Abstract
A 60-year-old woman with secondary chronic cluster headache had increased serum ferritin and serum transferrin saturation and was homozygous for the C282Y mutation in the HFE gene, which is indicative of hereditary haemochromatosis. Treatment with venesection that normalized her iron stores led to a radical improvement of her headache complaints that had been daily for several years. Later, the headache returned to some degree in spite of normal serum ferritin levels. Her cousin, a 33-year-old man who had had episodic cluster headache for several years, also had increased transferrin saturation and was compound heterozygous for two mutations, a genotype known to be associated with a slightly increased frequency of haemochromatosis. This is the first report of a headache disorder in a patient with hereditary haemochromatosis. The coexistence of the two disorders may be a mere coincidence, but the temporary improvement of headache from depletion of iron stores may indicate a causal relation, possibly mediated by iron deposits in pain-modulating centres in the brainstem.
Introduction
Hereditary haemochromatosis (HH) is an autosomal recessive disease causing increased iron absorption. Two mutations in the haemochromatosis gene (HFE gene) on chromosomal locus 6p21.3 have been identified. In Norway, homozygosity for the C282Y mutation accounts for 88% of newly diagnosed cases, whereas compound heterozygosity for the C282Y/H63D mutation accounts for about 5% of the cases (1).
The increased iron absorption may over time give progressive organ affection due to iron deposits in the liver, pancreas with diabetes mellitus, and increased skin pigmentation. Iron overload may also induce neuronal damage in the brain, and an association between iron overload and epilepsy has been suggested (2). Arthropathy, however, which is present in more than 25% of cases, is often not related to iron deposits (3). Headache has till now never been related to HH.
The present paper reports the coexistence of HH in a female with chronic cluster headache. Her headache responded partially to treatment of her HH. The history of a male cousin with episodic cluster headache and possibly HH is also presented.
Patients
Patient 1
This 60-year-old woman, who had been operated on for varicose veins and for gonarthritis, had had migraine with visual auras from puberty until childbirth about 30 years ago. Her mother and one of her three children had headache, possibly migraine. Fourteen years ago she started having nightly attacks, with severe pain in her right upper jaw, lasting a few minutes. During the first years, such attacks came sporadically, with intervals of several months, usually 2–3 h after falling asleep. Eleven years ago these intense attacks started to last longer and the pain was localized in the right eye and frontal region. During the attacks she paced the floor and there was lacrimation, conjunctival injection and a slight ptosis in the right eye, and rhinorrhea. She had phono- and photophobia, but no nausea or vomiting. Ten years ago she tried lithium, which seemed to be effective for 3 years, but attacks came back even on lithium therapy. Oxygen and ergotamine injections were effective against individual attacks.
Six years ago she started with verapamil, which also seemed to have a prophylactic effect for some time, and sumatriptan injections were remarkably effective even during very intense attacks. Attacks usually lasted between 30 and 60 min and came three to five times every night. Three years ago attacks again increased in frequency and she had up to eight attacks every 24 h. She now again tried various prophylactics (lithium, verapamil, valproic acid), also in combinations, without effect. Due to the high frequency of attacks, a tentative diagnosis of chronic paroxysmal hemicrania was made and indomethacin was given in daily doses of 75 mg and then 150 mg, but with no beneficial effect. Prednisolone 50 mg/day was partly effective, but when the dose was tapered off below 15 mg attacks returned. She still had a good effect from sumatriptan injections, and she also had an effect from oxygen on some milder attacks. During the last 7 or 8 years she also had periodic pain in the neck radiating to the right hand. EEG and cerebral MRI performed 8 years ago, and a cerebral angiography performed 2 years ago, were all normal.
Extensive blood samples taken 2.5 years ago were quite normal, except for serum (s)-ferritin, which was moderately increased (364 µg/L, upper reference limit 150 µg/L). For this, she was referred to a haematologist, who found a s-transferrin saturation of nearly 75%, indicating that she had haemochromatosis (decision limit for women: 50%). On clinical examination there was no liver enlargement, signs of diabetes or excessive skin pigmentation. A gene test showed that she was homozygous for the C282Y mutation on locus 6p 21.3 in the HFE gene. During this period she was also examined for hip pain and an orthopaedic examination indicated bilateral coxarthrosis. She had a hip joint prosthesis a year ago on the left side.
Due to her haemachromotosis, venesection was started 2.5 years ago, with the removal of 500 ml venous blood almost weekly for 2 months. S-ferritin had then dropped to below 100 µg/L. The pain continued almost unchanged during this period and she used several daily doses of sumatriptan injections and also oxygen. Approximately 2 weeks after the last venesection, however, she consulted a so-called healer three times. She felt that his treatment relieved her of body tension. In the same period, approximately 10 weeks after the first venesection, the attack frequency was dramatically reduced, and a few weeks later she herself discontinued all prophylactic and acute medication. After 6 months she again got some mild attacks; s-ferritin had then increased to more than 100 µg/L. She contacted the hospital and after a new series of venesections the attacks disappeared once more. She had had long periods on sick leave during the years with chronic cluster headache, but she could resume 50% of work 1 year after the headache was improved. Also, her hip pain got somewhat better. One and a half years after she got better the attacks appeared once more, up to three times a day, even though the s-ferritin levels had been below 100 µg/L on several occasions, and she is once more on sick-leave. During the last months she has also had some attacks of migraine without aura, treated with sumatriptan. She clearly distinguishes these attacks from her cluster headache.
Patient 2
This 33-year-old man is the cousin of patient 1. He had previously been healthy, except for this headache. He has four elder siblings who are healthy.
Twenty-six years ago he started to have attacks in the left side of the head, behind the eye and in the temporal region, always on the left side. Attacks lasted from 1 to 3 h and they could come both during the night and day. In the worst periods, there were several attacks every 24 h. Attack periods lasted from a few weeks to several months, usually in winter or spring. During the attacks there was lacrimation and conjunctival injection of the left eye and also nasal congestion. He also had a marked bradycardia, down to 40 beats/s. Neurological examination was normal. He had for several years tried ergotamine and oxygen with good effect on attacks. Verapamil as a prophylactic was ineffective. During the last 10 years the attacks tended to become more irregular, with quite short attack periods lasting from 3 days to 2 weeks in which there were up to four attacks per day. Then he would have sporadic attacks for several months before he had a new attack period. During the last year he had excellent effect from sumatriptan injections.
Blood sample analysis showed that he had a serum iron concentration of 36.7 µmol/L (reference limits 12.0–27.0); s-ferritin was 71 µg/L (reference limits 24–200 µg/L), and the s-transferrin saturation was 63% (normal in men <55%). A gene test showed that he was heterozygous for both the C282Y and H63D mutations, i.e. he was compound heterozygous. This patient has not been treated with venesection.
Discussion
To our knowledge, cluster headache has previously never been reported in patients with HH. There are also no reports of other headache types in HH patients. The two patients described here both fulfilled the International headache Society (IHS) criteria for cluster headache. Patient 1 had a secondary chronic type, whereas patient 2 had the episodic type. The fact that the two patients were relatives is not surprising as cluster headache is known to be 14 times more prevalent among first-degree relatives and twice as prevalent among second-degree relatives, compared with the general population (4).
As to the diagnosis of HH, no internationally accepted criteria exist. The diagnosis may be made on the basis of genotype (C282Y/C282Y, or C282Y/H63D), pathology (abnormal iron deposits in various organs), biochemical abnormalities (increased s-ferritin and/or increased s-transferrin saturation), or a certain clinical syndrome. Patient 1 undoubtedly had HH with moderately increased s-ferritin and markedly increased s-transferrin saturation, and she was homozygous for C282Y mutations in the HFE gene. She had arthropathy, which may be part of the clinical syndrome of HH, but she did not have signs of damage to the liver and pancreas, or increased skin pigmentation. In patient 2 the diagnosis is definitely less certain, although it is known that compound heterozygotes have a moderately increased risk for developing clinical haemochromatosis. As indicated by the normal s-ferritin level, iron stores were not increased, but the s-transferrin saturation level was clearly abnormal.
The fact that HH and cluster headache coexisted in patient 1 is no proof that the two disorders are related. The prevalence of HH in a Norwegian population sample has been found to be 0.7% (1), which is somewhat higher than in Australia (5). The genotype has the same frequency in men and women, but clinical disease is more than twice as frequent in men because females are protected from iron overload by blood loss during menstruation and childbirth. Cluster headache probably occurs in about 0.05–0.1% of the population, and is four to five times more frequent in men than in women. From these figures it can be calculated that HH and cluster headache would coexist by chance in 1/280 000 people(1/2000 (0.05%) × 1/140 (0.7%)). This means that approximately 15 Norwegians would have this combination of diseases even if the diseases are totally unrelated.
The fact that the cluster headache in patient 1 improved dramatically when her iron deposits had been normalized may indicate that the two conditions were related in this patient, but the temporal relation between the therapeutic phlebotomies and the improvement may be incidental. She became herself rather convinced of a causal relation between her cluster headache and the haemochromatosis, so later she demanded to have a series of venesections when the s-ferritin levels were above 150 µg/L and she started having mild attacks. During the last 8 months, however, she has had rather frequent attacks even though her s-ferritin level has been below 100 µg/L. The fact that there was, at best, only a partial response to normalization of iron stores does not disprove a causal relation between HH and cluster headache, as some other symptoms of HH, e.g. arthropathy, are also known to respond poorly to treatment.
Recently, elevated iron levels have been demonstrated with magnetic resonance imaging in the periaqueductal grey matter of patients with migraine and chronic daily headache as compared with controls (6). From that study, it can not be known whether the increased iron deposits were the cause of the headache. If it can be shown that increased iron in areas of the brain related to pain modulation plays a causal role in headache, it would seem highly probable that HH could worsen or even cause headache.
Most often symptoms and signs of haemochromatosis are caused by iron deposits leading to organ damage. Athropathy, however, may be present even if iron stores are normal (3, 7). There may be several explanations for this. One possibility is that the mutations may have other consequences than increased intestinal iron absorption. Hence, also headache could theoretically be related to HH even in cases where there are few or no signs of increased iron deposits. The present cases indicate that the relation between HH and cluster headache and also other headache types should be further investigated.