Sage Journals: Discover world-class research

Abstract

Hereditary haemorrhagic telangiectasia (HHT) is a genetic disorder characterized by epistaxis, telangiectasia and visceral vascular manifestations. It is associated with migraine with aura due to pulmonary arteriovenous malformations (pAVMs). The aim of the study was to evaluate headache prevalence in 106 consecutive HHT patients (67 female, 39 male, age 53.5 ± 14.5 years) and age- and gender-matched controls. An extensive clinical work-up was performed and headache prevalence was determined. Lifetime prevalence of migraine was higher in HHT patients (39.6%) than in controls (19.8%) [P < 0.001, χ² = 12.17, odds ratio (OR) 3.0; 95% confidence interval 1.6 < OR < 5.7]. A positive association was confirmed between HHT patients with pAVMs and migraine with aura (38.5% vs. 10%). Furthermore, HHT patients without pAVMs had a higher prevalence of migraine without aura (11.5% vs. 26.3%; χ² = 11.85; d.f. = 2; P = 0.003). We speculate that pathophysiological mechanisms, including dysfunction of the transforming growth factor-beta pathways and resulting vascular changes, contribute to the higher prevalence of migraine without aura in HHT patients without pAVMs.

Keywords

Migraine right-to-left-shunt hereditary haemorrhagic telangiectasia Rendu–Osler–Weber syndrome

Introduction

A strong association between migraine with aura and patent foramen ovale (PFO) has been suggested and a causal relationship between a right-to-left shunt and migraine has been discussed (1, 2). An association has recently been reported between migraine and hereditary haemorrhagic telangiectasia (HHT, Rendu–Osler–Weber syndrome), an autosomal-dominant disorder with a prevalence of 5–13/100000 (2–5). HHT comprises the classical triad of epistaxis, telangiectasia and a positive family history. The fourth and important aspect of HHT is visceral involvement, which is also part of the current clinical diagnostic criteria, the Curaçao criteria (6). Depending on the genotype, visceral involvement of the lungs in the form of pulmonary arteriovenous malformations (pAVMs) occurs in 0–59% of HHT patients (7, 8). Cerebral vascular malformations (CVMs) have been observed in 1–16% of cases (8–10), hepatic ones (HVMs) in 2–41% (9, 10) and gastrointestinal involvement (GIT) in 15–70% (9, 11). Vascular malformations can cause severe complications: the shunting of pAVMs can lead to embolic strokes and brain abscesses, and CVMs are a cause of inherited stroke in young adults (12, 13). Quality of life is significantly reduced in patients with HHT, and migraine also contributes to this loss in quality of life: in a study of HHT patients the presence of severe headaches correlated with lower scores in the Short Form-36 scales ‘Vitality’ and ‘Mental health’ (14).

Mutations in two genes encoding components of the transforming growth factor (TGF)-β signalling pathway, the endoglin gene on chromosome 9q33-34 (15) and the activin receptor-like kinase 1 (ACVRL1; also referred to as ALK-1) gene on chromosome 12q13 (16), are known to cause HHT-1 and HHT-2. A rare subtype associated with juvenile polyposis is caused by a mutation of SMAD4 (17), a protein that is also involved in the TGF-β pathway. Two new loci have recently been mapped to chromosome 5 for HHT-3 (18) and to chromosome 7 for HHT-4 (19); however, the corresponding genes have not yet been identified. Beside the known mutations, the diagnosis of HHT remains a clinical one due to locus heterogeneity and practical difficulties of screening families with a unique mutation.

As mentioned above, previous work has suggested an association between HHT and migraine based on cerebral (20) and on pulmonary malformations (4, 5). Concerning the higher prevalence of migraine with aura in patients with cardiac right-to-left shunt due to a PFO and the possible resolution or reduction of migraine after closure of the PFO (21), the aim of the present study was to investigate the prevalence of headache in consecutive HHT patients, to decipher the influence of pAVMs on migraine prevalence in HHT patients, and to determine whether the underlying endothelial changes, mainly due to mutations of the TGF-β signalling pathway, may influence migraine prevalence and severity.

Methods

Patient selection and inclusion criteria

Consecutive HHT patients (n = 106; 67 female, 39 male, age 53.5 ± 14.5 years) who attended the out-patient clinic of the Department of Otorhinolaryngology at the Saarland University, Homburg, between November 2002 and November 2005 and who fulfilled the definite Curaçao criteria were included in the study. The Curaçao criteria define HHT as definite if at least three out of the four criteria (epistaxis, telangiectasia, visceral lesions, positive family history) are present, as possible or suspected if two criteria are fulfilled and as unlikely if fewer than two criteria are present (6).

All subjects were of German White descent and were referred from different parts of Germany to one of the national German centres for HHT in Homburg.

During their visit the patients received an extensive physical examination, including a neurological and otorhinolaryngological examination. Patients completed standardized questionnaires, including a detailed headache questionnaire. They were seen by a neurologist, and missing data (e.g. lack in the family history, missing headache diary during the out-patient visit) were additionally ascertained by a standardized telephone interview after the visit. Headache was diagnosed according to the 2nd International Headache Society (IHS) criteria (22). Patients who fulfilled the IHS criteria for migraine with aura (MA) and who described co-occurrence of migraine attacks without aura (MoA) were classified as migraineurs with aura.

All patients were investigated with a non-invasive bubble contrast sonographic shunt test to determine a right-to-left shunt. Either carotid Doppler sonography (80.2%) or transthoracic echocardiography (19.8%) with contrast medium was performed. Transthoracic echocardiography had a higher sensitivity than the intravenous angiography in the detection of pAVMs (23); carotid Doppler sonography has been shown to have also high sensitivity to detect central shunting (24). Furthermore, most HHT patients received detailed screening for visceral involvement, including cranial magnetic resonance imaging (MRI) (83%), MRI of the chest (89%) and MRI or sonography of the abdomen (78%).

An independent age and gender pair-matched control group of healthy volunteers (67 female, 39 male, age 53.9 ± 14.5 years) was explored with reference to headache prevalence and other comorbidities. As control subjects, age- and gender-matched non-consanguine partners of the HHT patients and volunteers were interviewed.

All analyses were carried out using spss 13.0 (SPSS Inc., Chicago, IL, USA). Clinical symptoms and history data were compared between the HHT patients and migraineurs without HHT by Mann–Whitney u-test for non-parametric variables (gender, frequency of nausea, vomiting, photophobia, phonophobia, unilateral pain, pulsating pain, aggravation by physical activity, avoidance of routine physical activity) and t-test for parametric variables (age, mean duration of untreated attack, mean number of attacks per month, mean intensity of pain).

Results

As defined by the inclusion criteria, all 106 HHT patients fulfilled the definite Curaçao criteria. The clinical symptoms of the patients are presented in Table 1. The lifetime prevalence of migraine was significantly higher in HHT patients (39.6%) than in controls (19.8%) [P < 0.001, χ² = 12.17, odds ratio (OR) 3.0; 95% confidence interval 1.6 < OR < 5.7]. The increased lifetime prevalence of migraine is based mainly on a higher prevalence of MoA than of MA. Prevalence of tension-type headache was not different between the groups (see Table 2a). There was no patient who reported a trigemino-autonomic headache, either in the control group or in the HHT group. About two-thirds of the 106 HHT patients reported a dull, bilateral, frontal and symptomatic headache during and a few hours after severe epistaxis. Out of 106 HHT patients, 26 (24.5%) had pAVMs. A detailed view on this special group revealed a significantly higher prevalence of MA in HHT patients with pAVMs (38.5%) than in HHT patients without pAVMs (10%) and a lower prevalence of MoA (11.5% vs. 26.3%) (P = 0.003, see Table 2b). The size of the right-to-left shunt was evaluated by MRI angiography according to the recently published method (25). We did not find significantly larger pAVMs in patients with MA and HHT in comparison with HHT patients with MoA or without migraine.

Table 1

Clinical symptoms of the 106 consecutive HHT patients and their controls


	HHT patients		Controls
Characteristics	n	%	n	%

Total	106	100	106	100
Age	53.5 ± 14.5		53.9 ± 14.5
Women	67	63.2	67	63.2
Age of first symptoms	22.2 ± 16.3
Curaçao criteria
Epistaxis	105	99.1
Telangiectasia	104	98.1
Visceral manifestation	50	47.2
CVM	3	2.8
pAVM	26	24.5
HVM	21	19.8
GIT	20	18.9
Positive family history	98	92.5

CVM, cerebral vascular malformation; GIT, gastrointestinal involvement; HHT, hereditary haemorrhagic telangiectasia; HVM, hepatic vascular malformation; pAVM, pulmonary arteriovenous malformation.

Table 2a

Headache life-time prevalence in HHT patients and age-matched controls


	HHT		Controls
	n	%	n	%	χ²	OR	95% CI	P-value

Migraine	42	39.6	21	19.8	12.18	3.01	1.6 < OR < 5.65	< 0.001
MoA	24	22.6	12	11.3	4.82	2.29	1.08 < OR < 4.87	< 0.043
MA	18	17.0	9	8.5	3.44	2.21	0.94 < OR < 5.16	< 0.098
TTH	41	38.7	46	43.4	0.49	0.82	0.48 < OR < 1.42	< 0.58
Episodic TTH	35	33.0	45	42.5	2.01	0.67	0.38 < OR < 1.17	< 0.20
Chronic TTH	6	5.7	1	0.9	3.69	6.3	0.75 < OR < 53.26	< 0.12

Table 2b

Migraine life-time prevalence in HHT patients with and without pulmonary arteriovenous malformations (pAVM)


	No migraine	MoA	MA	n


HHT without pAVM	51 (63.7%)	21 (26.3%)	8 (10%)	80
HHT with pAVM	13 (50%)	3 (11.5%)	10 (38.5%)	26
n	64 (60.4%)	24 (22.6%)	18 (17%)	106

Pearson's χ² test: χ² = 11.85; d.f. = 2; P = 0.003.

HHT, hereditary haemorrhagic telangiectasia; MA, migraine with aura; MoA, migraine without aura; TTH, tension-type headache.

Characteristics of the migraine attack differed initially between HHT patients and controls in two symptoms: the mean duration of an untreated attack was shorter in HHT patients, and avoidance of routine physical activity was less frequent in HHT patients than in controls (Table 3), but after correction of multiple comparisons the P-values were no longer significant. We found no significant differences in the quality of the aura. All 18 HHT patients with MA reported a visual aura, and eight of the nine migraineurs without HHT. Sensory aura symptoms were reported by two patients in each group and speech disturbances during the aura by one patient in each group.

Table 3

Overview of the migraine characteristics of the HHT patients and controls


Migraine characteristics	HHT patients	Controls	P-value

Total	n = 42	n = 21
Age	50 ± 12.2	48.8 ± 16.5	0.78
Women	83.3%	85.7%	0.81
Mean duration of untreated attack, h	15.9 ± 8.9	24.7 ± 15.1	0.01
			P corrected > 0.05
Mean number of attacks per month	2.2 ± 1.0	1.9 ± 0.9	0.22
Mean intensity of pain (VAS 1-10)	7.4 ± 0.8	6.7 ± 0.7	0.1
Nausea	81%	81%	0.83
Vomiting	17%	33%	0.14
Photophobia	88%	86%	0.79
Phonophobia	69%	71%	0.85
Unilateral location	86%	81%	0.63
Pulsating pain	93%	100%	0.21
Aggravation by physical activity	93%	95%	0.72
Avoidance of routine physical activity	71%	95%	0.03
			P corrected > 0.05

HHT, hereditary haemorrhagic telangiectasia; VAS, visual analogue scale.

Discussion

In our 106 patients we found an increased migraine lifetime prevalence of 39.6% in comparison with the control subjects. The prevalence of migraine in HHT patients is as high as Thenganatt et al. (5) described recently (38%) in their HHT population with the same gender ratio (65% female vs. 63.2% in our study). The lifetime prevalence in our control subjects (19.8%) is comparable to data of other studies (26). In contrast to the data published in two recent studies (4, 5), the higher migraine prevalence of our study was based more on the higher prevalence of MoA than on an increased prevalence of MA. Analysis of all patients revealed only a trend towards a higher life-time prevalence of MA in HHT patients (P < 0.098). Subanalysis showed that HHT patients with pAVMs have a higher prevalence of MA. These results are in accordance with prior studies (4, 5). A new and interesting observation of our study is the finding of a higher prevalence of MoA in HHT patients without a pulmonary shunt. Discussing these results, we postulate that there are other risk factors independent of the pulmonary right-to-left shunt for developing migraine in HHT patients.

An additional reason for the higher prevalence of MoA in HHT patients could be a vasculopathy due to a pathological pathway of TGF-β (27). This hypothesis is based on the mutations in HHT patients encoding components of the TGF-β signalling pathway: endoglin, which has direct effects on endothelial cell adhesion and migration, and ACVRL1. Interestingly, two studies have shown increased serum TGF-β1 during migraine attacks (28) and during headache-free periods (29). Although the function of TGF-β as a proinflammatory cytokine in migraine is unclear, a role in the neurogenic inflammation of the vessels in migraine is conceivable, especially because TGF-β1 is involved in modulation of cell growth, differentiation and repair of endothelium and immunomodulation. Furthermore, in the non-vertebrate model of Caenorhabditis elegans, the UNC-2/Ca(2+) channel, which is the closest orthologue in C. elegans of human CACNA1A, affects TGF-β-dependent regulation of tryptophan hydroxylase (TPH), the rate-limiting enzyme in serotonin synthesis (30, 31). Mutations in UNC-2 lead to lethargic and uncoordinated phenotypes of these animals. ‘Suppressor analysis’ as a classical genetic technique to identify signalling pathways of a protein has revealed that normal UNC-2 function antagonizes a TGF-β pathway modulating locomotion in these animals via expression of TPH. The UNC-2/TGF-β pathway seems to be required for accumulation of normal serotonin levels and stress-induced modulation of TPH expression in serotonergic chemosensory neurons of these animals (30). By transgenic expression of the migraine-associated calcium channel, CACNA1A, in these UNC-2 mutant animals Estevez et al. (30) showed a functional substitution for UNC-2 due to suppression of the lethargic and serotonin-deficient phenotype and has suggested a regulatory link between CACNA1A function and levels of serotonin. Based on these findings, normal serotonin levels require a functional steady state between the regulation of TPH expression by this TGF-β pathway and the regulatory antagonizing of the TGF-β pathway by the UNC-2/CACNA1A-channel. Lack of antagonizing by UNC-2 may lead to a more active TGF-β pathway and impaired serotonin homeostasis. The demonstration that these channels share the ability to modulate tph expression through their effects on TGF-β signalling provides a specific example of how CACNA1A function may influence levels of the critical migraine neurotransmitter serotonin. Therefore, it is possible that this TGF-β pathway that regulates TPH2-dependent serotonin synthesis is an additional factor in the increased prevalence of MoA in our study.

Proinflammatory effects of TGF-β1 might explain migraine symptoms. Patients with migraine often complain of fatigue or lack of vigour during and between migraine attacks. These symptoms may be related to increased TGF-β1, which has been noted to play a possible role in central fatigue or chronic fatigue syndrome (32). Concerning our findings in HHT patients, the higher prevalence of MoA could be explained by abnormal (increased) TGF-β levels or impaired TGF-β pathway due to mutations in endoglin or activin receptor-like kinase. Increased TGF-β plasma levels were found in 29 of 31 HHT patients (33), but did not differ between HHT-1 and -2. Interestingly, systemic sclerosis, another disease with increased serum TGF-β levels (34), is also associated with migraine (35).

As hypothesized in prior studies (20), CVMs could not explain the increased migraine prevalence. In our study, only three of 88 HHT patients who received a cranial MRI scan had CVMs, and only one of these had MA. Other studies have reported that 1–16% of all HHT patients have CVMs (8–10), but this number is also too low to explain the increased number of migraine in HHT patients. Embolic phenomena due to pulmonary shunts may be other reasons for the increased migraine prevalence. We used a non-invasive bubble contrast sonographic test for the detection of pAVMs. This method for shunt detection is rated as the most sensitive examination technique for pAVMs with a higher sensitivity than intravenous angiography (23). Nevertheless, it is not known if negative shunting tests for gaseous bubbles also exclude microscopic shunting for small molecules such as serotonin, which are normally cleared by endothelial cells in the lungs. A functional pulmonary leak may lead to disturbed serotonin homeostasis in the brain.

In the Head-HUNT study a linear trend of decreasing headache and migraine prevalence with decreasing haemoglobin was reported (36). Due to many episodes of epistaxis, haemoglobin levels in HHT patients are often low and have to be monitored regularly. We analysed whether higher haemoglobin levels are associated with a higher prevalence of migraine, but the haemoglobin results between HHT patients with migraine (12.1 + 2.1 g/dl) and without migraine (12.2 + 3.0 g/dl) did not differ significantly.

An important difference from the recent studies in HHT patients (4, 5) concerning migraine prevalence is that data of our study are based on detailed neurological investigation and for missing data on an additional phone interview. The discrepant results of the studies might be explained by this quite accurate form of data acquisition. Especially visual disturbances, not fulfilling criteria of the visual aura, are sometimes misinterpreted as aura phenomena by patients and by the non-headache specialist and may therefore be incorrectly documented as MA in the charts.

In summary, we were able to confirm the positive association between HHT patients with pAVMs and MA. Furthermore, we have demonstrated a higher prevalence of MoA in HHT patients without pAVMs. We speculate that pathophysiological mechanisms, including dysfunction of the TGF-β pathways and resulting vascular changes, contribute to the higher prevalence of MoA.

References

Tepper

Sheftell

Bigal

. The patent foramen ovale-migraine question. Neurol Sci 2007; 28 (Suppl. 2):S118–23.

Post

Thijs

Schonewille

Budts

Snijder

Plokker

Westermann

. Embolization of pulmonary arteriovenous malformations and decrease in prevalence of migraine. Neurology 2006; 66:202–5.

Kjeldsen

Vase

Green

. Hereditary hemorrhagic telangiectasia (HHT): a population-based study of prevalence and mortality in Danish HHT patients. J Intern Med 1999; 245:31–9.

Post

Letteboer

Mager

Plokker

Kelder

Westermann

. A pulmonary right-to-left shunt in patients with hereditary hemorrhagic telangiectasia is associated with an increased prevalence of migraine. Chest 2005; 128:2485–9.

Thenganatt

Schneidermann

Hyland

Edmeads

Mandzia

Faughnan

. Migraines linked to intrapulmonary right-to-left shunt. Headache 2006; 46:439–43.

Shovlin

Guttmacher

Buscarini

Faughnan

Hyland

Westermann

CJ.

Diagnostic criteria for hereditary hemorrhagic telangiectasia (Rendu–Osler–Weber syndrome). Am J Med Genet 2000; 91:66–7.

Berg

Porteous

Reinhardt

Gallione

Holloway

Umasunthar

Hereditary haemorrhagic telangiectasia: a questionnaire based study to delineate the different phenotypes caused by endoglin and ALK1 mutations. J Med Genet 2003; 40:585–90.

Kjeldsen

Moller

Brusgaard

Vase

Andersen

. Clinical symptoms according to genotype amongst patients with hereditary haemorrhagic telangiectasia. J Intern Med 2005; 258:349–55.

Letteboer

Mager

Snijder

Koeleman

Lindhout

Ploos van Amstel

Westermann

. Genotype–phenotype relationship in hereditary hemorrhagic telangiectasia. J Med Genet 2006; 43:371–7.

10.

Bayrak–Toydemir

McDonald

Markewitz

Lewin

Miller

Chou

LS.

Genotype–phenotype correlation in hereditary hemorrhagic telangiectasia: mutations and manifestations. Am J Med Genet A 2006; 140:463–70.

11.

Plauchu

de Chadarevian

Bideau

Robert

. Age-related clinical profile of hereditary hemorrhagic telangiectasia in an epidemiologically recruited population. Am J Med Genet 1989; 32:291–7.

12.

Begbie

Wallace

GMF

Shovlin

. Hereditary haemorrhagic telangiectasia (Osler–Weber–Rendu syndrome): a view from the 21st century. Postgrad Med J 2003; 79:18–24.

13.

Cottin

Chinet

Lavolé

Corre

Marchand

Reynaud–Gaubert

Pulmonary arteriovenous malformations in hereditary hemorrhagic telangiectasia: a series of 126 patients. Medicine 2007; 86:1–17.

14.

Geisthoff

Heckmann

D'Amelio

Grünewald

Knöbber

Falkai

König

. Health-related quality of life in hereditary hemorrhagic telangiectasia. Otolaryngol Head Neck Surg 2007; 136:726–33.

15.

McAllister

Grogg

Johnson

Gallione

Baldwin

Jackson

CE.

Endoglin, a TGF-beta binding protein of endothelial cells, is the gene for hereditary haemorrhagic telangiectasia type 1. Nat Genet 1994; 8:345–51.

16.

Johnson

Berg

Baldwin

Gallione

Marondel

Yoon

SJ.

Mutations in the activin receptor-like kinase 1 gene in hereditary haemorrhagic telangiectasia type 2. Nat Genet 1996; 13:189–95.

17.

Gallione

Repetto

Legius

Rustigi

Schelley

Teipar

A combined syndrome of juvenile polyposis and hereditary haemorrhagic telangiectasia associated with mutations in MADH4 (SMAD4). Lancet 2004; 363:852–9.

18.

Cole

Begbie

Wallace

Shovlin

. A new locus for hereditary haemorrhagic telangiectasia (HHT3) maps to chromosome 5. J Med Genet 2005; 42:577–82.

19.

Bayrak–Toydemir

McDonald

Akarsu

Toydemir

Calderon

Tuncali

A fourth locus for hereditary hemorrhagic telangiectasia maps to chromosome 7. Am J Med Genet A 2006; 140:2155–62.

20.

Steele

Nath

Burn

Porteous

. An association between migrainous aura and hereditary haemorrhagic telangiectasia. Headache 1993; 33:145–8.

21.

Finsterer

Sommer

Stiskal

Stollberger

Baumgartner

. Closure of a patent foramen ovale: effective therapy of migraine and occipital stroke. Int J Neurosci 2005; 115:119–27.

22.

Headache Classification Subcommittee of the International Headache Society. The International Classification of Headache Disorders, 2nd edition. Cephalalgia 2004; 24 (Suppl. 1):1–160.

23.

Saluja

Henderson

White

Jr . Embolotherapy in the bronchial and pulmonary circulations. Radiol Clin North Am 2000; 38:425–48.

24.

Topcuoglu

Palacios

Buonanno

. Contrast M-mode power Doppler ultrasound in the detection of right-to-left shunts: utility of submandibular internal carotid artery recording. J Neuroimaging 2003; 13:315–23.

25.

Schneider

Uder

Koehler

Kirchin

Massmann

Buecker

MR angiography for detection of pulmonary arteriovenous malformations in patients with hereditary hemorrhagic telangiectasia. Am J Roentgenol 2008; 190:892–901.

26.

Stovner

Hagen

Jensen

Katsarava

Lipton

Scher

AI.

The global burden of headache: a documentation of headache prevalence and disability worldwide. Cephalalgia 2007; 27:193–210.

27.

Lebrin

Deckers

Bertolino

ten Dijke

. TGF-beta receptor function in the endothelium. Cardiovasc Res 2005; 65:599–608.

28.

Tietjen

Dore-Duffy

Beaumont

Athanas

Welch

KMA

. Ictal increase in vascular endothelial growth factor, and transforming growth factor beta 1 in young woman with migraine. Headache 1999; 39:383.

29.

Ishizaki

Takeshima

Fukuhara

Araki

Nakaso

Kusumi

Nakashima

. Increased plasma transforming growth factor-beta1 in migraine. Headache 2005; 45:1224–8.

30.

Estevez

Cowie

Gardner

. The voltage-gated calcium channel UNC-2 is involved in stress-mediated regulation of trypophan hydroxylase. J Neurochem 2004; 88:102–13.

31.

Estevez

. Intervertebrate modeling of a migraine channelopathy. Headache 2006; 46 (Suppl. 1):25–31.

32.

Inoue

Yamazaki

Manabe

. Transforming growth factor-beta activated during exercise in brain depresses spontaneous motor activity of animals. Relevance to central fatigue. Brain Res 1999; 846:145–53.

33.

Sadick

Riedel

Naim

Goessler

Hörmann

Hafner

Lux

. Patients with hereditary hemorrhagic teleangiectasia have increased plasma levels of vascular endothelial growth factor and transforming growth factor beta 1 as well as high ALK1 tissue expression. Haematologica 2005; 90:818–28.

34.

Fujimoto

Hasegawa

Hamaguchi

Komura

Matsushita

Yanaba

A clue for telangiectasis in systemic sclerosis: elevated serum soluble endoglin levels in patients with the limited cutaneous form of the disease. Dermatology 2006; 213:88–92.

35.

Pal

Gibson

Passmore

Griffiths

Dick

. A study of headaches and migraine in Sjögren's syndrome and other rheumatic disorders. Ann Rheum Dis 1989; 48:312–16.

36.

Aamodt

Borch–Iohnsen

Hagen

Stovner

Asberg

Zwart

. Headache prevalence related to haemoglobin and ferritin. The HUNT Study. Cephalalgia 2004; 24:758–62.

An Association of Migraine with Hereditary Haemorrhagic Telangiectasia Independently of Pulmonary Right-to-Left Shunts

Abstract

Keywords

Introduction

Methods

Patient selection and inclusion criteria

Results

Discussion

References