Cluster Headache: Diagnosis and Management Lee Kudrow, M.D. Director, California Medical Clinic for Headache, Encino, California. Reprint requests to: Lee Kudrow, M.D., California Medical Clinic for Headache Inc., 16542 Ventura Boulevard, Encino, California 91436. To preface this discussion of cluster headache, I would like to present an account of a cluster headache, written in the first person. Short of experiencing such an attack, the reader may witness the intensity of anticipation, pain, anxiety and frustration experienced by the cluster victim. "Following a period of perhaps several hours, during which time I feel quite elated and energetic, I experience a fullness in my ears, somewhat more on the right side than the left, having a character not unlike that which occurs during rapid descent in an airplane or elevator. I next become aware of a dull discomfort, an extension of ear fullness at the base of my skull - further extending over the entire head, on both sides, although somewhat more on the right. At this point, two or three minutes have elapsed; seemingly short but long enough for me to know that indeed a 'cluster' has begun and will ultimately get worse. Such anticipation causes me considerable consternation regarding any decision to continue my activities or cancel plans and find a place to be alone; giving way to a slowly increasing anxiety, fear, panic and withdrawal. I become aware of myself 'listening' for changes in my head. Is the cluster prematurely aborting itself, progressing further, or unchanging? A sudden stab, only fleeting, strikes my temple, then again - somewhere near the apex of my skull and upper molars in my face - always on the right side. It strikes me again, deep into the skull base and as quickly, changes location to a small area above my eyebrow. My nose is stuffed yet runs simultaneously. If I could sneeze I feel the attack would end. But in spite of all tricks I find myself unable to induce sneezing. "While the sharp stabs continue in this fashion, a slow crescendo of dull pain presents itself in an area of a hand's length and breadth over the eye and temporal region. The area of pain narrows into a smaller area but as if magnified, enlarges in intensity. I find myself bending my neck downward, though slightly, as if my head is being gently pushed from behind. My neck up to the base of my skull is tight and feels as if I were wearing a neck collar. I am compelled to remove my tie and loosen my shirt collar even though I know that it will not offer me even a modicum of relief. "In an attempt to alter this persistent discomfort, I drop my head between my legs, while seated. My face and eyes seem to fill with fluid but the pain remains unchanged. In spite of my suntan, as I look into the mirror, a gaunt, sickly pale face peers back. My right lid is only slightly drooping and the white of my eye is charted with many red vessels, giving it an overall color of pink. "Having difficulty standing in one place too long, I leave the mirror to continue my alternating pacing and sitting. "As usual, I am struck with the additional fear that the pain will never end, but dismiss it as impossible, since even if it were the case, I would surely kill myself. "The pain, now located somewhere behind my eye and slightly above, worsens. The pain is best described as a 'force' pushing with such incredible power through my eye that my head appears to be moving backward, yielding its resistance. The 'force' wanes and waxes, but the duration of successive exacerbations seems to increase. The cluster is at its peak, which is celebrated by an outpouring of tears from my right eye only. I have now been in cluster for 35 minutes - 10 minutes at its peak. "My wife peeks into the room in which I hold forth. I look up and see her expression of pity, frustration and helplessness. She sees my tortured face as I have seen it in the mirror at this stage before; a drooling mouth, agape, gray face wet on one side, an almost closed eyelid, and smelling of pain and anguish. She closes the door and leaves, feeling hurt for me, anger for the stupidity of medical science, and guilt - since deep within her mind is the suspicion that she is the cause of my suffering. "I cry for her, but more for myself. The pain is so incredible. Suddenly I am overwhelmed by a fury. I lift a chair high above my head and crash it to the floor. With a doubled fist I strike the wall. The pain persists. "Waning periods soon become longer in duration and I allow myself to suspect that the peak is behind me - but cautiously, since I have been too often disappointed. "Indeed, the pain is ending. The descent from the mountain of pain is rapid. The 'force' is gone. Only
Authors Romberg1
Table 1 Cluster Headache Eponyms, Misnomers and Other Appellations Nomenclature Date Eponyms Other Names 1840 1867 1878 1910
Description only
Möllendorff2 Eulenburg3 Sluder4 Bing5
1913
Bing's Headache Bing's Syndrome
Harris6 Harris7 Vail8 Gardner et al9
1926 1936 1932 1947
Horton et al10,11
1939 1952 1952
Kunkle et a112
Sluder's Syndrome
Horton's Headache Horton's Syndrome
Red Migraine Angioparalytic Hemicrania Sphenopalatine Neuralgia, Lower-Half Headache Erythroprosopalgia Migrainous Neuralgia Ciliary Neuralgia Vidian Neuralgia Greater Superficial Petrosal Neuralgia Erythromelalgia Histaminic Cephalgia Cluster Headache
severe pain remains. My nose and eye continue to run. The road back, as with all travel, covers the same territory - but faster. Stabbing, easily tolerated pain is felt. Then gone. Dull, aching fullness, neck stiffness - all disappear in turn, to be replaced by a welcome sensation of pins and needles over the right scalp area - not unlike after one's leg has been 'asleep.' Thus my head has awakened after a nightmare of torment. "Eye and nose dry, I let out a sigh. I collect my pile of wet tissues strewn all over the floor and deposit them in a wastepaper basket. The innocent chair now uprighted, I rub my slightly bruised fist. Thus, having ended the battle and cleaned up its field, I open the door and enter my pain-free world - until tomorrow." INTRODUCTION Cluster headache is known by many eponyms and other names.1-11 It was probably first described by Romberg,1 later by Eulenberg and more precisely by Harris,6,7 from whom the names "ciliary and migrainous neuralgia" derived. It is unlikely that Möllendorff,2 Sluder4 or Bing5 had described cluster headache; although they are often referenced in the literature in this regard. Because of Horton,10,11 this disorder became better known; and is often referred to as Horton's syndrome. Friedman13 had legitimatized the term "cluster headache" after Kunkle12 had described this condition as having a "clustering" character (Table 1 ). Classification of Cluster. There are essentially three major types of cluster headache (Table 2). Episodic, otherwise known as periodic, is the typical form. Attacks will occur daily for several weeks or months and then stop abruptly, to be followed by a rather extended remission period. Chronic cluster, on the other hand, is defined by the absence of a remission period. Primary chronic cluster refers to those individuals in whom remission periods have never occurred. If a patient with episodic cluster becomes chronic, he is described as having secondary chronic cluster headaches. Chronic paroxysmal hemicrania, or CPH, is a syndrome recently described by Sjaastad.14 It appears to be a form of chronic cluster in which attacks occur even more frequently, affecting primarily middle-aged women, and most curiously, it is responsive to aspirin, and more dramatically, to indomethacin. The entire chronic cluster group differs from episodic cluster in that attacks are more frequent and response to prophylactic medication is somewhat refractory. The third classification of cluster headache may be called atypical variants which include cluster-migraine and cluster-vertigo. Cluster-migraine appears to be a transitional state or link between cluster headache syndrome and migraine. It was most recently Table 2 Classification of Cluster Headache 1. Episodic (periodic) 2. Chronic A. Primary B. Secondary C. Chronic paroxysmal hemicrania (CPH) 3. Atypical Variant A. Cluster-migraine B. Cluster-vertigo
described by Medina and Diamond15 in which several cases were presented. It is characterized as migraine-type headache having the periodicity of the cluster syndrome. In other cases, the reverse of this is true. Cluster-vertigo, described by Gordon Gilbert16 represents patients who experience Meniere's-type vertigo, associated solely with cluster headaches. He described eight such cases. Incidence. The incidence of cluster headache remains unknown. Many investigators believe the incidence in the general population is under 1%. In clinic populations the incidence ratio of migraine: cluster is approximately 10:1 (Table 3). Table 3 Several Reports on the Incidence of Cluster Headache and Migraine No. Patients Ratio Author Migraine Cluster Migraine: Cluster Lieder17 52 4 13.0:1 Carroll18 89 16 5.6:1 Balla & Walton19 399 28 14.3:1 Ekbom20 400 16 25.0:1 Lance et al21 612 13 47.1:1 Heyck22 1890 48 39.4:1 Friedman23 2667 237 11.3:1 Kudrow 2835 425 6.7:1
Male:Female Ratios. Estimations of the male: female ratio for cluster headache in various clinic populations yield a range of between 4.5:1 to 6.5:1, indicating considerable male predominance (Table 4). Table 4 Male: Female Ratio and Mean Age at Onset of Cluster Headache Patients Onset Ratio Author Date No. Age Range M:F Friedman & Mikroupoulos13 1958 50 28 11-44 4.5:1 Ekbom24 1970 105 27.5 10-61 5.6:1 Lance & Anthony21 1971 60 8-62 6.5:1 Kudrow Present 425 29.6 1-63 5.1: 1
CLINICAL DESCRIPTION OF CLUSTER HEADACHE Cluster periods are defined as those periods during which attacks occur, generally lasting between 6 to 12 weeks. Remission periods have an average duration of approximately 12 months. There may be considerable variation. Attacks occur with a frequency of approximately one to three times a day, each lasting 45 minutes in duration, always unilateral, occulotemporal or occulofrontal in location, excruciating in severity, boring and nonthrobbing in character. The associated symptoms are also unilateral and consist of lacrimation, rhinorrhea or nasal stuffiness and partial Horner's syndrome which includes unilateral ptosis and miosis. Characteristically, vasodilator medications such as nitroglycerin and histamine will induce cluster attacks. Ekbom25 was able to induce an acute cluster attack in all of his subjects diagnosed as having cluster headache, using 1 mg. nitroglycerin sublingually. Frequently, but not always, alcohol will induce an acute cluster attack while the patient is in an active cluster period. Attacks are commonly induced upon awakening from a nap in the afternoon or upon awakening from sleep during the night, most commonly approximately 90 minutes after falling asleep. Dexter26 has shown that cluster attacks which occur during sleep hours are associated with the REM state. A major criterion used in the diagnosis of cluster headache at the California Medical Clinic for Headache is the behavior of the patient during the attack. Pacing, walking, sitting and rocking during the attack are activities which we consider pathognomonic of this disorder. There is no other primary headache disorder in which such behavior is an associated feature (Table 5). Periodicity. The onset of cluster periods has been reported by Ekbom24 and others to occur with seasonal periodicity. Specifically, spring and autumn are seasons of high incidence. Lance27 did not find this to be the case, since among patients who associated their cluster periods with seasons, the distribution was equally divided between all four
Cluster Periods Remissions Attacks Frequency Duration Location
Table 5 Cluster Headache Attack Profile 6-12 weeks* 12 months*
Severity Character Associated Symptoms
Induction Behavior in Attack * Average
values
1-3 day* 45 minutes* Unilateral Oculotemporal Excruciating Boring, nonthrobbing Unilateral Lacrimation Rhinorrhea Partial Horner's Vasodilators, alcohol, REM Walking, sitting, rocking
seasons. More curious is the attack periodicity in cluster headache. As noted earlier, an individual often awakens with an attack, from a nap, or two hours after falling asleep; it strikes especially during relaxation. The circadian accuracy in which these attacks occur is not at all understood. Attacks which occur 24 hours apart recur at exactly the same time. Attacks which occur twice a day are generally 12 hours apart, but more importantly again, at the same time of day. Physical Characteristics. In 1969, John Graham28 noted and reported that a great many cluster-headache individuals had certain and specific facial features. He described these as having a "leonine" appearance, deep skin furrows (especially the nasolabial and glabellar folds), and forehead wrinkles. He also described the presence of telangiectasia, often observed across the bridge of the nose. We have also noted, as had Dr. Graham, that there appears to be narrowing of the palpebral fissures. Quite often there are asymmetrical skin wrinkles and "orange-peel," thick skin; in all, the appearance of an alcoholic. However, Graham states that some of his patients who were nondrinkers had the typical cluster facies. He found that although women did not have all of these characteristics, they were somewhat masculine-looking. We have noted rather similar appearances among our women with cluster headache, although I am not sure it can be categorized as masculine. Moreover, we have found that a number of our male patients had characteristics more descriptive of acromegaly, that is, acromegalous in appearance (Table 6). Other Characteristics. Among 30 male patients with cluster headache, we found the mean height difference to be three inches taller than matched male controls. More curiously, we found an increased incidence of hazel eye color (38%). Indeed, cluster patients smoked more often and more cigarettes per Table 6 Physical Characteristics Often Observed Among Cluster Males FACIAL Ruddy Complexion Deep Furrows "Orange Peel" Skin Telangiectasia Narrowed Palpebral Fissures Asymmetric Creases Broad Chin, Skull Leonine Appearance GENERAL Rugged Appearance Tall, Trim Rarely obese Hazel Eye Color (1/3)
Table 7 Comparison of Cluster and Non-Cluster Groups with Respect to Stature, Eye Color, Hemoglobin, Smoking and Alcohol Use Factor Cluster Non-Cluster Test P-Values Height (in.) 71.4 68.5 t
Authors Romberg1
Table 1 Cluster Headache Eponyms, Misnomers and Other Appellations Nomenclature Date Eponyms Other Names 1840 1867 1878 1910
Description only
Möllendorff2 Eulenburg3 Sluder4 Bing5
1913
Bing's Headache Bing's Syndrome
Harris6 Harris7 Vail8 Gardner et al9
1926 1936 1932 1947
Horton et al10,11
1939 1952 1952
Kunkle et a112
Sluder's Syndrome
Horton's Headache Horton's Syndrome
Red Migraine Angioparalytic Hemicrania Sphenopalatine Neuralgia, Lower-Half Headache Erythroprosopalgia Migrainous Neuralgia Ciliary Neuralgia Vidian Neuralgia Greater Superficial Petrosal Neuralgia Erythromelalgia Histaminic Cephalgia Cluster Headache
severe pain remains. My nose and eye continue to run. The road back, as with all travel, covers the same territory - but faster. Stabbing, easily tolerated pain is felt. Then gone. Dull, aching fullness, neck stiffness - all disappear in turn, to be replaced by a welcome sensation of pins and needles over the right scalp area - not unlike after one's leg has been 'asleep.' Thus my head has awakened after a nightmare of torment. "Eye and nose dry, I let out a sigh. I collect my pile of wet tissues strewn all over the floor and deposit them in a wastepaper basket. The innocent chair now uprighted, I rub my slightly bruised fist. Thus, having ended the battle and cleaned up its field, I open the door and enter my pain-free world - until tomorrow." INTRODUCTION Cluster headache is known by many eponyms and other names.1-11 It was probably first described by Romberg,1 later by Eulenberg and more precisely by Harris,6,7 from whom the names "ciliary and migrainous neuralgia" derived. It is unlikely that Möllendorff,2 Sluder4 or Bing5 had described cluster headache; although they are often referenced in the literature in this regard. Because of Horton,10,11 this disorder became better known; and is often referred to as Horton's syndrome. Friedman13 had legitimatized the term "cluster headache" after Kunkle12 had described this condition as having a "clustering" character (Table 1 ). Classification of Cluster. There are essentially three major types of cluster headache (Table 2). Episodic, otherwise known as periodic, is the typical form. Attacks will occur daily for several weeks or months and then stop abruptly, to be followed by a rather extended remission period. Chronic cluster, on the other hand, is defined by the absence of a remission period. Primary chronic cluster refers to those individuals in whom remission periods have never occurred. If a patient with episodic cluster becomes chronic, he is described as having secondary chronic cluster headaches. Chronic paroxysmal hemicrania, or CPH, is a syndrome recently described by Sjaastad.14 It appears to be a form of chronic cluster in which attacks occur even more frequently, affecting primarily middle-aged women, and most curiously, it is responsive to aspirin, and more dramatically, to indomethacin. The entire chronic cluster group differs from episodic cluster in that attacks are more frequent and response to prophylactic medication is somewhat refractory. The third classification of cluster headache may be called atypical variants which include cluster-migraine and cluster-vertigo. Cluster-migraine appears to be a transitional state or link between cluster headache syndrome and migraine. It was most recently Table 2 Classification of Cluster Headache 1. Episodic (periodic) 2. Chronic A. Primary B. Secondary C. Chronic paroxysmal hemicrania (CPH) 3. Atypical Variant A. Cluster-migraine B. Cluster-vertigo
described by Medina and Diamond15 in which several cases were presented. It is characterized as migraine-type headache having the periodicity of the cluster syndrome. In other cases, the reverse of this is true. Cluster-vertigo, described by Gordon Gilbert16 represents patients who experience Meniere's-type vertigo, associated solely with cluster headaches. He described eight such cases. Incidence. The incidence of cluster headache remains unknown. Many investigators believe the incidence in the general population is under 1%. In clinic populations the incidence ratio of migraine: cluster is approximately 10:1 (Table 3). Table 3 Several Reports on the Incidence of Cluster Headache and Migraine No. Patients Ratio Author Migraine Cluster Migraine: Cluster Lieder17 52 4 13.0:1 Carroll18 89 16 5.6:1 Balla & Walton19 399 28 14.3:1 Ekbom20 400 16 25.0:1 Lance et al21 612 13 47.1:1 Heyck22 1890 48 39.4:1 Friedman23 2667 237 11.3:1 Kudrow 2835 425 6.7:1
Male:Female Ratios. Estimations of the male: female ratio for cluster headache in various clinic populations yield a range of between 4.5:1 to 6.5:1, indicating considerable male predominance (Table 4). Table 4 Male: Female Ratio and Mean Age at Onset of Cluster Headache Patients Onset Ratio Author Date No. Age Range M:F Friedman & Mikroupoulos13 1958 50 28 11-44 4.5:1 Ekbom24 1970 105 27.5 10-61 5.6:1 Lance & Anthony21 1971 60 8-62 6.5:1 Kudrow Present 425 29.6 1-63 5.1: 1
CLINICAL DESCRIPTION OF CLUSTER HEADACHE Cluster periods are defined as those periods during which attacks occur, generally lasting between 6 to 12 weeks. Remission periods have an average duration of approximately 12 months. There may be considerable variation. Attacks occur with a frequency of approximately one to three times a day, each lasting 45 minutes in duration, always unilateral, occulotemporal or occulofrontal in location, excruciating in severity, boring and nonthrobbing in character. The associated symptoms are also unilateral and consist of lacrimation, rhinorrhea or nasal stuffiness and partial Horner's syndrome which includes unilateral ptosis and miosis. Characteristically, vasodilator medications such as nitroglycerin and histamine will induce cluster attacks. Ekbom25 was able to induce an acute cluster attack in all of his subjects diagnosed as having cluster headache, using 1 mg. nitroglycerin sublingually. Frequently, but not always, alcohol will induce an acute cluster attack while the patient is in an active cluster period. Attacks are commonly induced upon awakening from a nap in the afternoon or upon awakening from sleep during the night, most commonly approximately 90 minutes after falling asleep. Dexter26 has shown that cluster attacks which occur during sleep hours are associated with the REM state. A major criterion used in the diagnosis of cluster headache at the California Medical Clinic for Headache is the behavior of the patient during the attack. Pacing, walking, sitting and rocking during the attack are activities which we consider pathognomonic of this disorder. There is no other primary headache disorder in which such behavior is an associated feature (Table 5). Periodicity. The onset of cluster periods has been reported by Ekbom24 and others to occur with seasonal periodicity. Specifically, spring and autumn are seasons of high incidence. Lance27 did not find this to be the case, since among patients who associated their cluster periods with seasons, the distribution was equally divided between all four
Cluster Periods Remissions Attacks Frequency Duration Location
Table 5 Cluster Headache Attack Profile 6-12 weeks* 12 months*
Severity Character Associated Symptoms
Induction Behavior in Attack * Average
values
1-3 day* 45 minutes* Unilateral Oculotemporal Excruciating Boring, nonthrobbing Unilateral Lacrimation Rhinorrhea Partial Horner's Vasodilators, alcohol, REM Walking, sitting, rocking
seasons. More curious is the attack periodicity in cluster headache. As noted earlier, an individual often awakens with an attack, from a nap, or two hours after falling asleep; it strikes especially during relaxation. The circadian accuracy in which these attacks occur is not at all understood. Attacks which occur 24 hours apart recur at exactly the same time. Attacks which occur twice a day are generally 12 hours apart, but more importantly again, at the same time of day. Physical Characteristics. In 1969, John Graham28 noted and reported that a great many cluster-headache individuals had certain and specific facial features. He described these as having a "leonine" appearance, deep skin furrows (especially the nasolabial and glabellar folds), and forehead wrinkles. He also described the presence of telangiectasia, often observed across the bridge of the nose. We have also noted, as had Dr. Graham, that there appears to be narrowing of the palpebral fissures. Quite often there are asymmetrical skin wrinkles and "orange-peel," thick skin; in all, the appearance of an alcoholic. However, Graham states that some of his patients who were nondrinkers had the typical cluster facies. He found that although women did not have all of these characteristics, they were somewhat masculine-looking. We have noted rather similar appearances among our women with cluster headache, although I am not sure it can be categorized as masculine. Moreover, we have found that a number of our male patients had characteristics more descriptive of acromegaly, that is, acromegalous in appearance (Table 6). Other Characteristics. Among 30 male patients with cluster headache, we found the mean height difference to be three inches taller than matched male controls. More curiously, we found an increased incidence of hazel eye color (38%). Indeed, cluster patients smoked more often and more cigarettes per Table 6 Physical Characteristics Often Observed Among Cluster Males FACIAL Ruddy Complexion Deep Furrows "Orange Peel" Skin Telangiectasia Narrowed Palpebral Fissures Asymmetric Creases Broad Chin, Skull Leonine Appearance GENERAL Rugged Appearance Tall, Trim Rarely obese Hazel Eye Color (1/3)
Table 7 Comparison of Cluster and Non-Cluster Groups with Respect to Stature, Eye Color, Hemoglobin, Smoking and Alcohol Use Factor Cluster Non-Cluster Test P-Values Height (in.) 71.4 68.5 t
