increase levels of endogenous 5-HT. Frequent side effects include orthostatic hypotension, insomnia and nausea. SUMMARY
The major finding of this analysis is that acute antimigraine agents (e.g., ergots, sumatriptan) share high affinity for 5-HT1D receptors. This receptor appears to be present in certain intracranial blood vessels. It is also found on nerve terminals where it inhibits the release of 5-HT and other neurotransmitters. Theoretically, 5-HT1D receptor agonists may acutely inhibit the release of vasoactive and/or pain-inducing substances in the perivascular space. Conceivably, drugs acting at this receptor would stop the progression of this perivascular process. In contrast, a number of prophylactic anti-migraine drugs share a relatively high affinity for 5-HT2 receptors in human brain. Although this receptor is also found in certain blood vessels, it is present throughout the nervous system. The receptor appears to mediate neuronal depolarizations at the cellular level. No hypothesis, at present, readily explains the effectiveness of prophylactic anti-migraine drugs based on this receptor. These data offer a novel approach to the analysis of anti-migraine agents. Drugs could be selected for use in clinical migraine studies based on their selectivity for a specific 5-HT receptor subtype. "Pure" drugs could be chosen which would essentially limit the number of possible sites of action for the drugs. For example, an agent which displays both high affinity and selectivity for 5-HT1D receptors could be clinically evaluated. Its effectiveness, or lack thereof, would indicate the importance of the specific 5-HT receptor site in the pathogenesis of migraine. Further attempts to determine a common mechanism of action for effective anti-migraine agents should facilitate the elucidation of the pathogenesis of this neurological syndrome. In addition, the site of action of these drugs should also be decided in the near future. The raphe nuclei, the trigeminal nuclei, pain pathways and the intracranial circulation are all likely to play important roles in this disorder. The selective pharmacological "tools" which will be developed in the near future should allow for a much clearer understanding of the pathophysiology of migraine. DISCUSSION
Dr. Raskin: We have a few minutes and if there are any questions directed at Dr. Peroutka, we might take one or two right now. Dr. David Codden? Dr. Codden: Thank you for a very comprehensive presentation, Steve. I have a question that has occurred to me recently about the absence, in my experience, of headaches in patients with carcinoid syndrome. I wonder which serotonin receptors are involved because I would think these people, at least some of them, would be very headache prone, but in my 25 years of experience I have not really seen this occur. I would appreciate your comments. Dr. Peroutka: That is a good question. First of all, carcinoid causes a diffuse increase in serotonin levels. So you could argue that migraine derives not from general increase in blood 5-HT or brain 5-HT but from an increased level of 5-HT perhaps in the "wrong places:" maybe perivascularly, maybe in the trigeminal system. So I think that a general increase in serotonin levels is probably not painful. For example, CSF levels show a wide variation between individuals in the levels of 5-HT in the nervous system. The standard deviation of CSF levels of 5-HT and 5-HIAA is huge. The levels do not correlate, to my knowledge, with pain sensitivity. Just as taking nitro-paste doesn't give you a headache despite a sense of pounding blood vessels, I would say that increasing serotonin, per se, is not painful. There must be something else wrong with the system. Dr. Raskin: The diarrhea of carcinoid syndrome is eminently treatable with the migraine drugs, methysergide and propanolol. Dr. Moskowitz: Steve, that was a very important presentation. Two questions: First, is there any information about the use of sumatriptan as a prophylaxis in migraine, because if your scheme is correct, then one would predict it perhaps is not, but if it is,
how would you fit that into the framework? Dr. Peroutka: I don't know of any data, but I'm sure there are people in this room who would know if there are any prophylactic studies. Can anybody answer that? Dr. Moskowitz: We talked about that at breakfast. My understanding is that there are no data regarding prophylaxis. Is that correct? Dr. Humphrey: There are no data on prophylaxis. Dr. Peroutka: On the other hand, it is interesting that some physicians use chronic ergots for migraine prophylaxis. Unlike sumatriptan, ergots are potent 5-HT2 blockers. Ergots are very interesting drugs, since they appear to be agonsits at 5-HT1D receptors and antagonists at 5-HT2receptors. So you can fit the ergots into the hypothesis since they have a multiple pharmacological effects, For example, ergonovine is probably a 5-HT2 blocker but also a 5-HT1D agonist. Dr. Moskowitz: The second question relates to the 3 types and 9 subtypes of 5-HT receptor as defined by binding studies and I am wondering, as a point of information, could you tell us how many of those 9 subtypes have physiological correlates? Dr. Peroutka: Pretty much all of the ones I discussed. They all have functional correlates right now. And three of them have been cloned: 5-HT1A, 5-HT1C, and 5-HT2. Dr. Couch: You mentioned two of the antidepressants, or actually three. You did not mention some of the other ones which have been touted in the past including Trazadone, Ludiomil, Asendin, and Vivactil. Could you comment on those compounds in relation to your scheme? Dr. Peroutka: Doxepin (Sinequan) is a very potent 5-HT2 blocker. In general, antidepressants modulate 5-HT2 receptor density. If you give antidepressants chronically, you see a down regulation of the 5-HT2 receptor. All antidepressants, regardless of class, lower the density and the functional activity of 5-HT2 receptors. We haven't done similar studies with the anti-migraine drugs. It would be interesting to see whether they chronically affect the 5-HT2 system. Whether the migraine drugs do the same thing, I don't know. But that is actually an interesting question and the data may even be available. I just cannot quote it directly. Acknowledgements: I thank Mary Keller for editorial assistance. This work was supported in part by the Stanley Foundation and NIH Grants NS 12151-14 and NS 23560-03. REFERENCES
1.
Rothlin E: Historical development of the ergot therapy of migraine. Int Arch Allergy 7:205-209, 1955.
2.
Sicuteri F: Prophylactic and therapeutic properties of 1-methyl-lysergic acid butanolamide in migraine. Int Arch Allergy 15:300-307, 1959.
3.
Peroutka S J: Vascular serotonin receptors: Correlation with 5-HT1 and 5-HT2 binding sites. Biochem Pharmacol 33:2349-2353, 1984.
4.
Lance JW: Fifty years of migraine research. Aust NZJ Med 18:311-317, 1988.
5.
Peroutka S J: 5-Hydroxytryptamine receptor subtypes. Ann Rev Neurosci 11:45-60, 1988.
6.
Hoyer D, Middlemiss DN: Species differences in the pharmacology of terminal 5-HT autoreceptors in mammalian brain. TIPS 10:130-132, 1989.
7.
Peatfield RC, Fozard JR, Rose FC: Drug treatment of migraine, in Rose FC (ed): Handbook of Clinical Neurology. New York, Elsevier, 1986, Vol 4, pp 173-216.
8.
Callaham M, Raskin N: A controlled study of dihydroergotamine in the treatment of acute migraine headache. Headache 26: 168-171, 1986.
9.
Raskin NH: Repetitive intravenous dihydroergotamine as therapy for intractable migraine. Neurology 36:995-997, 1986.
10.
Doenicke A, Brand J, Perrin VL: Possible benefit of GR43175, a novel 5-HTl-like receptor agonist, for the acute treatment of severe migraine. Lancet ii:1309-1311, 1988.
11.
Humphrey PPA, Feniuk W, Perren MJ, et al: GR43175, a selective agonist for the 5-HT1-like receptor in dog isolated saphenous vein. Br J Pharmacol 94:1123-1132, 1988.
12.
Heuring RE, Peroutka SJ: Characterization of a novel 3H-5-hydroxytryptamine binding site subtype in bovine brain membranes. J Neurosci 7:894-903, 1987.
13.
Waeber C, Dietl MM, Hoyer D, et al: Visualization of a novel serotonin recognition site (5-HT1D) in the human brain by autoradiography. Neurosci Letters 88:11-16, 1988.
14.
Peroutka SJ, Switzer JA, Hamik A: Identification of 5-hydroxytryptamine1D binding sites in human brain membranes. Synapse 3:61-66, 1989.
15.
Feniuk W, Humphrey PPA, Perren MJ: The selective carotid arterial vasoconstrictor action of GR43175 in anaesthetized dogs. Br J Pharmacol 96:83-90, 1989.
16.
Parsons AA, Whalley ET, Feniuk W, et al: 5-HTl-like receptors mediate 5-hydroxytryptamine-induced contraction of human isolated basilar artery. Br J Pharmacol 96:434-449, 1989.
17.
Connor HE, Feniuk W, Humphrey PPA: Characterization of 5-HT receptors mediating contraction of canine and primate basilar artery by use of GR43175, a selective 5-HTl-like receptor agonist. Br J Pharmacol 96:379-387, 1989.
18.
Peroutka S J, McCarthy BG: Sumatriptan (GR 43175) interacts selectively with 5-HT1B and 5-HT1D binding sites. Eur J Pharmacol 163:133-136, 1989.
19.
McCarthy BG, Peroutka S J: Comparative neuropharmacology of dihydroergotamine and sumatriptan (GR 43175). Headache 29:420-422, 1989.
20.
Heyck H: Pathogenesis of migraine. Res Clin Stud Headache 2: 1-28, 1989.
21.
Hiner BC, Roth HL, Peroutka SJ: Antimigraine drug interactions with 5-hydroxytryptamine1A receptors. Ann Neurol 19:511-513, 1986.
22.
Peroutka S J: Antimigraine drug interactions with serotonin receptor subtypes in human brain. Ann Neurol 23;500-504, 1988.
23.
Aghajanian GK, Sprouse JS, Rasmussen K: Physiology of the midbrain serotonin system, in Meltzer HY (ed): Psychopharmacology: The third generation of progress. New York, Raven Press, 1987, pp 141-150.
24.
Anthony M, Lance JW, Somerville B: A comparative trial of pindolol, clonidine and carbamazepine in the interval therapy of migraine. Med J Australia 1: 1343-1346, 1972.
25.
Sjaastad O, Stensrud P: Clinical trial of a beta-receptor blocking agent (LB 46) in migraine prophylaxis. Acta Neurol, Scandinav 48:124-128, 1972.
26.
Ekbom K, Lundberg PO: Clinical trial of LB-46 (d, 1-4-[2-hydroxy-3-isopropylaminopropoxy]) indol, an adrenergic beta-receptor blocking agent, in migraine prophylaxis. Headache 12:15-17, 1972.
27.
Stensrud P, Sjaastad O: Comparative trial of tenormin (atenolol) and inderal (propranolol) in migraine. Headache 20;204-207, 1980.
28.
Forssman B, Lindblad C J, Zbornikova V: Atenolol for migraine prophylaxis. Headache 23: 188-190, 1983.
29.
McArthur JC, Marek K, Pestronk A et al: Nifedipine in the prophylaxis of classic migraine: A crossover, double-masked, placebo-controlled study of headache frequency and side effects. Neurology 39:284-286, 1989.
30.
Albers GW, Simon LT, Hamik A et al: Nifedipine versus propranolol for the initial prophylaxis of migraine. Headache 29:215-218, 1989.
31.
Amery WK: Brain hypoxia: The turning-point in the genesis of the migraine attack? Cephalalgia 2:83-109, 1982.
32.
Peroutka S J: The pharmacology of calcium channel antagonists: A novel class of anti-migraine agents? Headache 23:278-283, 1983.
33.
Affolter H, Burkard WP, Pletscher A: Verapamil, an antagonist at 5-hydroxytryptamine receptors of human blood platelets. Eur J Pharmacol 108:157-162, 1985.
34.
Taylor JE, Defeudis FV: Inhibition of [3H]spiperone binding to 5-HT2 receptors of rat cerebral cortex by the calcium antagonists verapamil and D600. Eur J Pharmacol 106:215-216, 1985.
35.
Caviness VS, Jr, O'Brien P: Headache. New England J Med 446-450, 1980.
