Abstract
The mean age of onset of cluster headache (CH) is in the late third decade. Only few cases of childhood-onset (< 14 years) CH have been reported in the literature. We report the case of an 11-year-old boy who suffered from sudden attacks of shock-like, intense pain, localized in the right orbital region, with associated photophobia, phonophobia, conjunctival injection, lacrimation, nasal congestion, rhinorrhoea and psychomotor agitation. The episodes lasted 60–180 min, and the headache frequency was one to three per day. Physical and neurological examinations, magnetic resonance imaging and blood examinations were normal. The first bout lasted 8 months. Attacks were resistant to every symptomatic and partially to prophylactic treatment that has been tried. The second bout lasted approximately 2 months.
Introduction
Cluster headache (CH) is one of the most severe types of headache, characterized by periods of recurrent attacks of sudden and intense pain, localized in the orbital or temporal region, and associated with ipsilateral autonomic symptoms and signs (1).
Epidemiologically, CH is a relatively rare disease, affecting < 1% (i.e. 0.39%) (2) of the population, with a clear male preponderance (84%) (3), and the mean age of onset is 28–30 years (range 7–83 years) (4).
There are relatively few reports on the prevalence and clinical features in CH in children and adolescents, since only few population studies have also considered the paediatric population (5–8). To our knowledge, at paediatric age (0–18 years), CH has been considered to have the same clinical features as in adulthood. CH in this age range is rare, the estimated prevalence being 0.09–0.1% of the population. According to different studies (4–6,8,9), the sex ratio is approximately the same (M : F ∼3.2 : 1), but with a wide variation of range (1 : 1–6 : 1). Onset may be as soon as 3 years, but there is a relatively low number of cases with onset < 10 years old. A suspected case in a 1-year-old infant has also been described (10).
Case report
An 11-year-old boy was referred to our department due to recurrent severe, short-lasting and unilateral side-locked headache.
There were no close relatives with CH. Delivery, psychomotor and language development were normal. Since the age of 2.5 years he had suffered from episodes of deviation of the right eye and esotropia, for which he was initially treated with occlusion therapy. At 9 years, medial rectus and inferior oblique muscles in the right eye (the present symptomatic side) were operated upon.
Since 8 years of age, he had presented mild band-like headache episodes (∼2/year), without associated symptoms, lasting about 0.5 h; headaches were ketoprofen (80 mg)-responsive according to his parents.
At the age of 10.5 years, he started complaining of 30 min long attacks of sudden pain, described as sharp or shock-like, localized in the right orbital region. The intensity was severe, and pain was associated with photophobia, phonophobia, conjunctival injection, lacrimation, nasal congestion, rhinorrhoea and psychomotor agitation. The attacks had an initial daily fluctuating frequency (1–3/day), recurring at fixed hours of the day, and lasting in the full blown state for 60–180 min. The attack frequency has been reconstructed retrospectively collecting medical history on the basis of written notes by the parents during the months preceding our first observation.
In the clinical history reconstruction, occurrence of episodes was different from that subsequently observed, probably due to incorrect reporting by parents or to modifications in the timing of headaches, since sometimes three episodes per day were reported. Anyway, a sort of ‘grouping’ (or clustering) in specific hours of the day can be estimated.
Later on, using a headache diary, the daily frequency was two time-locked attacks, of 50–55 min duration. Attacks remained time-locked even during the beginning of daylight saving time in our country (Fig. 1). During the symptomatic period the patient was given different symptomatic and prophylactic pharmacological treatments and was under direct observation (as in-patient and out-patient) for 2 months, during the first cluster period. When admitted, several episodes were directly seen by different doctors, confirming the clinical diagnosis. Moreover, video documentation was obtained, but the video does not clearly show unilateral signs, because the boy was weeping/crying from pain.
Temporal pattern of headache attacks September 2006 to May 2007. x, single episode retrospectively reported by parents; grey box, headache attack occurrence, from headache diary.
The first bout lasted 8 months, with a progressive reduction in duration/frequency (Fig. 2)
Temporal pattern of duration of headache attacks.
Neurological examination was normal (weight 40 kg, height 156 cm). Blood examinations, including coagulation and inflammatory variables, showed no significant alterations. Brain computed tomography, magnetic resonance imaging (MRI) and angio-MRI were normal, except for mild asymmetry of the ventricular system (right>left). Electroencephalography showed slow posterior potentials during hyperpnoea (within normal range). Cardiological examination, ECG, echocardiography, transcranial Doppler, and echo-Doppler of the supra-aortic vessels were all normal. Ophthalmological examination showed normal fundus, absence of binocular vision, with alternate eye suppression, and hypermetropic astigmatism.
Three months after the onset of headache, while tapering off the corticosteroid therapy started prior to our observation, he presented a cutaneous macular eruption resembling mastocytosis. A cutaneous biopsy instead showed the presence of trichostasis spinulosa.
The second bout was at first characterized by episodes of ‘sense of heat’ in the same area of the head as the previous bout. Pain was absent, but mild and inconsistent autonomic unilateral signs (i.e. lacrimation) were noticed. The duration was between 2 and 6 h. These episodes were not daily for a while, progressively showing a daily frequency. After 1 month he presented headaches, beginning in the night at unspecified times, but always present at wakening, and always stopping at 10.30 h, despite therapy or different wakening times. The exact time of onset was not clearly specified, since the pain inconstantly awaked the patient from sleep, and sometimes he could fall, but the pain was always described as present at awakening.
The pain had the same characteristics as in the first bout. Associated symptoms/signs included photophobia, phonophobia and restless behaviour. No evidence of lacrimation, nasal congestion or rhinorrhoea was reported during the attacks in the second bout.
Each attack was followed by the onset of dizziness, lasting about 3 h.
At this stage the patient fulfilled the diagnostic criteria for episodic CH, having completed two bouts (8 and 2 months, respectively) with a 4-month remission in between (1).
The symptomatic treatments he was given [ketoprofen, paracetamol with codeine, metamizole, ketorolac, ibuprofen, oxygen (7–8 l/min, 10–15 min duration, administered through a non-rebreathing mask while sitting on bed), sumatriptan (6 mg subcutaneously) and octreotide (0.1 mg/ml subcutaneously] had no clear-cut effect, with the exception of oxygen, which showed partial efficacy on pain duration.
Summary of prophylactic treatment used in our patient
AE, adverse events.
During the observation time, the patient continued to present two headache attacks per day, lasting from 50 to 60 min.
During the period March–May 2007 the patient was seen and followed up in another headache centre, where he was put on combination therapy with rivastigmine, methysergide, olanzapine, Al-Mg hydroxide and L-acetylcarnitine. This treatment continued for almost 2 months, with the exception of olanzapine (suspended after 1 week, for psychomotor agitation) and methysergide (suspended after 1 month, for psychomotor agitation). In April, the frequency was one attack/day and in May the period was over, with progressively decreasing attack duration (Fig. 2). Despite this temporal pattern, the patient was seen in the same headache centre and was again given rivastigmine, parenteral methylprednisolone (4 mg/day), chlorpromazine (6 mg/day) and ranitidine (150 mg/day). On discharge, a therapy with rivastigmine, trimipramine, clonazepam and L-acetylcarnitine was tried for a period.
During all this period the patient has been followed in another headache centre. The therapies administered were all decided there, in agreement with parents.
Demographic data of published reports
Since elements of psychiatric comorbidity were found in the medical history, a psychodiagnostic evaluation was performed. Psychological data were obtained from individual interviews, drawings, semistructured diagnostic interview (Schedule for Affective Disorders and Schizophrenia for School-Age Children—Present and Lifetime Version) and administration of projective tests (Rorschach, Blacky Picture Test). From both the interviews and the tests clear components of anxiety appeared, especially regarding the separation from the parental figures, traits of obsessive-compulsive behaviour and an evident inhibition of phantasmatic activity, with a marked recourse to concreteness.
Moreover, during the second cluster period, there was more evidence of a secondary gain from the symptoms, such as a greater care from parents and other family members and a reduction in school activities.
In our opinion, the psychological picture does not affect the clinical pattern of our patient.
Discussion
To the best of our knowledge, there are 15 studies (n = 128), dealing with patients under 18 years [i.e. five population studies (110 patients) and 12 case reports (18 patients), considering that two studies are both population studies and case reports]. According to these studies, the prevalence of CH at paediatric age is estimated to be 0.09–0.1%.
The mean age of onset in paediatric age studies is 11–14 years, but the range could differ (i.e. 3–18 years), since several cases with onset under 10 years have been reported (5,6,8,10–20).
Genetic factors appear to be involved in the same proportion as in the adult population, as a pedigree positive for CH has been found in approximately 10% of cases, whereas a pedigree positive for migraine-type headache was found in 25% of cases. Male : female ratio seems to show almost the same proportion (M : F ∼3.2:1) as in adulthood.
Prevalence of episodic and chronic forms shows the same values as at adult age (80–90% vs. 10–20%). No data have been reported about the incidence of primary chronic forms, although six probable cases have been described (8,11,13,17).
These patients could present a single cluster period, or recurrence of clusters, separated by periods from 5–7 weeks to 6 months. The frequency of headaches attacks could vary from 3/week to 10/day. In this perspective it has been noted that the frequency of headaches/cluster periods is smaller in childhood. Similarly, the duration of the single cluster period is shorter (3,5,8,11).
It has also been observed that the temporal pattern shows a trend towards a gradual increase of frequency and duration of symptoms in adult life; however, the number of patients reported is too limited to draw conclusions.
It has also been reported that autonomic symptoms and signs seem to be less evident in children than in adults, although these data need to be confirmed (11).
Several treatment alternatives have been tried in the different case reports. In all of them the first- or second-line medication always turn out to be effective. To the best of our knowledge, no case of pharmacoresistance has been reported. According to these data, the most effective symptomatic treatments are oxygen (11,13,15,17), sumatriptan (15,20) and acetylsalicylic acid (16,21). Prophylactic treatments tried in literature are prednisone/prednisolone (12,21), indomethacin (16), pizotifen (15), verapamil (15,17,20), methysergide (11,21,22), loratadine (23), astemizole (23) and flunarizine (14). No controlled study has been reported.
In comparison with the reported cases, our patient appeared to be resistant to the majority of the most effective symptomatic and prophylactic medications.
According to the criteria of the International Classification of Headache Disorders, 2nd edn (1), our patient should be considered as having episodic CH, but presents peculiar features different from other cases of CH in childhood reported in the literature. This patient could represent a ‘variant’ clinical picture.
First, a prolonged first cluster period (8 months) seems unusual; at the beginning of the disease, short cluster periods are more frequently described, with the exception of the primary chronic forms. In addition, the remission was rather short (4.5 months). This temporal pattern may suggest a trend to chronic form for our patient, but needs to be confirmed at forthcoming follow-up.
Second, our patient represents the first case of CH in childhood resistant to symptomatic treatment, even with adult doses. Besides, the efficacy of the therapy with rivastigmine, methysergide, olanzapine and L-acetylcarnitine in aborting the bout is questionable. The progressive reduction in duration of episodes resembles more a spontaneousremission of a cluster period than a therapy effect (Fig. 2).
Third, the young age of onset and the severity of pain (no different from adults) already at this stage is another aspect not reported in previous case reports.
The last atypical pattern of our patient is the autonomic involvement. The sense of heat in the area of pain without pain and inconstant autonomic signs represent a pattern in concordance with what is seen in adult dissociation between pain and autonomic symptoms and signs.