Custer headache (CH), which French authors refer to as suicide headache, also goes by the name of ‘alarm clock’ headache, owing to the day-to-day regularity of the attacks, which often reoccur even at regular times of the day. The “cluster headache” definition, coined by Kunkle in 1952, is the one which has prevailed over the years, even though there have been several names whereby this disorder has been described. As from the early 20th century, many different denominations where coined, in an attempt to define CH by referring both to the peripheral nerves which appeared to be involved in the pain fit (ciliary, pterygopalatine, petrosal and vidian nerves), and the local autonomic signs associated with the pain (rubra and angioparalytic migraines, with hemifacial vasodilatation having a sympathetic genesis , autonomic facial cephalalgia, etc.). However, these definitions only described partial aspects of cluster headache, as it is viewed today: it was thanks to Horton that, even prior to World War II, an exhaustive description of this disorder was supplied, which resulted in the definition “Horton’s histaminic cephalalgia” becoming very popular and remaining for decades the most widely used expression. Just as popular, was the vascular headache definition, which is still in use but should, on the contrary, be completely abandoned, especially as a result of the data acquired over the last few years, providing evidence as to the crucial role played in the physiopathology of CH by the central nervous system and by the hypothalamus in particular. With the term cluster headache Kunkle meant to metaphorically describe the typical tendency of these fits to cluster together within very specific time intervals (6/12 weeks, on average), which he defined “cluster periods”. Indeed, such fits cluster together with calendar-like predictability (with an annual or seasonal recurrence) as far as the active phases of the disease are concerned (“cluster periods”), and with a “clock-like” precision (circadian/ultradian recurrence) as far as the fits themselves are concerned (“cluster attacks”). This cyclical sequence is not only seasonal, with a typical recrudescence in early spring and autumn, but is also related to the changes in lifestyle routine, with a clear connection for instance to working activity, to periods in which resting periods are not regular, or to temperature or time-zone changes. The observations on the connection existing between the frequency of the fits and the seasons of the year (which is likely to be related to light time fluctuations) are in any case not the only ones: normally, the length of an active period ranges between 1 and 2 months, but, in order to qualify for the ‘cluster’definition, a minimum length of one week is needed, with a maximum duration of a year. The reoccurrence of the fits over the 24 hours is just as cyclical: the attacks do not usually last more than 3 hours and may reoccur several times throughout the day, almost according to a set timetable, especially during early afternoon or the first part of the night, displaying a clear relation with the waking/sleeping cycle and with meal times, rather than with the type and amount of food eaten. At least half of the patients suffering from CH has experienced being suddenly awakened during sleep by the occurrence of a painful fit. Night-time fits, which tend to be not as frequent as those occurring during daytime, appear to be related to the REM sleeping phases and only 10% of patients display exclusively nocturnal fits, whereas 30% regularly suffer from them during postprandial rest. According to the criteria adopted by the International Headache Society (IHS), occurrence frequency ranges from one fit every two days to 8 fits per day, and their lengths ranges between a minimum of 15 minutes and a maximum of 180 minutes. The periods lapsing between 1 and 3 o’clock p.m., around 9 p.m. and between 1 and 2 a.m. have been reported to be those associated with greater risk. Some authors have related them to certain lifestyle habits, coinciding with times of greater relaxation and with postprandial rest, also in connection with a greater consumption of alcoholic drinks.
EPIDEMIOLOGY
Cluster headaches affect approximately 0.1% of population, with a distinct
male prevalence, which is even more evident in the chronic form: the various
surveys display a 3/1 male/female ratio. Cluster patients often also display
typical physical features, such as rubicund colouring, deep wrinkles, “orange-peel”
skin, telangiectase, tightened palpebral fissures, broad skull and chin
and a significantly higher stature than in healthy population. An apparent
preference is reported for medium-high socio-economical strata, and since
also the few women affected by this disease are mostly engaged in professions
involving high responsibilities, some American authors have coined the term
manager headache, and have even outlined the typical features of the preferred
victim: a professionally very committed male, belonging to a medium-high
socioeconomic stratum, a businessman, an executive, a leader with a stressful
life, high stature and sturdy build, most often hazel-eyed and with a broad
face (some refer to ‘leonine looks’), often a heavy smoker and drinker,
not particularly hypochondriac and often married with a “short and domineering
woman”. However, female cluster headaches also seem to have been increasing
over the last few years: as women start taking up increasingly demanding
professional positions and lifestyles, with ever increasing responsibilities,
they are in all likelihood also inheriting the burdens of sex equality.
The prevalence studies on cluster headaches we are referring to are those
conducted in 1978 by the Swedish researcher Karl Ekbom and in 1986 by Giovanni
D’Alessandro, involving all the 21,792 individuals (both male and female)
making up the population of the Republic of San Marino at the time: among
these, only one case of female cluster headache was found, which corresponded
to a 0.009% rate, and, compared to the 0.13% rate reported for the male
population, it involved a 14:1 male/female cluster headache ratio. The only
similar study conducted more recently (1999) by Gian Camillo Manzoni has
however shown a female CH rate of 0.04%, which is almost fourfold compared
to the survey carried out 13 years earlier by D’Alessandro. Even though
women have always displayed a cluster headache of their “own” (the CPH variant,
which stands for Chronic Paroxysmal Hemicrania, originally described by
Sjaastad and Dale in 1974), displaying painful fits, which are much more
frequent and shorter, a 1999 study indicated that also the true female cluster
headache differs from the male variant: the onset takes place earlier (on
average around 25.3 years of age versus 29), the “cluster” and remission
periods are shorter, ptosis and miosis are less frequent than in men, whereas
nausea and vomiting appear more often.
FEATURES
The average age at which cluster headache makes its appearance is around
20-30 years, with a peak at the age of 20 in patients with episodic cluster
headache, whereas the onset of the chronic form displays a bimodal progress
with two incidence peaks: in the 10-29 age bracket and in the 50-59 age
bracket. A research study conducted on 3,102 males and 3,527 females from
the ad hoc Group for CH in the age of development by the Italian Society
for the Study of Cephalagias (SISC), has shown that also in Italy CH is
very rare in youth: only 0.03% of young people of either sexes appears to
be affected by it, which confirms that this pathology tends to affect adults.
Also as far as the clinical variants of CH are concerned, we now have consolidated
epidemiological data: the episodic form represents 90% of cases, whereas
the chronic form represents only 10%. Ever since the onset of the latter
form, the periodic clusters of fits are missing. Occasionally, fits may
switch from one side of the head to the other or last longer, for instance
by exceeding 2 hours. In these cases it is easy to understand the importance
that autonomic symptoms have for diagnosis differentiation. The episodic
form is characterised by active periods with a high frequency of fits, with
months or even years of remission in-between, whereas the chronic form is
characterised by fits with no remission spells. Approximately 80% of the
patients maintain the episodic form for the whole course of the disease,
approximately 13% switches to the chronic form and in the remaining 7% of
cases combined forms are observed. Without the aid of drugs, only a few
patients display remissions exceeding 3 years, whereas in most cases clusters
occur with a rate ranging between 1 fit every 2 years and 2 per year.
THE
CLINICAL PICTURE
CHs display absolutely distinctive features owing to the extreme severity
of the clinical picture: the burning pain with its overwhelming intensity
dramatically impairs the patients’ quality of life. In 56% of cases, pain
is described as shooting, agonizing or stabbing, in 38% of cases as pulsating
or hammering and in 30% of cases as pervasive and acute. The fit is associated
with a typical sequence of autonomic signs and symptoms that are always
homolateral to the pain: lacrimation, rhinorrhoea, conjunctival injection,
partial Horner, forehead perspiration and rhythm disorders. The intensity
of the pain is such that it can become intolerable to the point that even
cases of attempted suicide have been reported. The image of the subject
writhing in pain or continuing to nervously walk up and down in the dark,
with abundant lacrimation from one eye and with his/her face twitched into
a mask of pain, typically represents the way this type of headache manifests.
Pain distribution is always on one side: most frequently on the right-hand
side (49.1%), the left-hand side being affected in 35.4% of cases, whereas
in 15.5% of cases there is an alternation in the side involved. Pain is
most frequently localised in the peri- and retroocular area, but often the
temporal, frontal, face and maxillary regions are also affected. When fits
are particularly severe and /or frequent, a feeling of unease may continue
to affect the area of interest also in-between critical phases. In approximately
10% of patients the range of symptoms becomes chronic, as it manifests on
a daily basis: this dramatically impairs the patient’s quality of life.
THERAPY
Ever since the early ’90s, the drug of choice for the treatment of painful
crises has been sumatriptan, a selective agonist of the serotoninergic receptors
5HT 1b/1d, which, when administered subcutaneously, proves effective within
a few minutes in 95% of cases. In slight-and medium-intensity fits, oxygen
is also effective, if inhaled at a speed of 7-10 l/min. for about 15 minutes.
A valid alternative is represented by indomethacin, administered either
by intramuscular injection or by 50 or 100 mg suppositories; this drug is
especially effective in women’s clinical variant. (CPH). Nasal application
of lidocain (4-6%) is also possible to anesthetise the sphenopalatine ganglion,
even though this technique is less used and more complicated to carry out.
Both for the episodic and for the chronic form, the prophylactic drug of
first choice is verapamil, whose employment has been finally validated in
a dosage of 360 mg/day through a double-blind trial study versus placebo,
conducted two years ago by our group: crises decrease both in frequency
and intensity, and they finally disappear, as a rule, within 3-7 days. Before
starting the therapy, it is advisable to undergo an ECG, since this drug
displays contraindications in the event of major bradycardia or of atrioventricular
block. Lithium carbonate is also widely employed: whilst being contraindicated
in patients who are hypertensive, naturopathic or suffer from thyroid or
kidney disorders, it is more effective in chronic forms than in the episodic
variants. The usually employed dosage is of 900 mg/day: to avoid possible
side effects (diarrhoea, polyuria., nervous excitement, tremor or hypothyroidism),
lithiemia plasmatic levels should always be checked to make sure they remain
within the 0.4-0.8 mEq/L range. A double-blind trial study comparing verapamil
(360 mg/day) and lithium (900 mg/day) has proved that the two drugs are
equally effective, but verapamil displays fewer side effects and a shorter
latent period.
THE
OTHER TREATMENTS AND THE ASSOCIATION THERAPIES
In the episodic form, the corticosteroids prednisone and desametasone are
effective and display a rapid preventive action. Prednisone is employed
in a 50-60 mg/day dosage for 2-3 days, with a subsequent dosage reduction
in the amount of 10 mg/day every 2-3 days. In the chronic form, these drugs
produce a rapid reduction in the number of fits, especially in the initial
phases of the treatment, when preventive drugs have not started to produce
their action yet. In some cases, a cycle of intravenously administered desametasone
is required, with a gradually decreasing dosage, starting with 12 mg in
saline solution, subsequently reducing the dosage by 4 mg every 3 days.
This administration method is designed for drug-resistant cases with a high
crisis frequency, whilst adequate gastric protection needs to be provided
with specific drugs. In monotherapy non-responders, drug associations can
be indicated: verapamil + lithium, verapamil + steroids and, in chronic
forms, lithium + steroids. When identifiable, possible trigger factors,
such as alcohol and smoke, should be also eliminated; these however can
only influence the crises during the cluster periods, that is when the disease
is already active. The prophylactic treatment is certainly effective in
the chronic forms, whereas in the recurring or cyclic forms there is always
a chance that the crisis period may spontaneously end, rather than ending
as a result of the therapy. The prophylaxis should be started early, especially
in episodic forms, and should subsequently be continued for at least two
weeks, also after the crisis has disappeared. Withdrawal should be gradual
and, in the event the crises reappear, one should again gradually increase
amounts up to the therapeutic dosage. When the next cluster phase appears,
the treatment should always be resumed in the same manner.
HYPOTHALAMIC
INVOLVEMENT
The symptom remission periods appear bizarre and unpredictable, with healthy
intervals ranging from a minimum of two months to several years. This anomalous
feature, which constantly distinguishes cluster headaches, has always been
the object of research studies and continues to be an open issue allowing
for data and hypotheses which increasingly point at an involvement of the
brain structures that have to do with the regulation of the neuroendocrine
rhythms. The seasonal regularity with which the cluster periods reoccur
and, above all, the rhythmicity of the painful fits which occur at specific
times of the day and of the night has led to the hypothesis of an involvement
of the biological clock located in the hypothalamus, which is also confirmed
by the rhythmical alterations detected by our group in the prolactin and
cortical levels and by the results of the studies conducted on the rhythm
of melatonin. In 1998, the German neurologist Arne May detected , through
PET, that during the pain fit, homolaterally to the pain, an activation
of hypothalamic cells occurred and, the following year, through functional
magnetic resonance, he demonstrated that in that very area an alteration
of neuronal density also occurs. Therefore, on the whole, this research
approach has led to the identification of a possible brain target for the
treatment of cluster headaches. Last year, our group published on the New
England Journal of Medicine the results of a trial study which, for the
first time, demonstrated the effectiveness of a stereotaxic neurosurgical
technique, called deep brain stimulation, which controls cluster headache
fits which do not respond to pharmacological treatment, through the implant
of an electrode in the rear hypothalamus area. The treatment consists in
a stereotaxic inhibitory stimulation of the hypothalamic area during the
active phase of the disease. This operation should be limited to cases to
be selected through careful clinical investigation and a strict assessment
protocol. The first implant was carried out on 14 November 2000 in a patient
who for a long time had enjoyed no benefit from usual prophylactic therapies.
All the other patients subsequently subject to this technique also suffered
from untreatable chronic cluster headache and had responded to no prophylactic
treatment (verapamil, lithium, prednisone, intravenous desametasone, indomethacin,
metisergide, valproic acid, etc.).
DEEP
BRAIN STIMULATION
In the rear hypothalamus nucleus, the surgeon implants a microelectrode
which, in all likelihood, exerts an inhibitory function on the neuronal
hyperactivity of the cluster phase. The technique, besides fully controlling
pain, has proved to be absolutely safe; however it should only be implemented
on patients on whom total ineffectiveness or non-response to pharmacological
treatment has been proved with reasonable certainty. As a rule these patients,
besides suffering of course from a chronic form, need to meet the following
requirements: 1. Have a strictly unilateral cephalalgia; 2. Display complete
refractoriness to medical therapy or clinically significant contraindications
to pharmacological treatment; 3. Not suffer from personality disorders.
The deep brain stimulation method has already been used for some time in
the treatment of various other conditions, such as for instance in the field
of movement disorders, of Parkinson’s Disease or of peripheral pain, where
the stimulating electrode is positioned in the periaqueductal grey (PAG)area
, whereas with central pain , the ventroposterolateral (VPL) thalamus area
is used. The choice of the rear hypothalamus area for the treatment of chronic
cluster headache was based on the findings of neuroimaging and clinic neuroendochrinology
studies, which have led us to the hypothesis that the rear hypothalamus
area was a more accurate target for deep brain stimulation treatment. Over
the last ten years, this technique has proved safe, with very rare side
effects and a negligible mortality rate, with the further advantage of a
possible complete reversibility of the operation. At over two years’ distance,
we have observed a persistent effectiveness without side effects, and to
date 6 additional patients suffering from untreatable chronic cluster headache
have been subject to this operation. All these patients have displayed a
gradual reduction of the crises up to their full disappearance, without
suffering from side effects and with the recovery of a quality of life which
had previously been very bad. Translated by Interpres sas
Gennaro
Bussone
Primario Neurologo dell’Unità Operativa “Cefalee e malattie cerebro vascolari”
Istituto Nazionale Neurologico Carlo Besta
Milano
