Experience of the Irish National Centre for hereditary haemorrhagic telangiectasia 2003–2008

Respiratory Medicine (2010) 104, 1218e1224 available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/rmed

Experience of the Irish National Centre for hereditary haemorrhagic telangiectasia 2003e2008 C.F. Ni Bhuachalla a, T.M. O’ Connor b,*, M. Murphy a, N. Colwell b, A. Brady a a b

Dept. of Radiology, National HHT Centre, Mercy University Hospital, Cork, Ireland Dept. of Respiratory Medicine, National HHT Centre, Mercy University Hospital, Grenville Place, Cork, Ireland

Received 13 November 2009; accepted 6 March 2010 Available online 1 April 2010

KEYWORDS Hereditary haemorrhagic telangiectasia; Rendu-Osler-Weber syndrome; Pulmonary arteriovenous malformation; Cerebral arteriovenous malformation; Embolisation

Summary Hereditary haemorrhagic telangiectasia (HHT) is a group of autosomal dominant disorders of vascular structure. The Irish National Centre for HHT at the Mercy University Hospital, Cork, Ireland was founded in 2003. From 2003 to 2008, screening of 164 patients with contrast echocardiography, thoracic computerised tomography (CT) and cerebral magnetic resonance imaging (MRI) has identified 88 patients with definite HHT, 72 (82%) of whom had epistaxis, 70 (80%) had telangiectasia and 81 (92%) had a first-degree relative with HHT. We sought to describe the manifestations of HHT in an Irish population and to determine differences between internationally reported data. The HHT patient database was analysed to describe demographics, clinical manifestations and interventional procedures performed in all referred patients. Contrast echocardiography and/or CT were performed in 86 patients with definite HHT, identifying 27 patients (31%) with pulmonary arteriovenous malformations (pAVMs). Nineteen patients with single or multiple pAVMs had 28 embolisation procedures performed, with 1e6 pAVMs embolised per procedure. Cerebral MRI was performed in 78 (89%) patients and 2 (2.3%) had cerebral arteriovenous malformations (cAVMs). HHT prevalence is thought to be 1 in 2500e8000, suggesting that there are many undiagnosed cases in Irish patients. Internationally published data suggest a prevalence of 15e35% for pAVMs and 10e23% for cAVMs in patients with HHT. While the prevalence of pAVMs in our group is consistent with these data, the prevalence of cAVMs is considerably lower, suggesting that Irish patients with HHT may differ genotypically and phenotypically from those in other countries. ª 2010 Elsevier Ltd. All rights reserved.

* Corresponding author. Tel.: þ353 21 4935325; fax: þ353 21 4935350. E-mail address: [email protected] (T.M. O’ Connor). 0954-6111/$ - see front matter ª 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.rmed.2010.03.003

Hereditary haemorrhagic telangiectasia in Ireland

Introduction Hereditary haemorrhagic telangiectasia (HHT) is a group of autosomal dominant multisystem vascular dysplasias characterised by recurrent epistaxis with cutaneous and mucosal telangiectases (Fig. 1) and visceral arteriovenous malformations (AVMs).1 It is historically known as Rendu-Osler-Weber syndrome, after three physicians who, in the late 19th and early 20th century, recognised this hereditary disorder involving nosebleeds and characteristic cutaneous and mucosal red spots. The title HHT was devised by Hanes in 1909.1 HHT affects 1 in 8000e10,000 individuals but is subject to ethnic and geographical variation, with prevalence reported as 1 in 2500 in some regions of France.2 Penetrance and expressivity are also subject to variation, even within a single affected family,2 that is, an affected individual with epistaxis, telangiectasia and pulmonary AVMs may have a sibling with epistaxis only. A number of HHT subtypes have been identified including HHT type 1, associated with an endoglin gene mutation (chromosome 9) and HHT type 2 associated with an ALK-1 (AC-VRL1) gene mutation (chromosome 12). Other chromosomes involved in HHT include chromosomes 5, 7 and 18 (MADH4 or SMAD4 gene).2,3 These genes encode proteins expressed on vascular endothelial cells which are involved with signalling by the transforming growth factor-beta (TGF-b) family. TGF-b signalling affects cellular growth and differentiation. Thus abnormal angiogenesis in HHT is thought to occur due to aberrant TGF-b signalling caused by disturbed protein function from endoglin and ALK-1 genetic mutations.4 Genetic testing may help to diagnose HHT in a patient with an affected relative who has a known HHT mutation.5 The understanding of HHT and its complications continues to evolve since this condition was first recognised. The Scientific Advisory Board of the HHT Foundation International established clinical criteria for the diagnosis of HHT (the Curac ¸ao criteria) in 1998 (Table 1). HHT diagnosis is considered definite if three criteria are present,

Figure 1

1219 possible or suspected if two criteria are present and unlikely if fewer than two criteria are present.6 More recently, international guidelines for the diagnosis and management of HHT have been published.7 HHT is a relatively common but under-recognised condition. The importance of screening for this disorder must be emphasised as up to 10% of affected individuals suffer disability or premature death as a result of complications associated with HHT.5 These complications are often due to an AVM’s increased likelihood of rupture when compared with a regular blood vessel. In the case of pulmonary AVMs (pAVMs) (Fig. 2) complications also arise from right-to-left shunting of blood leading to cerebrovascular accident or brain abscess due to bland or septic emboli entering the cerebral circulation.8 Therefore, all patients with suspected or confirmed HHT, as well as first-degree relatives of definite cases, should be referred to a specialist centre for appropriate screening.2 The Irish National HHT Centre was established in the Mercy University Hospital (MUH) in Cork, Ireland in January 2003. This centre aims to identify patients with this condition, to identify all affected family members through screening, to identify the manifestations of HHT in each person affected and to manage these appropriately. Screening tests for visceral AVMs include contrast echocardiography, computerised tomography (CT) of thorax and magnetic resonance imaging (MRI) of brain. Interventions for epistaxis, pAVMs and cerebral AVMs (cAVMs) are carried out as necessary and include nasal cautery, nasal septal dermoplasty, nasal grafting and AVM embolisation. This paper reports the experience of the Irish National HHT centre from 2003 to 2008, with particular reference to the prevalence of AVMs in this patient population compared with other international data.

Methods Patients referred to the Irish National HHT Centre were invited to attend for clinical history and physical examination; further investigations in those with clinical features of HHT

67 Year male with HHT and multiple mucocutaneous telangiectasia.

1220 Table 1

C.F. Ni Bhuachalla et al. Curac ¸ao criteria for the diagnosis of HHT.

Spontaneous and recurrent epistaxis Multiple telangiectasia at characteristic sites including face, lips, oral mucosa and fingers Visceral involvement: gastrointestinal telangiectasia (with or without bleeding) or visceral AVMs (pulmonary, cerebral, hepatic or spinal) Family history: first-degree relative affected (in accordance with these criteria) HHT diagnosis is considered definite if three criteria are present, possible or suspected if two criteria are present and unlikely if fewer than two criteria are present.6

Evidence of HHT Follow-up Rescreen if cerebral white matter lesions on MRI Contrast Echocardiography Negative

Positive for pAVMs

Negative

included contrast echocardiography and contrast-enhanced cerebral MRI, according to an established management algorithm (Fig. 3). Among the patients referred, 69% were screened because of a family history of HHT and the remainder had epistaxis, telangiectasia and/or pAVMs detected by other physicians and were therefore screened. Contrast echocardiography, also known as an ‘echobubble’ study, involves injecting 10 ml of agitated saline into a peripheral vein and observing its course through the cardiac chambers with transthoracic echocardiography. Under normal circumstances, minute air bubbles appear rapidly in the right heart and then dissipate as they are filtered in the pulmonary circulation. An intracardiac shunt manifests as bubbles appearing in the left heart chambers within one cardiac cycle of appearing in the right side of the heart. When the patient has a pAVM, bubbles will be visualised in the left heart chambers after three to 8 cardiac cycles as the bubbles are shunted through the AVM.9 Patients with a positive echo-bubble in this study proceeded to non-contrast CT thorax (Siemens Somatom volume zoom CT) to identify pAVMs. Pulmonary AVMs with feeding arteries of more than 3 mm diameter were embolised. Feeding arteries were localised by selective pulmonary angiography. The feeding arteries were then selectively cannulated and embolised using stainless steel

Figure 2 Typical pAVM in a patient with definite HHT with a large feeding vessel.

Cerebral MRI if > 7yrs

Thoracic CT

Feeding artery <3mm

Feeding artery ≥3mm

Positive Antibiotic prophylaxis for dental/surgical procedures when pAVMs not excluded by negative echobubble

Follow-up at 5 yr interval (3 yr interval age 12-18, frequently in pregnancy)

Embolize Repeat CT at 6 months to confirm occlusion

Figure 3 Screening and management algorithm for patients with suspected HHT.

coils (Cook, Denmark) or Amplatzer plugs (AGA Medical, Plymouth, Maine, USA). Follow-up is by non-contrast CT thorax, usually six months after a successful embolisation, and 3e5 yearly thereafter.2 Rapid pAVM enlargement can occur during puberty and pregnancy. Increased frequency of screening is recommended during these intervals.10 A typical pAVM with a large feeding vessel is shown in Fig. 2. Figs. 4 and 5 show a pAVM and embolisation of a pAVM (both using pulmonary angiography). The MRI studies of patients in whom cAVMs were identified were reviewed by a consultant neuroradiologist with extensive neuro-vascular embolisation experience; his assessment of the need for further intervention for cAVMs was followed. Patients with mucocutaneous manifestations, epistaxis or GI telangiectasia were assessed as needed by a Consultant Dermatologist, ENT surgeon or Gastroenterologist. In general, screening for liver disease in HHT is not routinely recommended, as most affected patients are asymptomatic.2 Data were recorded prospectively from patients with suspected HHT. Details recorded included demographics of the index case and of family members affected, clinical features of the index case, screening tests carried out and their results, treatment of pAVM, cAVM and epistaxis, and follow up information. The database, at the time of analysis, consisted of one hundred and sixty-four patients from 43 families from all over Ireland. Patients were male and female, ranging in age from 7 to 71 years. Statistical analysis was carried out using SPSS software (SPSS, Chicago, IL).

Hereditary haemorrhagic telangiectasia in Ireland

1221

Results

Figure 4 pAVM with feeding artery and draining vein (pulmonary angiography).

Between 2003 and 2008, screening of one hundred and sixtyfour patients and their families identified 88 patients (54%) with definite HHT (based on the presence of 3 or more Curac ¸ao criteria), 26 patients (16%) with possible or suspected HHT (based on the presence of 2 Curac ¸ao criteria) and 50 patients (30%) unlikely to have HHT (based on the presence of fewer than 2 Curac ¸ao criteria) (Table 2). Results of cAVM and pAVM screening were incorporated into the clinical diagnosis for each patient. Among the screened population, forty-two percent were male, 84% had a family history of HHT, 56% had a history of epistaxis and 57% had cutaneous telangiectasia. Among the group of eighty-eight patients with definite HHT, 53% were male, 92% had a family history of HHT, 82% had a history of epistaxis and 80% had cutaneous telangiectasia (Table 3). Contrast echocardiography and/or non-contrast CT were performed in eighty-six patients, identifying 27 patients (31%) with pAVMs. Nineteen patients with single or multiple pAVMs had 28 embolisation procedures performed, with 1e6 pAVMs embolised per procedure. While pleuritic chest pain was common (20%) after pAVM embolisation, the only serious complication was migration of a 12 mm Cook Nester coil to the left lower limb during attempted embolisation in a patient with a large pAVM with a 13 mm feeding artery. The device was successfully retrieved with a snare and the pAVM was embolised without further complications. Contrast-enhanced cerebral MRI was performed in seventy-eight patients (89%), 1 of whom had a small cAVM requiring follow up only. One patient had a previously diagnosed cAVM which had bled many years previously, resulting in a significant neurological deficit. It was felt that no new intervention was required for this cAVM. No patient had other vascular malformations of cerebral vessels. Thirty-two families (out of 43 assessed) with at least one member with definite HHT were identified. Among the group of twenty-six patients with possible or suspected HHT, 31% were male, 85% had a family history of HHT, 69% had a history of epistaxis and 65% had cutaneous telangiectasia (Table 3). Contrast echocardiography and/or non-contrast CT were performed in twenty-six patients, identifying 5 patients (19%) with pulmonary pAVMs, none of which fulfilled the criteria for embolisation. Contrastenhanced cerebral MRI was performed in twenty-five patients (96%), none of whom had cAVMs. Among the group of fifty patients unlikely to have HHT, 28% were male, 68% had a family history of HHT, 4% had a history of epistaxis and 12% had cutaneous telangiectasia (Table 3). Contrast echocardiography and/or non-contrast CT were performed in forty-four patients, but no pAVMs were detected. Contrast enhanced cerebral MRI was performed in thirtyeight patients (76%), none of whom had cAVMs. No cases of pulmonary hypertension were detected, either by echocardiography during screening or by direct measurement of pulmonary artery pressure in those who had pAVM embolisation.

Discussion Figure 5 Complete embolisation of feeding artery with stainless steel coils (pulmonary angiography).

One hundred and sixty-four subjects were screened in the Irish National HHT centre from 2003 to 2008. More than half

1222

C.F. Ni Bhuachalla et al.

Table 2 Baseline population.

characteristics

of

Characteristics

Number

Total screened Gender Male Female Age at screening (yrs) 0e25 26e50 51e75 Family history of HHT Epistaxis Telangiectasia HHT Diagnosis Definite Possible/Suspected Unlikely

164

the

screened %

69 95

42.1 57.9

58 79 27 137 92 93

35.4 48.1 16.5 83.5 56.1 56.7

88 26 50

54 16 30

of those screened were diagnosed with definite HHT (54%), and one third of this group had pAVMs of which 70% required embolisation. Two patients (2.3%) with definite HHT had cAVMs, one that had already bled and 1 that was considered too small to require embolisation. Other complications such

Table 3

HHT-specific diagnosis and management.

Number Age at screening (yrs) <18 yrs at screening Gender Male Female Family History of HHT Epistaxis Epistaxis treatment Nasal cautery Septodermoplasty Skin Graft Telangiectasia Contrast Echo Positive study CT Thorax pAVMs present Multiple pAVMs Patients embolised Patients reembolised Embolisation procedures MRI Brain cAVMs present White matter change

Definite HHT

Suspected HHT

Unlikely HHT

88 41.4  14.69 3 (3.4%)

26 37.3  17.59 4 (15.4%)

50 29.8  15.06 10 (20%)

47 (53%) 41 (47%) 81 (92%)

8 (31%) 18 (69%) 22 (85%)

14 (28%) 36 (72%) 34 (68%)

72 (82%)

18 (69%)

2 (4%)

16 2 1 70 73 46 61 27 12 19 9 28

5 0 0 17 25 16 15 5 2 0 0 0

(18%) (2%) (1%) (80%) (83%) (52%) (69%) (31%) (14) (22%) (10%)

78 (89%) 2 (2.3%) 9 (10%)

(19%)

0 0 0

(65%) (96%) (61%) (58%) (19%) (8%)

6 40 8 13 0 0 0 0 0

25 (96%) 0 3 (12%)

(12%) (80%) (16%) (26%)

38 (76%) 0 1 (2%)

as epistaxis requiring active management, anaemia from GI telangiectasia causing blood loss and AVMs in other organs were infrequent. Among those with possible HHT, 19% had small pAVMs that require follow up and none had cAVMs. Symptoms and signs of HHT tend to develop late in childhood and early adulthood.12 Epistaxis is the most common symptom,13 as was the case in our population. pAVMs are prevalent in fifteen to 35% and cAVMs in 10e23% of patients with HHT.2,5,8,11,14,15 pAVMs are associated with haemoptysis, haemothorax, migraine, seizure, transient ischaemic attacks, cerebrovascular accidents and brain abcesses.2,5,8,9,16 The pathogenesis of pAVM associated neurological events arises from the loss of the filtering capacity of the pulmonary circulation with pulmonary rightto-left shunting resulting in paradoxical emboli passing directly into the cerebral circulation.2,5 The importance of detecting AVMs is underscored by a large observational study carried out by Easey et al. which found that patients with HHT were twenty-three times more likely to have a neurological event.17 These neurological complications occur in up to twenty-seven percent of HHT patients but are more commonly a result of a pAVM embolic event (61%) than a direct complication of a cAVM (28%).14 Screening of subjects in our centre detected a pAVM prevalence of thirty-one percent and a cAVM prevalence of 2.3% in those with definite HHT. Cerebral MRI was performed in eighty-six percent of our referred population and 89% of those with definite HHT and was deferred only in those of young age or with other contraindications to MRI. While the prevalence of pAVMs in Irish patients is consistent with international series, the prevalence of cAVMs is considerably lower. This could relate to the small size of our study population but may also suggest that Irish patients with HHT may differ genotypically and phenotypically from those in other countries. Another recent study from Germany has suggested a lower than expected prevalence of cAVMs at 5.3% in HHT patients.18 Furthermore, there appears to be a much lower prevalence of both pAVMs and cAVMs in patients with mutations in the ALK-1 gene (HHT2), compared with those with mutations in the endoglin gene (HHT1).19 While genetic screening for known HHT mutations has only been performed in our centre in 2 families to date (both of whom were positive for endoglin/HHT type 1 mutations), this study highlights the importance of prospective genetic screening combined with continued audit to determine if the apparent low prevalence of cAVMs is sporadic or real. In our centre, patients with negative contrast echocardiography do not have further screening for pAVMs, with the exception of children in whom echocardiography is repeated in the late teenage years. Patients with positive contrast echocardiography proceed to unenhanced CT thorax. Where no pAVM is detected, a repeat CT thorax is scheduled three to 5 years later, on the assumption that small pAVMs may exist that are not yet detectable on CT thorax. Where a pAVM is detected, the size of the AVM and in particular the size of the feeding artery will determine management. Where a pAVM has a feeding artery greater than 3 mm in diameter, embolisation is performed and a follow up CT Thorax is scheduled after 6 months to assess involution of the pAVM sac or replacement with a linear scar.2,9,10 Subsequent CT thorax follow up can then occur at three to 5 year intervals. Rapid pAVM enlargement can occur during puberty or pregnancy20

Hereditary haemorrhagic telangiectasia in Ireland and increased frequency of follow up is recommended during these periods.10 Prophylactic antibiotic therapy is recommended prior to any dental or surgical procedure in patients with positive contrast echocardiography or in whom pAVMs cannot be definitely excluded. Such antibiotic therapy is used to reduce risk of bacterial embolisation leading to cerebral abscess.5,11 Thoracic CT detected thirty-two patients with single or multiple pAVMs, all of whom had definite or possible HHT. Nineteen patients with single or multiple pAVMs had 28 embolisation procedures performed, with 1e6 pAVMs embolised per procedure. All patients who required embolisation had definite HHT (three or more Curac ¸ao criteria). Nine patients required repeat embolisation due to persistence of a pAVM on CT thorax after initial embolisation. Persistent pAVMs can result from recanalization of a previously embolised vessel, growth of a missed or previously small accessory artery, or development of a collateral flow around the occlusion. This collateral supply may be derived from a pulmonary artery or from systemic arterial collateral flow beyond the embolisation (for example, from a bronchial artery).2 Complications associated with pAVM embolisation include pleuritic-type chest pain, pulmonary infarction, air embolism, device migration and deep vein thrombosis.9,21 This study had several limitations. Firstly, the numbers of patients screened thus far are relatively small. Assuming a HHT prevalence of 1 in 2500e8000 in an Irish population, there are 500e1600 prevalent cases in Ireland, indicating that only a small number of cases have been diagnosed and screened for visceral AVMs in our centre. Of course, other incident cases will have been diagnosed sporadically in other centres. However, it is clear that the majority of patients with HHT remain undiagnosed in Ireland, as is likely to be the case in other countries, and it may be that a more comprehensive dataset for Ireland would indicate a higher prevalence of cAVMs. Secondly, in the present study not all patients underwent screening for cAVMs (cerebral MRI was deferred in younger patients) and this could cause a bias as cAVMs can be present in younger patients. Thirdly, while patients with a family history of HHT and pAVMs, but without epistaxis and telangiectasia were considered to have suspected HHT (by Curac ¸ao criteria), it is almost certain that such patients have definite HHT, meaning that some patients may have been misclassified. Internationally-published data suggest a prevalence of fifteen to 35% for pAVMs and 10e23% for cAVMs in patients with HHT. While the prevalence of pAVMs in our group is consistent with these data, the prevalence of cAVMs is considerably lower, suggesting that Irish patients with HHT may differ genotypically and phenotypically from those in other countries.

Acknowledgements The authors would like to acknowledge the support of the following: The Grace Nolan Foundation, HHT Foundation International, Professor Bob White, Dr. Marie Faughnan, Ms. Kate Henderson. Dr. John Bourke, Mr. Gerry O’Leary, Mr. John Russell, Dr. Martin Buckley, Dr. Paul Brennan, Professor Andrew Green, Dr. Brian Hennessy, Dr. Claire Buckley.

1223 Competing Interests There are no competing interests with regards to this submission. Funding This submission has not received any funding.

References 1. Guttmacher AE, Marchuk DA, White RI. Hereditary haemorrhagic telangiectasia. N Engl J Med 1995;333: 918e24. 2. Brady AP, Murphy MM, O’ Connor TM. Hereditary haemorrhagic telangiectasia: a cause of preventable morbidity and mortality. Ir J Med Sci, doi:10.1007/s11845-008-0220-5. 3. Shovlin CL, Letarte M. Hereditary haemorrhagic telangiectasia and pulmonary arteriovenous malformations: issues in clinical management and review of pathogenic mechanisms. Thorax 1999;54:714e29. 4. Lerut J, Orlando G, Adam R, Sabba C, Pfitzman R, Klempnauer J, Belghiti J, Pirenne J, Thevenot T, Hillert C, Brown CM, Gonze D, Karam V, Boillot O. Liver transplantation for hereditary haemorrhagic telangiectasia. Ann Surg 2006; 244:854e62. 5. Begbie ME, Wallace GMF, Shovlin CL. Hereditary haemorrhagic telangiectasia Osler-Weber-Rendu syndrome: a view from the 21st century. Postgrad Med J 2003;79:18e24. 6. Shovlin CL, Guttmacher AE, Buscarini E, Faughnan ME, Hyland RH, Westermann CJ, Kjeldsen AD, Plauchu H. Diagnostic criteria for hereditary haemorrhagic telangiectasia (Rendu-Osler-Weber syndrome). Am J Med Genet 2000;91: 66e7. 7. Faughnan ME, Palda VA, Garcia-Tsao G, Geisthoff UW, McDonald J, Proctor DD, Spears J, Brown DH, Buscarini E, Chesnutt MS, Cottin V, Ganguly A, Gossage JR, Guttmacher AE, Hyland RH, Kennedy SJ, Korzenik J, Mager JJ, Ozanne AP, Piccirillo JF, Picus D, Plauchu H, Porteous ME, Pyeritz RE, Ross DA, Sabba C, Swanson K, Terry P, Wallace MC, Westermann CJ, White RI, Young LH, Zarrabeitia R. International guidelines for the diagnosis and management of hereditary hemorrhagic telangiectasia. J Med Genet; 2009 Jun 29 [Epub ahead of print]. 8. Kjeldsen AD, Oxhoj H, Andersen PE, Elle B, Jacobsen JP, Vase P. Pulmonary arteriovenous malformations: screening procedures and pulmonary angiography in patients with hereditary haemorrhagic telangiectasia. Chest 1999;116:432e9. 9. Khurshid I, Downie GH. Pulmonary arteriovenous malformations. Postgrad Med J 2002;78:191e7. 10. Saluja S, Henderson KJ, White RI. Embolotherapy in the bronchial and pulmonary circulations. Radiol Clin North Am 2000; 38:425e48. 11. Pollack JS, Saluja S, Thabet A, Henderson K, Denbow N, White Jr RI. Clinical and anatomic outcomes after embolotherapy of pulmonary arteriovenous malformations (abstract). J Vasc Interv Radiol 2006;17:35e45. 12. Plauchu H, de Chadarevian JP, Bideau A, Robert JM. Agerelated clinical profile of hereditary haemorrhagic telangiectasia in an epidemiologically recruited population. Am J Med Genet 1989;32:197e291. 13. AAssar OS, Friedman CM, White Jr RI. The natural history of epistaxis in hereditary haemorrhagic telangiectasia. Laryngoscope 1991;101:977e80. 14. Fulbright R, Chaloupka JC, Putman CM, Sze GK, Merriam MM, Lee GK, Fayad PB, Awad IA, White Jr RI. MR of hereditary

1224 hemorrhagic telangiectasia; prevalence and spectrum of cerebrovascular malformations. Am J Neuroradiol 1998;19: 477e84. 15. Matsubara S, Manzia JL, ter Brugge K, Willinsky RA, Montanera, Faughnan ME. Angiographic and clinical characteristics of patients with cerebral arteriovenous malformations associated with hereditary hemorrhagic telangiectasia. Am J Neuroradiol 2000;21:1016e20. 16. Post MC, Letteboer TGW, Mager JJ, Plokker TH, Kelder JC, Westermann CJJ. A pulmonary right-to-left shunt in patients with hereditary haemorrhagic telangiectasia is associated with an increased prevalence of migraine. Chest 2005;128: 2485e9. 17. Easey AJ, Wallace GMF, Hughes JMB, Jackson JE, Taylor WJ, Shovlin CL. Should asymptomatic patients with hereditary haemorrhagic telangiectasia (HHT) be screened for cerebral vascular malformations? Data from 22 061

C.F. Ni Bhuachalla et al.

18.

19.

20.

21.

years of HHT patient life. J Neurol Neurosurg Psychiatr 2003;74:743e8. Massmann A, Fries P, Wirth M, Seidel R, Geisthoff UW, Buecker A, Schneider GK. Detection of cerebral arteriovenous malformations with magnetic resonance imaging in patients with hereditary hemorrhagic telangiectasia. Hematol Meet Rep 2009;3:17e8. Letteboer TG, Mager JJ, Snijder RJ, Koeleman BP, Lindhout D, Ploos van Amstel JK, Westermann CJ. Genotypeephenotype relationship in hereditary haemorrhagic telangiectasia. J Med Genet 2006;43:371e7. Ference BA, Shannon TM, White RI, Zawain M, Burdge CM. Life threatening pulmonary haemorrhage with pulmonary arteriovenous malformations and hereditary hemorrhagic telangiectasia. Chest 1994;106:1387e90. White RI, Pollack JS, Wirth JA. Pulmonary arteriovenous malformations: diagnosis and transcatheter embolotherapy. J Vasc Interv Radiol; 1996:787e804.