Pulmonary arteriovenous malformations: a radiological and clinical investigation of 136 patients with long-term follow-up

Clinical Radiology xxx (2018) 1e7

Contents lists available at ScienceDirect

Clinical Radiology journal homepage: www.clinicalradiologyonline.net

Pulmonary arteriovenous malformations: a radiological and clinical investigation of 136 patients with long-term follow-up P.E. Andersen a, b, *, P.M. Tørring c, S. Duvnjak a, b, O. Gerke d, e, H. Nissen f, A.D. Kjeldsen b, g a

Department of Radiology, Odense University Hospital, Denmark Department of Clinical Research, University of Southern, Denmark c Department of Clinical Genetics, Odense University Hospital, Denmark d Department of Nuclear Medicine, Odense University Hospital, Denmark e Department of Biostatistics, University of Southern, Denmark f Department of Cardiology, Odense University Hospital, Denmark g Department of Otorhinolaryngology Head and Neck Surgery, Odense University Hospital, Denmark b

art icl e i nformat ion Article history: Received 5 April 2018 Accepted 5 July 2018

AIM: To assess the clinical outcome of patients with and without hereditary haemorrhagic telangiectasia (HHT) after embolisation of pulmonary arteriovenous malformations (PAVM) from a single national centre. MATERIALS AND METHODS: The present register-based observational study including all patients with PAVM treated with embolisation at a reference centre for HHT and PAVM was undertaken over a 20-year period. Demographic data, HHT genotyping, clinical presentation, and outcome were registered. Patients with HHT were compared to the patients without HHT. Clinical examination, contrast-enhanced echocardiography, and computed tomography (CT) were used to assess the clinical outcome at follow-up. RESULTS: One hundred and thirty-six patients with 339 PAVM underwent embolisation during the study period: 22 did not have HHT; 62% had HHT1, 10% had HHT2, 4% had JP-HHT, 8% had clinical HHT without identified genetic mutations. Solitary PAVM were more common among patients without HHT than with HHT. Mean follow-up after the first embolisation was 58 months. Mean age at first embolisation was 46.5 years, and at last follow-up 51.8 years. The clinical success without shunt at follow-up was 87%. The 30-day mortality related to the embolisation was 0%. Twenty patients died during follow-up (mean age 69 years). Most patients could be treated during one session, but many will need a long follow-up with repeated clinical examinations and embolisation. CONCLUSION: The majority of patients referred for embolisation of PAVM had HHT. Multiple PAVM is associated with HHT. Patients with PAVM should be screened for HHT and patients with HHT for PAVM. Embolisation is a safe procedure with high clinical success. Ó 2018 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.

* Guarantor and correspondent: P. E. Andersen, Department of Radiology, Odense University Hospital/University of Southern Denmark, J.B. Winsløwsvej 4, DK-5000 Odense C, Denmark. Tel. þ45 65412188; mob.: þ45 21207031; fax þ45 65907244. E-mail address: [email protected] (P.E. Andersen). https://doi.org/10.1016/j.crad.2018.07.096 0009-9260/Ó 2018 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.

Please cite this article in press as: Andersen PE, et al., Pulmonary arteriovenous malformations: a radiological and clinical investigation of 136 patients with long-term follow-up, Clinical Radiology (2018), https://doi.org/10.1016/j.crad.2018.07.096

2

P.E. Andersen et al. / Clinical Radiology xxx (2018) 1e7

Introduction Pulmonary arteriovenous malformations (PAVM) are anastomotic shunts between the pulmonary arteries and veins causing right-to-left shunts. Thus, PAVM may cause paradoxical systemic embolisation to the brain (inducing thrombosis or brain abscess), decreased oxygenation of the blood, or may rarely rupture. Generally, there is an indication to treat these malformations with embolisation to prevent complications.1e10 PAVM are classified as simple, complex, or rarely, diffuse.11,12 Simple PAVM have one supplying artery, whereas complex PAVM have more supplying arteries. PAVM may occur sporadically or in association with hereditary haemorrhagic telangiectasia (HHT).1,13e16 About 90% of patients with PAVM have underlying HHT13,17: however, PAVM can occur in a variety of acquired medical conditions, such as hepatic cirrhosis, chest trauma, actinomycosis or following cardiopulmonary surgery in children. The remainder of PAVM are believed to be idiopathic and tend to be solitary.18 HHT is an autosomal dominant inherited vascular disease with a prevalence of about 15.5/100,000.19 HHT patients may develop mucocutaneous telangiectatic lesions and visceral arteriovenous malformation. Three diseasecausing genes have been identified: ENG (MIM 131195) causing HHT1, ACVRL-1 (MIM 601284) causing HHT2, and SMAD-4 (MIM 600993), which causes the rare JP-HHT syndrome (juvenile polyposis syndrome in combination with HHT).29 Approximately 30% of adult patients with HHT will have PAVM that needs treatment, and some will have AVM in the brain, in the liver, or in the bowel.5,21e23 The aim of the present study was to investigate the clinical outcome after embolisation of PAVM with longterm follow-up of an unselected consecutive patient cohort from a single national centre. Further aims were to provide demographical, epidemiological, clinical, and radiological characterisation of patients with PAVM; to describe the phenotypic differences between PAVM patients with and without HHT; and to disclose the clinical outcome at long-term follow-up with comparison of HHT genotypes and non-HHT patients.

Materials and methods Study design The present study was a single centre retrospective register-based observational study,24 which included all consecutive patients with the intention to treat PAVM with embolisation during the period 22 October 1996 to 31 December 2016. Patients were enrolled prospectively in the study, and a cohort study identifying these patients was extracted from medical and radiological records. There were no exclusion criteria. Patients’ age and gender were recorded. Type, location and number of PAVM, treatment sessions, HHT genotyping, clinical data, contrast-enhanced echocardiography (CE), and radiological examinations (chest radiography, computed tomography [CT] and pulmonary angiography) have been recorded.

Odense University Hospital is the Danish reference centre for HHT diagnostics and PAVM treatments, and almost all Danish patients with diagnosed HHT and/or PAVM are referred to this centre for clinical diagnosis, genetic examination, and management of associated vascular malformations.

Clinical evaluation The clinical examination included evaluation of telangiectatic lesions, family history, medical history regarding epistaxis, gastrointestinal bleeding, and neurological symptoms. The diagnosis of HHT was based on the clinical Curacao criteria, and/or by the presence of a pathogenic variant in one of the known HHT genes.25 HHT genotyping was performed in HHT-suspected patients, as described previously.20,26 CE was performed as described earlier.17,27e30 In an apical four-chamber view, the appearance of microbubbles after 6e8 heart cycles into the left atrium was recorded and graded on an arbitrary scale from 1e4, with grade 1 being few bubbles <10, grade 2 moderate amounts of bubbles, grade 3 large amounts of bubbles but less than on the right side, and grade 4 similar opacification of right and left ventricle.

Interventional procedures Embolisation was performed by two radiologists (PEA and SD) both with >15 years of experience in interventional radiology and both with the European Qualification in Interventional Radiology (EBIR). Embolisation was performed as described previously.6,8,15,31 Antibiotics were given only in selected cases in high-risk patients and heparin only in prolonged procedures. Pulmonary pressure measurements were recorded only in selected cases and not performed routinely. When a PAVM with feeding artery 3 mm or in some cases even smaller31 was detected, embolisation was performed with use of detachable silicone balloons,32 standard, detachable or hydro-coils, or vascular plugs.33e35 Exchange of guidewires and catheters was performed submerged to avoid air embolism to the systemic arteries, especially the coronary arteries. Usually only one lung was treated during one session in cases with multiple bilateral PAVM. There were no absolute contraindications to embolisation. Relative contraindications were high pulmonary arterial pressure, significant hepatic arteriovenous shunt and cardiac failure, pregnancy, renal insufficiency, and previous allergic reactions to contrast media. The study was approved by the Danish Data Protection Agency (file no. 15/10194) according to the Danish Act on Processing of Personal Data (Act no. 429 of 31 May 2000), and Danish Health and Medicines Authority (file no. 33013-974/1).

Outcomes The follow-up plan was established at multidisciplinary team (MDT) meetings. Follow-up was terminated when the

Please cite this article in press as: Andersen PE, et al., Pulmonary arteriovenous malformations: a radiological and clinical investigation of 136 patients with long-term follow-up, Clinical Radiology (2018), https://doi.org/10.1016/j.crad.2018.07.096

P.E. Andersen et al. / Clinical Radiology xxx (2018) 1e7

shunt was grade 0e1 at CE (in 76 patients with mean follow-up 16.9 months; standard deviation [SD] 30.5), and based on clinical findings (history, state of health, functional dyspnoea, physical capacity, oxygenation) and CT evaluation. Further, follow-up was terminated if the patient refused, repeatedly missed planned visits, or died. Continued follow-up was indicated if the patients were not fully grown, became pregnant, for control of growth of small PAVM,36,37 or for other reasons than pulmonary (e.g., nose bleedings or other arteriovenous malformations than PAVM). If the shunt was grade 2 at CE, continued followup including CT of the chest was planned and further embolisation performed or planned, if indicated. The latest patient files were reviewed, and patient status was recorded as follows1: patient still alive, but no longer in follow-up2; patient still alive and still in follow-up3; patient no longer alive. It was recorded if the patients were still part of the follow-up regimen at the HHT centre at the time of death. Data including clinical, CE, and radiological findings were recorded. All radiological images and medical and imaging reports were saved for the present follow-up investigation and after cessation of patient enrolment on 31 December 2016 all patient files were perused. Shunts of grades 0e1 at CE, obviation of surgical intervention for PAVM, and no embolisation-related complications were considered clinically successful outcomes. Complications related to the embolotherapy were recorded and analysed. Complications were defined as permanent damages or disability secondary to the embolisation procedure, need for supplementary anaesthesiological assistance, or prolonged admission to the hospital, and graded according to the DindoeClavien classification.38 Cause of death was registered.

Statistical analysis Quantitative data are expressed as meanSD or mean and range, whereas qualitative data were shown as frequencies and percentages. Comparison of two means was performed using Student’s t-test. Comparisons of frequencies were performed using Fisher’s exact test and twosample test of proportions. All statistical tests were twotailed and the significance level was 5%. Stata/MP 15 (StataCorp LP, College Station, TX, USA) was used for data analysis.

Results Study population

last embolisation was 38.3 and 13 months, respectively (Table 1). In 20 cases the patients were followed-up until death. The 136 patients were treated at 229 sessions for 339 PAVM (Table 1). About half of the patients had one PAVM. In 158/229 (69%) of the treatment sessions, patients were treated for a single PAVM, in 48 (21%) the patients were treated for two PAVM, in 12 (5.2%) for three PAVM, and in 11 (4.8%) for four or more PAVM during a single session. Mean 1.7 (range 1e9) treatment sessions were performed per patient. Eighty-four patients were treated at one session (Table 1) and the remaining 52 patients at 145 sessions, with 34 (25%) at two sessions, eight (5.9%) at three sessions, and 10 (7.4%) at four or more sessions. Ninety-five (70%) of the patients had repeated treatments because of multiple PAVM, PAVM in both lungs, revascularisation, growth of small PAVM, appearance of additional arterial supply, or new PAVM. Four-fifth of PAVM were simple and the remaining complex (Table 2). Two patients had diffuse disease with a vast number of small PAVM and were treated for the biggest PAVM. One hundred (73.5%) of the patients were treated in one lung and the remaining in both lungs. A mean of 2.5 PAVM per patient were treated (Table 2). There were 179 (53%) PAVM located in the right lung and 160 (47%) in the left (p¼0.30). Further, there were more PAVM in the lower lobes than in the upper/middle lobes (63.4% versus 36.6%; p<0.0001).

PAVM characteristics and HHT genotypes One hundred and fourteen (83.8%) of the PAVM cohort had HHT, and 22 did not (Table 3); 74.6% of the HHT cohort had HHT1, 11.4% had HHT2, 4.4% had JP-HHT, and 9.6% had HHTu (clinical HHT based on the Curacao criteria, but without identified mutations in the known HHT-causing genes). There were more PAVMs per patient among HHT patients (315/114 ¼ 2.8 per patient) than among non-HHT patients (24/22 ¼ 1.1 per patient; p<0.0001; Tables 2 and 3). Seventy percent with a single PAVM had HHT and 30% did not, whereas 98% of those with multiple PAVM had HHT Table 1 Characteristics of the pulmonary arteriovenous malformations patient cohort. Characteristics

MeanSD (range) or n (%)

Age at first embolisation (years) Age at last follow-up (years) Follow-up after first embolisation (months)

46.517.9 (11e79) 51.817.9 (15e92) 58.063.6 (0.3e241), median 38.3 35.547 (0.5e217), median 13 70 (51.5%) 28 (20.6%) 14 (10.3%) 24 (17.6%) 84 (61.8%) 52 (38.2%) 136 (100%)

Follow-up after last embolisation (months)

On 31 December 2016, 116 patients, who had undergone embolisation for PAVM, were alive. Thus, the point prevalence of patients with treated PAVM was 116/5,749,000e2/ 100,000. The accumulated all-cause mortality rate was 14.7% (20/136). The PAVM cohort consisted of 136 patients of which 59% were females. Mean age at first embolisation and last follow-up was 46.5 and 51.8 years, respectively. Median follow-up after the first embolisation and after the

3

Patients with 1 PAVM Patients with 2 PAVMs Patients with 3 PAVMs Patients with  4 PAVMs Patients treated at one session Patients treated at > 1 session Total number of patients with PAVM

PAVM, pulmonary arteriovenous malformation; SD, standard deviation.

Please cite this article in press as: Andersen PE, et al., Pulmonary arteriovenous malformations: a radiological and clinical investigation of 136 patients with long-term follow-up, Clinical Radiology (2018), https://doi.org/10.1016/j.crad.2018.07.096

4

P.E. Andersen et al. / Clinical Radiology xxx (2018) 1e7

Table 2 Lesion morphological baseline data from the HHT and non-HHT cohorts. Category

Number of patients PAVMs per patient (mean and range) RUL RML RLL LUL LLL Right lung Left lung Lower lobes Upper- and middle lobes Simple PAVMs Complex PAVMs PAVMs

HHT HHT1

HHT2

JP-HHT

HHTu

Total all HHT types

85 (62.5%) 2.9 24 32 73 31 88 129 119 161 87 202 (59.6%) 46 (13.6%) 248 (73.2%)

13 (9.6%) 1.5 2 8 7 0 3 17 3 10 10 13 (3.8%) 7 (2.1%) 20 (5.9%)

5 (3.7%) 2 0 2 1 1 6 3 7 7 3 8 (2.4%) 2 (0.6%) 10 (2.9%)

11 (8.1%) 3.4 4 5 8 5 15 17 20 23 14 29 (8.6%) 8 (2.4%) 37 (10.9%)

114 (83.8%) 2.8 (1e13) 30 (9.5%) 47 (14.9%) 89 (28.2%) 37 (11.8%) 112 (35.6%) 166 (52.7%) 149 (47.3%) 201 (63.8%) 114 (36.2%) 252 (74.3%) 63 (18.6%) 315 (92.9%)

Non-HHT

Total

22 (16.2%) 1.1 (1e2) 3 (12.5%) 3 (12.5%) 7 (29.2%) 4 (16.6%) 7 (29.2%) 13 (54.2%) 11 (45.8%) 14 (58.3%) 10 (41.7%) 18 (5.3%) 6 (1.8%) 24 (7.1%)

136 (100%) 2.5 (1e13) 33 (9.7%) 50 (14.8%) 96 (28.3%) 41 (12.1%) 119 (35.1%) 179 (52.8%) 160 (47.2%) 215 (63.4%) 124 (36.6%) 270 (79.6%) 69 (20.4%) 339 (100%)

HHT, hereditary haemorrhagic telangiectasia; LLL, left lower lobe; LUL, left upper lobe; PAVM, pulmonary arteriovenous malformation; RLL, right lower lobe; RML, right middle lobe; RUL, right upper lobe; SD, standard deviation.

Table 3 Clinical data for HHT/non-HHT cohorts (frequencies and percentages). Category

HHT1

HHT2

JP-HHT

HHTu

Total all HHT types

Non-HHT

Total

Patients Patients Patients Patients Patients Death

45 40 33 52 85 11 (12.9%)

10 3 10 3 13 2 (15.4%)

4 1 3 2 5 1 (20.0%)

4 7 4 7 11 2 (18.2%)

63 (55.3%) 51 (44.7%) 50 (43.9%) 64 (56.1%) 114 (83.8%) 16 (14.0%)

21 (95.5%) 1 (4.5%) 21 (95.5%) 1 (4.5%) 22 (16.2%) 4 (18.2%)

84 (61.8%) 52 (38.2%) 71 (52.2%) 65 (47.8%) 136 (100%) 20 (14.7%)

treated at one session treated at more than one session with 1 PAVM with >1 PAVM total

HHT, hereditary haemorrhagic telangiectasia; PAVM, pulmonary arteriovenous malformation.

and 2% non-HHT. All but one patients with non-HHT had a single PAVM. Consequently, more patients with non-HHT were treated at one session compared to patients with HHT (46.3% versus 15.4%; p<0.0001). There was a similar distribution of simple and complex PAVM across the HHT types and the non-HHT group, but there were more complex PAVM in HHT1 than in the other HHT types and more in the HHT group than in non-HHT. PAVM were equally observed in the lower lobes and the upper/middle lobes in the HHT group (63.8% resp. 36.2%) compared to the non-HHT group (58.3% resp. 41.7%; Table 2). PAVM were also comparably distributed in the right and left lung in the HHT group (52.7% right 47.3% left) compared to the non-HHT group (54.2% right, 45.8% left). Furthermore, PAVM were diagnosed more often in HHT1 (73.2%) than in HHT2 (5.9%; p<0.0001).

Clinical outcomes Treatment was clinically successful in 118 (86.8%) patients with CE 0e1 (Table 4) with 97 in the HHT group and 21 in the non-HHT group (p¼0.004). Seventeen patients died during follow-up without a clinical shunt at the last hospital visit. Forty-two patients were monitored continuous for growth of small PAVM (14 patients) every 5th year with low-dose CT, or for reasons other than pulmonary shunt. Seven patients waiting for embolisation of PAVM were considered neither clinically successful nor

unsuccessful. Eleven patients (8.1%) were considered clinically unsuccessful at follow-up with no difference between the HHT group and the non-HHT group. Eight were in continuous control with CE >1 or no CE performed, stayed away from hospital visits, or controls were stopped according to the patients’ wishes. Three patients died during follow-up without CE performed or CE>1 (Table 4). There were no significant differences between the HHT and nonHHT groups regarding clinical success at long term followup (Table 4). The 30-day embolisation related mortality was zero, and the cumulated 30-day mortality was 0.7% because a 78-year-old female patient died 4 days after successful embolisation of three big PAVM. She had previously had multiple cerebral emboli and sepsis, and died for reasons not related to the embolisation procedure. Twenty patients (14.7%) died during follow-up (mean age 7010.76 years, range 52e92.5 years). They were similarly distributed between the HHT and non-HHT groups (14% versus 18%). Five patients died of cerebral complications, four died of heart attacks, three died due to malignant diseases, one following drug abuse, and six of old age of otherwise unknown reasons.

Complications Transitory self-limited cardiac arrhythmias were common during catheterisation crossing the right atrium and ventricle with no further intervention necessary. Slight

Please cite this article in press as: Andersen PE, et al., Pulmonary arteriovenous malformations: a radiological and clinical investigation of 136 patients with long-term follow-up, Clinical Radiology (2018), https://doi.org/10.1016/j.crad.2018.07.096

P.E. Andersen et al. / Clinical Radiology xxx (2018) 1e7

5

Table 4 End-points by HHT/non-HHT cohorts (frequencies and percentages). Endpoints

HHT1

HHT2

JP-HHT

HHTu

Total all HHT types

Non-HHT

Total

Terminated (with no clinical shunt and CE 0e1) Dead during follow-up (with no clinical shunt and CE 0e1) Under continuous control (with no clinical shunt and CE 0e1) Successful (with no clinical shunt and CE 0e1) Under control with CE > 1 or without CE performed; stayed away or controls stopped according to patients’ wish Dead during follow-up without CE performed or CE > 1 Insufficient Planned PAVM embolisation Total

31 8 32 77 5

8 2 3 13 0

3 1 1 5 0

1 2 5 9 2

43 (37.7%) 13 (11.4%) 41 (36.0%) 97 (85.1%) 7 (6.2%)

16 (72.7%) 4 (18.2%) 1 (4.5%) 21 (95.5%) 1 (4.5%)

59 (43.4%) 17 (12.5%) 42 (30.9%) 118 (86.8%) 8 (5.9%)

3 8 6 85

0 0 0 13

0 0 0 5

0 2 1 11

3 (2.6%) 10 (8.8%) 7 (6.1%) 114 (100%)

0 (0%) 1 (4.5%) 0 (0%) 22 (100%)

3 (2.2%) 11 (8.1%) 7 (5.1%) 136 (100%)

CE, contrast echocardiography; HHT, hereditary haemorrhagic telangiectasia; PAVM, pulmonary arteriovenous malformation.

elevation of temperature and pleural reaction postembolisation was also common, but needed no other treatment than acetyl salicylic acid. Five patients were under elective anaesthetic surveillance, two because of bradycardia, one (previously treated for cerebral AVM) because of spasms in the extremities, and two because of young age. Three patients had minor and self-limited haemoptysis during or shortly after the embolisation, possibly caused by the guidewires, but without further treatment needed. One patient had a lobectomy 10 months after the embolisation because of persisting chest pain similar to before the embolisation. In one case, a test occlusion balloon was used prior to the embolisation of a big PAVM to evaluate right ventricular performance. Embolotherapy was afterwards performed without complication. In one case, a coil dislodged and stuck in the tricuspid valve. It was not possible to capture the coil with a snare and the patient has been monitored by echocardiography ever since and there have been no complications after 10 years follow-up. Two patients had cerebral abscesses without clinical consequences during or after the treatment course. One patient had a (otherwise uncomplicated) biopsy performed (for a lung mass) before the diagnosis of PAVM was established. Another patient had haemothorax with pleura drainage and two patients had minor self-limited haemoptysis prior to PAVM embolisation. One patient was treated emergent with embolotherapy because of massive haemoptysis and reembolised 13 weeks later because of recurrent major haemoptysis with embolisation of another PAVM with good result. There were no instances of air embolism or paradoxical clot embolisation during the procedures, and no access site thrombosis or persistent haemoptysis after PAVM closure. Thus, according to the DindoeClavien classification,38 only grade IeII complications were observed in this study.

Discussion Most of the demographic and epidemiological data published so far are based on rather small cohorts, short follow-up, and incomplete data, because PAVM is a rare disease with an incidence of about 2e3/100,000.39 Using low-dose CT screening, an incidence of about 38/100,000 was estimated,40 but many of these patients are without

indication for embolisation. In the present study, the cohort consisted of patients with the indication-to-treat PAVM according to clinical symptoms and size of the feeding arteries and the point prevalence has been estimated to be 2/ 100,000. a During the 20-year study period, 136 patients with 339 PAVM were detected. The data are very precise and complete with few missing data. The follow-up was long because the patients were also followed in the HHT centre for reasons other than PAVM. The mean number of PAVM in each patient was 2.5 with most located in the lower lobes, as shown by others41 with a slight predominance for the right lung. Most PAVM in this study were simple (80%) in agreement with 80e95% found by others2,18,42 and most patients had a single PAVM, especially in the non-HHT group. Thus, the majority of patients were treated for a single PAVM at one session. Multiple PAVM were more often found in the HHT group and many patients (42%) needed multiple treatments because of multiple PAVM located in both lungs, retreatments because of re-canalised PAVM, growth of small PAVM, systemic arterial supply to the PAVM, or the development of new PAVM. A significant rate of recanalisation was found, which varied between 5e15%.43e46 Thus, a long follow-up period with clinical and radiological controls with CE and low-dose CT or magnetic resonance imaging (MRI) is often necessary.36,47 Of the patients treated, 84% had HHT, which is similar to previous findings.3,4,13,17,46,48 The present report gives the incidence of HHT in patients with PAVM from the Danish national referral centre for both HHT and PAVM treatment. PAVM are identified in around 50% of patients with HHT1, 13% of the HHT2 patients, and 50% of JP-HHT syndrome patients.20 In the present study, 75% of the patients with HHT (and PAVM) had HHT1. PAVM has been shown to occur comparably often in the whole PAVM cohort (inclusive nonHHT), respectively in the HHT cohort (exclusive non-HHT): HHT 1 (63% versus 75%), HHT2 (10% versus 11%), JP-HHT (4% versus 4%), and HHTu (8% versus 10%; all p¼ns; Table 2). In other vascular malformations, such as cerebral and venous malformations, hereditary cases are characterised by multiple lesions compared to sporadic lesions that are primarily solitary.1,41,49,50 In the present study, it seems also to be the case for PAVM in HHT, and thus the patients without HHT were treated more often with a single session compared with the patients with HHT. The clinical outcome

Please cite this article in press as: Andersen PE, et al., Pulmonary arteriovenous malformations: a radiological and clinical investigation of 136 patients with long-term follow-up, Clinical Radiology (2018), https://doi.org/10.1016/j.crad.2018.07.096

6

P.E. Andersen et al. / Clinical Radiology xxx (2018) 1e7

showed a success of 87% of the patients with shunts of grades 0e1 with no difference between the HHT and the non-HHT groups. Eight percent were deemed unsuccessful with a significant residual shunt (grade >1 at CE) or had dropped-out of the controls. CE is a very sensitive method with a very high negative predictive value, but is not of value in patients with small non-embolised PAVM for follow-up controls. Low-dose CT plays a role in these patients, especially in children, for detection of potential growth of small PAVMs with following indication for embolisation. Control intervals are individualised, depending on age and size of PAVM, of pregnancy and co-morbidity, but in most cases with intervals of 5 years or more. Patients with PAVM should be screened for HHT and patients with HHT for PAVM. Limitations of the study are the inherited possibility for recall-bias and selection bias in the caseecontrol design. A strength of this study is that this it represents a national referral centre for HHT diagnostics and PAVM treatments. Thus, the study includes comprehensive data of many patients with a long follow-up and few drop-outs. In conclusion, the majority of patients referred for embolisation of PAVM have HHT. Most patients can be treated in one session, but repeated controls for a prolonged period and repeated embolisation will often be necessary. Complications related to the embolisation are of low grade, and the clinical success without shunt at long-term follow-up is high.

References 1. Cartin-Ceba R, Swanson KL, Krowka MJ. Pulmonary arteriovenous malformations. Chest 2013;144(3):1033e44. 2. Pugash RA. Pulmonary arteriovenous malformations: overview and transcatheter embolotherapy. Can Assoc Radiol J 2001;52(2):92e102. 3. Mager JJ, Overtoom TT, Blauw H, et al. Embolotherapy of pulmonary arteriovenous malformations: long-term results in 112 patients. J Vasc Interv Radiol 2004;15(5):451e6. 4. Pollak JS, Saluja S, Thabet A, et al. Clinical and anatomical outcomes after embolotherapy of pulmonary arteriovenous malformations. J Vasc Interv Radiol 2006;17(1):35e44. 5. Shovlin C. Pulmonary arteriovenous malformations. Am J Respir Crit Care Med 2014;190(11):1217e28. 6. Andersen PE, Kjeldsen AD. Interventional treatment of pulmonary arteriovenous malformations. World J Radiol 2010;2:339e44. ISSN 19498470. 7. Lacombe P, Lacout A, Marcy P-Y, et al. M. Diagnosis and treatment of pulmonary arteriovenous malformations in hereditary haemorrhagic telangiectasia: an overview. Diagn Intervent Imag 2013;94(9):835e48. 8. Andersen PE, Kjeldsen AD. Clinical and radiological long-term follow-up after embolization of pulmonary arteriovenous malformations. Cardiovasc Intervent Radiol 2006;29(1):70e4. 9. Hsu CC, Kwan GN, Thompson SA, et al. Embolisation for pulmonary arteriovenous malformation. Cochrane Database Syst Rev 2015 Jan 29;1:CD008017, https://doi.org/10.1002/14651858. 10. White Jr RI. Pulmonary arteriovenous malformations: how do I embolize? Tech Vasc Interv Radiol 2007;10(4):283e90. 11. Meek ME, Meek JC, Beheshti MV. Management of pulmonary arteriovenous malformations. Semin Intervent Radiol 2011;28(1):24e31. 12. Pierucci P, Murphy J, Henderson KJ, et al. New definition and natural history of patients with diffuse pulmonary arteriovenous malformations: twenty-seven-year experience. Chest 2008;133(3):653e61. 13. Cottin V, Plauchu H, Bayle J-Y, et al. Pulmonary arteriovenous malformations in patients with hereditary haemorrhagic telangiectasia. Am J Respir Crit Care Med 2004;169(9):994e1000.

14. Chick JFB, Reddy SN, Pyeritz RE, et al. A survey of pulmonary arteriovenous malformation screening, management, and follow-up in hereditary haemorrhagic telangiectasia centers of excellence. Cardiovasc Intervent Radiol 2017;40(7):1003e9. 15. Andersen PE, Kjeldsen AD, Oxhøj H, et al. Embolotherapy for pulmonary arteriovenous malformations in patients with hereditary haemorrhagic telangiectasia (RendueOslereWeber syndrome). Acta Radiol 1998;39(6):723e6. 16. Kjeldsen AD, Oxhøj H, Andersen PE, et al. Pulmonary arteriovenous malformations. Screening procedures and pulmonary angiography in patients with hereditary haemorrhagic telangiectasia. Chest 1999; 116(2):432e9. 17. van Gent MWF, Post MC, Snijder RJ, et al. Real prevalence of pulmonary right-to-left shunt according to genotype in patients with hereditary haemorrhagic telangiectasia: a transthoracic contrast echocardiography study. Chest 2010;138(4):833e9. 18. Wong HH, Chan RP, Klatt R, et al. Idiophatic pulmonary arteriovenous malformations: clinical and imaging characteristics. Eur Resp J 2011;38(2):368e75. 19. Kjeldsen AD, Vase P, Green A. Hereditary haemorrhagic telangiectasia: a population-based study of prevalence and mortality in Danish patients. J Intern Med 1999;245(1):31e9. 20. Tørring PM, Brusgaard K, Ousager LB, Andersen PE, Kjeldsen AD. National mutation study among Danish patients with hereditary haemorrhagic telangiectasia. Clin Genet 2014;86(2):123e33. 21. Lesca G, Olivieri C, Burnichon N, et al. Genotypeephenotype correlations in hereditary hemorrhagic telangiectasia: data from the FrencheItalian HHT network. Genet Med 2007;9(1):14e22. 22. Kjeldsen AD, Kjeldsen J. Gastrointestinal bleeding in patients with hereditary haemorrhagic telangiectasia. Am J Gastroenterol 2000;95(2): 415e8. vian JP, Bideau A, et al. Age-related clinical profile 23. Plauchu H, de Chadare of hereditary haemorrhagic telangiectasia in an epidemiologically recruited population. Am J Med Genet 1989;32(3):291e7. 24. von Elm E, Altman DG, Egger M, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Ann Intern Med 2007;147(8):573e7. 25. Shovlin CL, Guttmacher AE, Buscarini E, et al. Diagnostic criteria for hereditary haemorrhagic telangiectasia (RendueOslereWeber syndrome). Am J Med Genet 2000;91(1):66e7. 26. Kjeldsen AD, Møller TR, Brusgaard K, Vase P, Andersen PE. Clinical symptoms according to genotype among patients with hereditary haemorrhagic telangiectasia. J Intern Med 2005;258(4):349e55. 27. Gazzaniga P, Buscarini E, Leandro G, et al. Contrast echocardiography for pulmonary arteriovenous malformations screening: does any bubble matter? Eur J Echocardiogr 2009;10(4):513e8. 28. Parra JA, Bueno J, Zarauza J, et al. Graded contrast echocardiography in pulmonary arteriovenous malformations. Eur Respir J 2010;35(6):1279e85. 29. Karam C, Sellier J, Mansencal N, et al. Reliability of contrast echocardiography to rule out pulmonary arteriovenous malformations and avoid CT irradiation in pediatric patients with hereditary haemorrhagic telangiectasia. Echocardiography 2015;32(1):42e8. 30. Gossage JR. The role of echocardiography in screening for pulmonary arteriovenous malformations. Chest 2003;123(2):320e2. 31. Andersen PE, Kjeldsen AD. Embolization of pulmonary AVMs of feeding arteries less than 3 mm: reports of two cases and a 8-year follow-up without embolization. Acta Radiol Short Rep 2012;1(2):1e5. https:// doi.org/10.1258/arsr.2012.120001. 32. Andersen PE, Kjeldsen AD. Long-term follow-up after embolization of pulmonary arteriovenous malformations with detachable silicone balloons. Cardiovasc Intervent Radiol 2008;31(3):569e74. 33. Tau N, Atar E, Mei-Zahav M, et al. Amplatzer vascular plugs versus coils for embolization of pulmonary arteriovenous malformations in patients with hereditary haemorrhagic telangiectasia. Cardiovasc Intervent Radiol 2016;39(8):1110e4. 34. Andersen PE, Kjeldsen AD. Occlusion of pulmonary arteriovenous malformations by use of vascular plug. Acta Radiol 2007;48(5):496e9. € 35. Kucukay F, Ozdemir M, Senol E, et al. Large pulmonary arteriovenous malformations: long-term results of embolization with Amplatzer vascular plugs. J Vasc Interv Radiol 2014;25(9):1327e32.

Please cite this article in press as: Andersen PE, et al., Pulmonary arteriovenous malformations: a radiological and clinical investigation of 136 patients with long-term follow-up, Clinical Radiology (2018), https://doi.org/10.1016/j.crad.2018.07.096

P.E. Andersen et al. / Clinical Radiology xxx (2018) 1e7 36. Ryan DJ, O’Connor TM, Murphy MM, et al. Follow-up interval for small untreated pulmonary arteriovenous malformations in hereditary haemorrhagic telangiectasia. Clin Radiol 2017;72(3):236e41. 37. Faughnan ME, Palda VA, Garcia-Tsao G, et al, HHT Foundation InternationaleGuidelines Working Group. International guidelines for the diagnosis and management of hereditary haemorrhagic telangiectasia. J Med Genet 2011;48(2):73e87. 38. Dindo D, Demartines N, Clavien PA. Classification of surgical complications: a new proposal with evaluation in a cohort of 6336 patients and results of a survey. Ann Surg 2004;240(2):205e13. 39. Khurshid I, Downie GH. Pulmonary arteriovenous malformation. Postgrad Med J 2002;78(918):191e7. 40. Nakayama M, Nawa T, Chonan T, et al. Prevalence of pulmonary arteriovenous malformations as estimated by low-dose thoracic CT screening. Intern Med 2012;51(13):1677e81. 41. Lee EY, Boiselle PM, Cleveland RH. Multidetector CT evaluation of congenital lung anomalies. Radiology 2008;247(3):632e48. 42. White Jr RI, Mitchell SE, Barth KH, et al. Angioarchitecture of pulmonary arteriovenous malformations: an important consideration before embolotherapy. AJR Am J Roentgenol 1983;140(4):681e6. langer C, Chartrand-Lefebvre C, Soulez G, et al. Pulmonary arteriovenous 43. Be malformation (PAVM) reperfusion after percutaneous embolization: sensitivity and specificity of non-enhanced CT. Eur J Radiol 2016;85(1):150e7.

7

44. Lee DW, White Jr RI, Egglin TK, et al. Embolotherapy of large pulmonary arteriovenous malformations: long-term results. Ann Thorac Surg 1997;64(4):930e40. 45. Remy-Jardin M, Wattinne L, Remy J. Transcatheter occlusion of pulmonary arterial circulation and collateral supply: failures, incidents, and complications. Radiology 1991;180(3):669e705. 46. Milic A, Chan RP, Cohen JH, et al. Reperfusion of pulmonary arteriovenous malformations after embolotherapy. J Vasc Interv Radiol 2005;16(12):1675e83. 47. Vorselaars VM, Velthuis S, Snijder RJ, et al. Follow-up of pulmonary right-to-left shunt in hereditary haemorrhagic telangiectasia. Eur Respir J 2016;47(6):1750e7. 48. Shovlin CL, Jackson JE, Bamford KB, et al. Primary determinants of ischaemic stroke/brain abscess risks are independent of severity of pulmonary arteriovenous malformations in hereditary haemorrhagic telangiectasia. Thorax 2008;63(3):259e66. 49. Morrison L, Akers A. Cerebral cavernous malformation, familial. In: Adam MP, Ardinger HH, Pagon RA, et al., editors. GeneReviews. Seattle, WA. Seattle: University of Washington; 1993-2018. 50. Boon LM, Vikkula M. Multiple cutaneous and mucosal venous malformations. In: Adam MP, Ardinger HH, Pagon RA, et al., editors. GeneReviews. Seattle, WA. Seattle: University of Washington; 19932018.

Please cite this article in press as: Andersen PE, et al., Pulmonary arteriovenous malformations: a radiological and clinical investigation of 136 patients with long-term follow-up, Clinical Radiology (2018), https://doi.org/10.1016/j.crad.2018.07.096