ICHD-3 Beta Definition of Cluster Headache
A. At least 5 attacks fulfilling B-D below.
B. Severe unilateral orbital, supraorbital, and/or temporal pain lasting 15 to 180 min.
C. Attack is associated with at least one of the following signs on the side of pain:
1. Conjunctival injection
3. Nasal congestion
5. Forehead and facial sweating
8. Eyelid edema
D. Frequency: from one every other day to eight per day
E. At least one of the following:
1. History, physical, and neurological examinations do not suggest disorders in groups 5-11 of IHS classification.
2. History and/or physical and/or neurological examinations do suggest other disorder, but it is ruled out by appropriate investigations.
3. Such disorder is present, but tension-type headache does not occur for the first time in close temporal relation to the disorder.
During the cluster headache in addition to pain the patient may experience symptoms on one side of the face around the eye, upper cheek, or temple. These symptoms may be:
drooping of the upper eyelid (ptosis)
smallness of one pupil (miosis)
sweating above the eye
redness of the eye (conjunctival injection)
tearing of one eye (lacrimation)
nasal congestion or drainage of clear fluid (rhinorrhea)
These “sinus” type symptoms many times bring the patient to the general doctor or ear nose throat surgeon with a self-made diagnosis of “sinus headache” with resultant treatment with repeat doses of various antibiotics or occasional sinus surgery.
The pain begins quickly, without warning. It is excruciating in intensity and explosive in quality. Rarely the pain is pulsatile. Patients describe the headaches as “killer” headache[i] or “suicide headache.” The pain may be “boring, stabbing, burning, like a knife, like a hot poker in the eye” and sometimes “pulsating, throbbing, squeezing, or aching.” At the end of the attack the symptoms resolve in 1-2 minutes.
The pain of cluster headache is located in one temple and behind or above one eye. Sometimes the face, neck, ear, or one side of the head may be involved. Cluster headache pain is persistently one sided in location. Migraine without aura headaches are unilateral 60-70 % of the time but cluster headaches are always one sided. Migraine may switch sides back and forth, the patient usually stating that one side is more prominent, but cluster headaches are strictly one sided and switches sides in 15% of patients usually for the duration of a bout. Very rarely the pain will switch sides during the cluster period.
The patient may have characteristic autonomic symptoms during the attack. Lacrimation from the eye affected with pain is the most common associated symptom. The autonomic nervous system is a wired nervous system connecting the brain through the spinal cord and ganglia to a target organ. For instance, with tearing the autonomic nervous system connects to the lacrimal gland to produce tears in the eye. The autonomic nervous system fibers enter the eye wrapped around the carotid artery, and go to specific anatomical areas. These autonomic fibers innervate the pupil causing dilatation or constriction, the nasal turbinates to produce mucus, and a small muscle in the upper eyelid which helps hold the eyelid up. Autonomic fibers also innervate the sweat glands over the forehead and the blood vessels which course over the surface of the conjunctiva--the white part of the eye.
All of the so called autonomic symptoms of cluster headache reflect altered, usually temporarily, dysfunction of the different fibers. Not every cluster headache patient has all the symptoms on the list, but commonly they may have three or four of the cardinal symptoms. Usually family members will comment about the upper eyelid drooping, a medical condition called ptosis. A small pupil on one side may be noted by a family member or the patient, but only if they are very observant. Redness of the conjunctiva, which is the white part of the eye, may be noted by the patient if they look at themselves closely in the mirror. Tearing, called lacrimation, or nasal dripping and congestion are usually pretty obvious. In my experience forehead sweating is not a prominent symptom and may not be noted by the patient.
Sometimes a patient with chronic cluster headache may develop an upper lid/pupillary syndrome related to damage to the tiny autonomic fibers in the carotid artery on that side called Horner’s Syndrome. Patients with Horner’s syndrome present with mild upper lid drooping so that the lid may fall down below the colored part of the eye. This is usually not bad enough to obscure vision and may just be a few millimeters of change. Also Horner’s syndrome is associated with smallness of the pupil on one side, a condition called miosis. There may also be lack of sweating, called anhydrosis, over the upper forehead and so many a medical student memorizes the classical physical findings of Horner’s Syndrome--ptosis, miosis, and anhydrosis.
Nausea, vomiting, and sensitivity to light and sound may occur with cluster headache, but these symptoms are not as prominent as they are in migraine without aura.
Cluster headaches come without warning and reach a peak within 2-15 minutes. This is different from a typical migraine without aura attack which may take half an hour to several hours between onset and peak of headache pain. Cluster headaches are very severe, quick onset, one sided headaches which consist of pain around the eye, temple, or cheek. Cluster headaches may track the clock, coming at the same time every day. This periodicity is a key feature of cluster headache; the attacks of pain recur at the same hour each day for the duration of the cluster bout.
The attacks characteristically occur one to two hours after going to sleep in half of patients. An attack at this time corresponds with dreaming and REM stage sleep. Some patients have several attacks at night consistently interrupting sleep. Some patients get daytime attacks associated with napping or relaxation. Seventy-five percent of attacks occur between 9 p.m. and 10 a.m.
The duration of the attacks ranges from 15 to 180 minutes, with most cases lasting about 30 minutes to 2 hours with a mean of 45 minutes. The attacks may occur 1 to 8 times a day and the patient is pain free between attacks.
Attacks--the cluster period
The headaches come in time periods called “clusters” which usually last 6-12 weeks. The term “cluster” here means that the headaches cluster together in time, much like grapes cluster together on the vine.
Thus, the patient may state that he had 6 weeks of headaches in March and April of 2004, but then the headaches completely stopped. Following this there were no headaches in all of 2005, but they started again in March of 2006 and brought the patient to the doctor.
Although the usual cluster period is 1 ½ to 3 months, 10% of patients develop chronic cluster type headaches all year round and in these patients the “cluster” term is meaningless.
Also cluster patients may cycle between intermittent or the typical "cluster" pattern and chronic daily cluster headaches. Chronic cluster patients may evolve into the episodic form without treatment.
Although some cluster periods occur in the spring and fall, other researchers have found cluster cycles in February and June that seem to occur at the time of increase in daylight hours. Cluster attacks occur 7 to 10 days before the longest and shortest days in the year, suggesting that the pineal gland located in the center of the brain may be involved. The pineal gland responds to ambient light and helps set the sleep cycle.
Behavior during attacks
The patient usually gets up and paces around the room. Sometimes they may sit but they don't lie down in bed in a quiet room with the lights out like a typical migraine patient would do and they don't miss work like a migraine patient would.
Blau in 1993[ii] wrote an article in The Lancet on the “Behavior during a cluster Headache.” Blau stated:
Walking with the trunk slightly bent forwards and clutching the head, or sitting and rocking backwards and forwards with the hands pressed on or near the painful site, was the most common behavior during attacks.
Self-trauma, a common feature, included pressing a finger, thumb, or fist into the affected eye or adjacent temple, hitting the forehead against a hard object (wall, floor, radiator, or mantelpiece), or rubbing or pressing the forehead on the affected side on the carpet or chair; others repeatedly hit or painfully scratched the head distant from the painful site; one rubbed his thighs until they became red and sore. Clenching fists so that the finger nails dug into the palms, applying intense heat (“a boiling cup of tea or a hot water bottle”), or ice-cold objects to the forehead were also described. In two patients with lower-half cluster headache (pain radiating to the upper jaw instead of upwards over the head), one inserted and twisted a steel knife blade between the painful upper teeth, while another pushed a finger nail as hard as he could into a tooth socket in the painful region. Three pushed a cotton wool tip up the affected nostril while another blew his nose very hard, each to provoke the rhinorrhoea that heralded the end of their attacks.
Dr. Karl Ekbom[iii]
Ekbom[iv] noted the inability of cluster headache patients to remain still during attacks, in contrast with migraine patients who want to lie completely still.
Another interesting feature is that cluster headache may be aggravated by alcohol consumption only during the cluster period in about half the patients. If a migraine patient reacts to a certain alcoholic drink, the migraine may come on anytime the patient imbibes that type of alcohol. The cluster headache patient may say that a beer or glass of red wine will immediately set off a headache during the 6-12 weeks when the headaches come, but alcohol has no effect on inducing headache between cluster attacks.
The vasodilator, nitroglycerin, known to treat vasospastic angina pectoris characteristic of coronary artery disease, given as 1 mg sublingually triggers an attack during a bout. Histamine may also provoke an attack.
Distinguishing patient features
The patients often have a driven, type A workaholic nature. They tend to be ambitious, efficient, conscientious, striving, compulsive, self-controlled, self-sufficient, reserved, and tense. They also tend to smoke cigarettes and drink alcohol more than the usual person.
I recall a patient I saw years ago who had cluster headache and the typical driven lifestyle some of these patients have. In his small-town he was the sheriff, voluntary fireman, and emergency ambulance attendant. He did these jobs on the side while running his family business during the day. He carried four beepers.
At the June 1985 Second International Headache Congress in Copenhagen Dr. John R. Graham[v] of the Headache Research Foundation, Jamaica Plain, Massachusetts stated that “Most patients with cluster headaches are men who are considered ‘Macho.’” Dr. Graham thought that “the few women who got cluster headaches tended to have square, boyish faces, and to be masculine in appearance.”
At the same meeting Dr. Lee Kudrow,[vi] of the California Medical Clinic for Headache, Encino reported that men with cluster headaches were as much as 3 inches taller than the average and had characteristic facial features with deep asymmetrical facial creases.
There is also a relationship between cluster headache and obstructive sleep apnea.[vii]
The title itself, cluster headache, causes confusion in the lay and in some doctors’ minds. Several times a month I’ll talk with a headache patient who will tell me that their doctor said they had “cluster headache,” because the patient had several headaches in a row. Usually the patient has migraine without aura that is partially treated with acute therapy medication and the patient has one big long headache with peaks and valleys. Many of these patients have a diagnosis of medication overuse headache and have central sensitization.
There was a lot of confusion in the literature when the syndrome of what is now universally called cluster headache was first recognized. At the start of my career in the late sixties I was taught to call this type of headache “vasodilating” headache which I did for a few years until the “cluster” term rose to prominence and then stuck.
Other names offered for cluster headache have been migrainous neuralgia, histamine headache, histamine cephalalgia, Horton’s headache, Ciliary neuralgia, Sluder’s sphenopalatine neuralgia, geniculate neuralgia, Raeder’s paratrigeminal neuralgia, erythroprosopalgia, Raeder's syndrome, vidian neuralgia, red migraine (because the white part of the eye turns red), periodic migrainous neuralgia, ciliary (migrainous) neuralgia, greater superficial petrosal neuralgia, and anterior migraine.
Cluster headache is now firmly established as a distinctive syndrome. Finally it was understood that all of these conditions had similar symptoms and that different doctors in different parts of the world had been describing the same thing.
Cluster headache occurs in approximately 69 cases per 100,000 people, and is far less common than migraine. Men are affected more than women at a rate of 6:1. It usually occurs between the ages of 20 to 50 years with a mean of 30 years; however, it may begin as early as the first decade and as late as the eighth decade. Attacks in women with cluster headache usually do not correlate with their menses, generally stop during pregnancy, but like migraine may be started by the use of oral contraceptives. Women may have both cluster headache and migraine but they are commonly misdiagnosed only with migraine.
Cluster in women
Rozen, et al,[viii] writing in the Journal of Neurology, Neurosurgery, and Psychiatry in 2001 studied the clinical features of cluster headache in women and found that their experience was similar to men. Women develop cluster headache earlier than men with the mean age of onset at 29 years in women as opposed to 31 years in men. Women had 2 peaks of onset in the 2nd and 5th decade, whereas men peaked at 31 years. Women had more “migrainous symptoms” with cluster headache, especially vomiting, but both men and women had frequent sensitivity to light and sound. Rozen, et al,[ix] pointed out that the symptoms of photophobia and phonophobia are not included in ICDH II cluster headache criteria but he thought that the criteria needed revision to include them.
Cluster headaches used to occur mainly in men and a ratio of men/women of 6/1 was commonly quoted in the past, but recent articles suggest that women are getting the syndrome more often, perhaps related to the stress of women leaving the home and entering the workplace. Bahra, et al,[x] wrote in Neurology in 2002 on “Cluster headache. A prospective clinical study with diagnostic implications.” They stated that “The overall male-to-female ratio in this sample was 2.5:1, and this has decreased with time.” Also Manzoni[xi] wrote an article in 1997 in Headache entitled “Male Preponderance of Cluster Headache is Progressively Decreasing Over the Years.”
Age at onset
The typical age of onset is the third decade which would be 20 to 30 years old but the range extends from 20 to about 60 years old. Many persons with cluster headache have family members with migraine.
The history of the various terms used to describe this painful disorder has a long serpentine story weaving back into the eighteenth century. Most authors give early credit to the description made by von Mollendorff,[xii] who in 1867 described “red migraine with hemicrania, homolateral redness of face, injection of the eye, lacrimation, and dilatation of the temporal arteries." However, an earlier report was made by Gerhard van Swieten[xiii] who gave a full description of episodic cluster headache which fulfills IHS criteria in 1745 in his textbook of clinical medicine. Van Swieten was the founder of the then leading medical center, the Vienna school. His article, lost because it was originally written in Latin, was found again and translated in 1992.
Wilfred Harris (1869-1960) [xiv]
In 1926 Wilfred Harris[xv] described a special recurrent headache which he called ”Periodic Migrainous Neuralgia.” In 1936[xvi] in the British Medical Journal Harris wrote about the subject again using the title of “Ciliary (Migrainous) Neuralgia.” Harris was the first to describe successful treatment of the acute attack of cluster headache with ergotamine.[xvii] In his 1936 article Harris offered his “Definition of Migrainous Neuralgia”:
I have employed the term “migrainous neuralgia” to describe cases of recurrent neuralgia affecting the temple or the side of the forehead and often both jaws, sometimes extending to the back of the head, usually strictly unilateral. The pain in these cases is often most intense and excruciating, and may start suddenly or gradually and subside in the same way, the duration of the pain varying from ten minutes to half an hour, or often for five or six hours and occasionally for twenty-four hours for periods of six to eight weeks yearly. Nausea occasionally accompanies the pain, suggesting its association with migraine, but vomiting is rare, and visual spectra and transient hemianopia are never met with as in the usual form of migraine.
Case II from Harris’ 1936 article on Ciliary (Migrainous) Neuralgia follows:
In one case, indeed, a man of 40 suffered almost continually for three weeks, having had only eight hours; sleep during that period. His first attack began when he was aged 37, and the longest interval between attacks was two months, the severe pain lasting usually one or two weeks, being continuous and limited strictly to the left eyeball.
The pain is described by various sufferers as “intensely severe”; “very excruciating”; “the eye feels as though being hammered or rolled on the ground”; “the eye feels as if it were being pulled out, or turned inside out”; “like boiling water behind the eye.”
Bayard T. Horton (1895-1980) [xviii]
Horton, et al,[xix] spoke about a new syndrome which they originally named “erythromelalgia of the head” in 1939 but changed to “histamine cephalalgia” in 1941.[xx] Horton[xxi] at the Mayo Clinic called the syndrome “histamine headache” in 1939. One of Horton’s cases from his 1939 article follows:
Vasodilating pain in the left eye, associated with a left hemicrania, and precipitated by the drinking of beer--A man, aged 58 years, gave a history of recurring bouts of pain in the left eye of twelve months’ duration. The attacks occurred either during the day or night and often awakened him from a sound sleep.
The onset of the pain was gradual and was localized behind the left eye. It was of a constant, sharp character, but no throbbing sensation was present. When the pain lasted more than one and a half hours, it extended to the left temporal and occipital regions and into the neck. It decreased in the orbital region as it increased in the occipital region. This pain often persisted for four or five hours, without associated nausea or vomiting. The drinking of alcoholic beverages, particularly beer, invariably precipitated an attack of pain in the left eye after an interval of thirty to fifty minutes. The pain was reproduced on three different occasions by having the patient drink a bottle of beer. Compression of the left common carotid artery gave momentary relief.
Gardner, et al,[xxii] introduced “greater superficial petrosal neuralgia” in 1947 and suggested treatment by resecting the nerve. “Cluster headache” was used by Kunkle, et al,[xxiii] in 1952 and then by Friedman and Mikropoulos[xxiv] in 1958.
In 1956, Sir Charles Symonds[xxv] writing in Brain gave a more complete account of the syndrome of cluster headache in an article entitled “A Particular Variety of Headache” which helped to educate the general medical audience and the public about this little known malady. Symonds stated about the syndrome:
Its essential features are the occurrence of paroxysms of headache, that is to say, pain of sudden onset and transitory duration, which occur in bouts lasting as a rule for several weeks with long intervals of perfect freedom. In the paroxysm the pain is felt mainly in the supra-orbital region or in and behind the eye, though it may spread beyond this region. It is, however, strictly unilateral. It is of agonizing severity, but very rarely lasts longer than two hours and often less. During a bout there is usually at least one paroxysm in each 24 hours; there may be more. In the intervals between paroxysms there is complete relief. The bout having ended, there is no further complaint of headache until the occurrence of the next bout after an interval of freedom, which is rarely less than 6 months and may be several years. No local cause for the pain is to be discovered in the shape of disease of the eyes, nose, nasal sinuses, ears, scalp or skull, or the sensory nerves supplying the region where the pain is felt.
A report of Symonds’ case 1 follows:
Male, aged 39 when first seen in 1948. At the age of 35 began to have attacks of pain behind the left eye described as like pain in a tooth exposed to cold, reaching peak intensity in 15 minutes, remaining thus for an hour or an hour and a half, and rapidly disappearing, but sometimes followed by a dull ache for an hour or two. At the height of the attacks the eye would water and become bloodshot and the left side of the nose would feel blocked. The attack would usually begin between 2 a.m. and 3 a.m. and would be repeated for two or three nights in succession. After an interval of six or eight weeks the experience would be repeated.
When seen again in November 1955, he reported that the nature of the attacks had remained unaltered, but that the series or bout had gradually become longer, lasting up to a fortnight, and the intervals between bouts had also lengthened to six months. Then in the summer 1952 he had a bout lasting four weeks followed by complete freedom for eighteen months. There was then a bout lasting two weeks, again followed by freedom for eighteen months, when he began a bout which had lasted for five weeks when he was seen. In this the paroxysms had occurred at irregular times, mainly during the night, but occasionally during the day. On two occasions there had been three paroxysms in the twenty-four hours. But on several occasions he had gone twenty-four hours without a paroxysm.
Cluster headache is now considered to be one of the trigeminal autonomic cephalalgias, a group of primary headaches which includes paroxysmal hemicrania and Short-lasting Unilateral Neuralgiform headache attacks with Conjunctival injection and Tearing (SUNCT). The area of the brain involved with cluster headache is the posterior hypothalamic gray matter. In contradistinction the area of the brain involved with migraine is located in the brainstem.
Goadsby and Lipton[xxvi] stated in a review of this subject in Brain in 1999:
The short-lasting primary headache syndromes may be conveniently divided into those exhibiting marked autonomic activation and those without autonomic activation. The former group comprises chronic and episodic paroxysmal hemicrania, short-lasting unilateral neuralgiform headache with conjunctival injection and tearing (SUNCT syndrome) and cluster headache. These headache syndromes are compared with other short-lasting headache disorders, such as hypnic headache, and persistent headache with milder autonomic features such as hemicrania continua. Cluster headache is included with the shorter-lasting headaches to attempt a nosological analysis of these syndromes. The paroxysmal hemicranias are characterized by frequent short-lasting attacks of unilateral pain usually in the orbital, supraorbital or temporal region that typically last minutes. The attack frequency usually ranges from 5 to 40 attacks per day. The pain is severe and associated with autonomic symptoms such as conjunctival injection, lacrimation, nasal congestion, rhinorrhoea, ptosis, or eyelid oedema. Almost all reported cases respond to treatment with Indomethacin, but respond poorly to other treatments including other non-steroidal anti- inflammatory drugs. A recent case study demonstrated the release of both trigeminal and parasympathetic neuropeptides during a bout of pain in the same pattern previously described in cluster headache.
Research like the above led to the ICHD 3 Beta[xxvii] classification:
Cluster headache and other trigeminal autonomic cephalalgias
3.1 Cluster headache
3.1.1 Episodic cluster headache
3.1.2 Chronic cluster headache
3.2 Paroxysmal hemicrania
3.2.1 Episodic paroxysmal hemicrania
3.2.2 Chronic paroxysmal hemicrania (CPH)
3.3 Short-lasting unilateral neuralgiform headache attacks with conjunctival injection and tearing (SUNCT)
3.4 Probable trigeminal autonomic cephalalgia
3.4.1 Probable cluster headache
3.4.2 Probable paroxysmal hemicrania
3.4.3 Probable SUNCT
Goadsby[xxviii] writing in Lancet Neurology in 2002 on “Pathophysiology of cluster headache: a trigeminal autonomic cephalgia,” stated:
Cluster headache is a form of primary neurovascular headache with the following features: severe unilateral, commonly retro-orbital, pain accompanied by restlessness or agitation, and cranial (parasympathetic) autonomic symptoms, such as lacrimation or conjunctival injection. It occurs in attacks typically of less than 3 h in length and in bouts (clusters) of a few months during which the patient has one or two attacks per day. The individual attack involves activation of the trigeminal-autonomic reflex; thus, such headaches can be broadly classified with the other trigeminal-autonomic cephalgias, such as paroxysmal hemicrania and the syndrome of short-lasting unilateral neuralgiform headache attacks with conjunctival injection and tearing. Observations of circadian biological changes and neuroendocrine disturbances have suggested a pivotal role for the hypothalamus in cluster headache. Functional neuroimaging with PET and anatomical imaging with voxel-based morphometry have identified the posterior hypothalamic grey matter as the key area for the basic defect in cluster headache.
There also are patients who have both cluster headache and trigeminal neuralgia (tic douloureux). This may be called the “cluster-tic syndrome.” The patient should receive both diagnoses and be treated for both conditions.[xxix]
Horton[xxx], at the Mayo Clinic and well experienced with cluster headache said: "Error in diagnosis is usually due to the fact that the physician has not had the opportunity to observe the patient in the course of a spontaneous or induced attack."
What Horton described is usually no longer possible since the doctor and the patient now interface at the clinic in the doctor’s office usually between attacks and the patient describes a history of something that had happened in the past, rather than the doctor seeing the patient at home in his native environment and observing an attack.
[i] Waldman SD. Cluster: The “Killer” Headache. National Headache Foundation Head Lines, July/August 2007;157:2-3.
[ii] Blau JN. Behaviour during a cluster headache. The Lancet. 1993;342:723-25.
[iv] Ekbom KA. Clinical comparison of cluster headache and migraine. Acta Neurol Scand.. 1970;43(suppl 41):1-48.
[v] Graham K. International Medical News Service Second International Headache Congress. June 1985.
[vi] Kudrow L. International Medical News Service Second International Headache Congress. June 1985.
[vii] Weintraub JR. Cluster headaches and sleep disorders. Curr Pain Headache Rep. 2003 Apr;7(2):150-6.
[viii] Rozen TD, Niknam RM, Shechter AL, Young WB, Silberstein SD. Cluster headache in women: clinical characteristics and comparison with cluster headache in men. J Neurol Neurosurg Psychiatry 2001;70:613-617.
[ix] Rozen TD, Niknam RM, Shechter AL, Young WB, Silberstein SD. Cluster headache in women: clinical characteristics and comparison with cluster headache in men. J Neurol Neurosurg Psychiatry 2001;70:613-617.
[x] Bahra A, May A, Goadsby PJ. Cluster headache. A prospective clinical study with diagnostic implications. Neurology 2002;58:354-361.
[xi] Manzoni GC. Male Preponderance of Cluster Headache is Progressively Decreasing Over the Years Headache: The Journal of Head and Face Pain. 1997;37(9),588–589.
[xii] Friedman AP, Mikropoulos HE. Cluster Headaches. Neurology. 1958; 8:653-663.
[xiii] Isler H. Episodic cluster headache from a textbook of 1745: van Swieten’s classic description. Cephalalgia. 1993;13:172-174.
[xv] Harris, W. Neuritis and Neuralgia. Oxford Medical Publications. 1926, p. 307.
[xvi] Harris, W. Ciliary (Migrainous) Neuralgia and its Treatment. British Medical Journal. March 7th, 1936, 457-60.
[xvii] Boes CJ, Capobianco DJ, Matharu MS, Goadsby PJ. Wilfred Harris' early description of cluster headache. Cephalalgia. 2002;22(4):320–326.
[xix] Horton BT, McLean AR, Craig WM. The use of histamine in the treatment of specific types of headache. Proc Staff Meet Mayo Clin. 1939;14:257
[xx] Horton BT. The use of histamine in the treatment of specific types of headache. JAMA. 1941;116:377
[xxi] Horton B. Symposium: Head and Face Pain. Medicine. Transactions American Academy of Ophthalmology and Otolaryngology. 1944. September-October: 23-33.
[xxii] Gardner WJ, Stowell A, Dutlinger R. Resection of the greater superficial petrosal nerve in the treatment of unilateral headache. J Neurosurg. 1947;4:105–114.
[xxiii] Kunkle EC, Pfeiffer JB, Wilhoit WM, Hamrick LW. Recurrent brief headache in "cluster” pattern. Trans. Am Neurol Assoc 1952;77:240-43.
[xxiv] Friedman AP, Mikropoulos HE. Cluster Headache. Neurology. 1958;8:653-663.
[xxv] Symonds C. A Particular Variety Of Headache. Brain. 1956:79:217-232.
[xxvi] Goadsby PJ and Lipton RB. A review of paroxysmal hemicranias, SUNCT syndrome and other short-lasting headaches with autonomic feature, including new cases. Brain. 1997;120(1):193-209.
[xxvii] The International Classification of Headache, 2nd Edition. Cluster Headache and other trigeminal autonomic cephalalgias. Cephalalgia. 2004. 24; (Supplement 1):44-47
[xxviii] Goadsby PJ. Pathophysiology of cluster headache: a trigeminal autonomic cephalgia. Lancet Neurol. 2002;1(4):251-257
[xxix] The International Classification of Headache, 2nd Edition. Headache induced by food components and additives. Cephalalgia 2004. 24; (Supplement 1):45
[xxx] Horton BR. Histaminic cephalalgia. JAMA. 1956;160:468-469.