Imaging in Arteriovenous Lung Malformation
- Author: Sat Sharma, MD, FRCPC; Chief Editor: Eugene C Lin, MD more...
First described in 1897, pulmonary arteriovenous malformation (PAVM) is an abnormal communication between the pulmonary artery and the pulmonary vein. PAVMs are usually congenital in origin; however, they may be acquired in a variety of conditions, such as hepatic cirrhosis, schistosomiasis, mitral stenosis, trauma, actinomycosis, and metastatic thyroid carcinoma.
Patients with PAVM may have hypoxemia and paradoxical embolization complications, incuding stroke and brain abscess.
After the initial description of telangiectasia and epistaxis by Henry Jules Rendu, in 1896, Sir William Osler reported a family known to have hereditary hemorrhagic telangiectasia (HHT). In 1907, Frederick Weber described other manifestations of this disorder; since then, HHT has come to be known as Rendu-Osler-Weber syndrome (or, alternately, as Osler-Weber-Rendu syndrome). Approximately 70% of PAVMs are associated with HHT, and about 15-30% of individuals with HHT have a PAVM. (See images below.)
Diagnostic approach in a patient with a suspected PAVM
A pulmonary arteriovenous malformation may be suspected in the following clinical situations:
An incidental finding of a solitary pulmonary nodule on chest radiographs
The presence of mucocutaneous telangiectases
A situation in which the patient presents with clinical findings of dyspnea, hemoptysis, hypoxemia, polycythemia, clubbing, cyanosis, cerebral embolism, or brain abscess
Whenever a PAVM is suspected, the presence of a right-to-left shunt should be confirmed by the performance of a 100% oxygen study, contrast-enhanced echocardiography, or radionuclide perfusion lung scanning. A definitive diagnosis is established by means of direct imaging of the PAVM with a contrast-enhanced study, such as a computed tomography (CT) scan or a pulmonary angiogram.
Chest radiography reveals some abnormality in most patients. However, further evaluation is needed, with a test to confirm the presence of a right-to-left intrapulmonary shunt and with an imaging study to confirm the presence of PAVM.
Contrast-enhanced echocardiography is extremely sensitive in detecting clinically important PAVMs. If contrast-enhanced echocardiography is not available, radionuclide perfusion lung scanning may be used.
Contrast-enhanced CT scanning remains the criterion standard in the diagnosis of PAVM. Although pulmonary angiography is less sensitive than contrast-enhanced CT scanning, it is performed to accurately define the anatomy, specifically before therapeutic embolization is performed. Although pulmonary angiography may also be a criterion standard for confirmation of a PAVM, angiography is required only when further intervention is planned. Otherwise, in most situations, contrast-enhanced CT scanning is sufficient to confirm the diagnosis.
Although magnetic resonance imaging (MRI) is reportedly useful in the diagnosis of PAVM, it may be less useful than contrast-enhanced CT scanning. Therefore, if the plain chest radiographs suggest a PAVM, contrast-enhanced CT scan remains the preferred examination for confirming its presence.
Limitations of techniques
Chest radiographs may suggest a PAVM, but these images are not helpful in distinguishing between the various causes of a lung nodule. Contrast-enhanced echocardiography may be useful for distinguishing between intracardiac and intrapulmonary shunts, although this identification may be difficult at times.
Contrast-enhanced CT scans may not depict microscopic PAVMs, but the diagnostic yield with spiral CT scanning has been improving. Pulmonary angiography is less sensitive in identifying small or microscopic PAVMs, and MRI has significant limitations in screening for small lesions and in differentiating PAVM from lesions of other causes.
Parra et al compared the results of transthoracic contrast echocardiography (TTCE; divided into 4 grades depending on degree of opacification of the left ventricle after administration of contrast) with the results of thoracic CT scanning in 95 patients with HHT who were examined with both modalities to detect PAVMs. Of the 95 patients, none with normal or grade 1 TTCE had detectable PAVMs on thoracic CT. Grades 2, 3, and 4 were associated with PAVMs according to thoracic CT in 25, 80, and 100% of cases, respectively. The authors therefore concluded that there is a significant association between the TTCE grade and the detection of a PAVM by thoracic CT. They additionally noted that there were significant associations between TTCE grade and the cardiac cycle when the contrast was first visible in the left atrium.
Al-Saleh et al prospectively evaluated the diagnostic yield of a screening protocol for pulmonary and cerebral AVMs in children with a definite or potential diagnosis of HHT. According to the authors, their data showed that children with a definite HHT diagnosis have a high frequency of pulmonary AVMs, even if they are clinically asymptomatic. In contrast, there were no AVMs detected in patients who did not fulfill the diagnostic criteria for HHT. Screening tests for AVMs included chest CT and brain MRI. Of 61 children with a definite clinical or genetic diagnosis of HHT who were asymptomatic for visceral AVMs at first baseline assessment, 15 (25%) had a pulmonary or cerebral AVM on initial screening tests, with pulmonary AVMs predominating in pediatric HHT patients. Of 55 children who fulfilled only 1 or 2 HHT clinical diagnostic criteria, none had pulmonary or cerebral AVMs on initial screening tests.
Chest radiography is recommended for the initial evaluation of patients with heredtary hemorrhagic telangiectasia. Pulmonary arteriovenous malformation may well be an incidental finding on a chest radiograph. A common radiographic finding is a round or oval mass of uniform opacity. The opacity may have sharply defined borders with occasional lobulation. The mass is usually 1-5 cm, and linear shadows are adjacent to the opacity; these are the feeding vessels. PAVMs are commonly present in the lower lobes, and approximately 50% of patients have 2-8 lesions. (See the images below.)
Degree of confidence
When the classic radiographic features are present on the chest radiograph in a patient with suspected PAVM, the diagnosis is certain. However, chest CT scanning is invariably required to confirm the finding.
Patients with microvascular telangiectases may have normal chest radiographic results. Of the 27 patients reported on in one series, the plain chest radiographic findings were strongly suggestive of PAVM in 6 patients, somewhat suggestive in 5 patients, and not suggestive in 6 patients. CT scanning may be more accurate than plain chest radiography in diagnosing PAVM. In a patient who has clinical features suggestive of PAVM but has a normal chest radiograph, further evaluation using other imaging modalities should be undertaken.
Contrast-enhanced CT scanning is the preferred imaging modality for confirming the diagnosis of PAVM. One study reported better sensitivity for pulmonary arteriovenous malformations with ultrafast, contrast-enhanced CT scanning than with pulmonary angiography. CT scans depicted 98% of PAVMs in 20 patients, whereas angiography depicted only 60% of PAVMs.
Three-dimensional (3D), spiral CT scanning produces images of vascular structures that are continuously reconstructed. In one study, spiral CT scanning proved to be a better investigative tool than unilateral pulmonary angiography. The accuracy of 3D spiral CT scanning is reported to be more than 95%, and it may also be useful in identifying smaller PAVMs. (See the CT scans below.)
Multidetector-row CT scanning
The development of multidetector-row CT (MDCT) scanning has resulted in a revolution in imaging capabilities. When compared with single–detector-row CT (SDCT) scanning, CT angiography with MDCT scanning provides superior imaging capabilities. The advantages include an increased per-second scanning area, the ability to retrospectively select the scanning width, the ability to routinely provide 1-mm collimation, and the possibility of higher resolution with no blooming of section thickness.[9, 10, 11, 12]
MDCT scanning also has several practical advantages in the evaluation of PAVM. These include volume imaging, with 3D displays of the volume data, and retrospective reconstruction of the CT scanning data in any plane with high resolution.
Maximum-intensity projection (MIP) and surface rendering were the 3D techniques that were primarily applied to CT angiographic applications in the past. However, volume rendering is becoming the preferred 3D rendering technique for CT angiography.
Surface rendering, the earliest method for 3D display, is commonly available and utilizes each voxel in the data set. The voxel intensity is compared with a threshold value, thereby defining the surface of the object, and the remaining data are discarded. Surface contours are typically modeled with surface shading; therefore, the resulting image is simplified and may provide a misleading representation of the structure. Conversion of data from a volume to a surface sacrifices a large portion of the data and limits the usefulness of surface-rendered medical images.
MIP has been extensively utilized in creating angiographic images from CT scanning and MRI data. Each voxel is evaluated along a line through the image, and the value of the corresponding display pixel is selected based on the value the maximum voxel. The resulting images are typically not displayed with surface shading and increase the background mean of the image, thereby enhancing the structures. The problem with MIP is that volume averaging along with the algorithm may lead to artifacts. Despite its artifacts and deficiencies, MIP has been extensively evaluated and usually provides accuracy superior to that of surface rendering for CT angiography.
This technique renders the entire volume of data rather than just the surfaces, with the total contributions of each voxel being utilized; therefore, more information is displayed than would be provided in a surface model. Volume technique is the most advanced form of 3D rendering currently available for creating accurate, clinically useful medical images and has revolutionized the field of vascular imaging.
It may be difficult to differentiate a PAVM from a vascular tumor on a CT scan. In this situation, a false-positive diagnosis of PAVM is possible. Because prolonged breath holding is required, a large PAVM may be difficult to visualize on 3D spiral CT scans. Furthermore, CT scans may not depict microscopic or small PAVMs.
Magnetic Resonance Imaging
Conventional MRI may be useful in detecting PAVMs. However, compared with other techniques, it has reduced sensitivity and specificity. Because rapidly flowing blood results in an absent or minimal MRI signal, a PAVM may be indistinguishable from normal vascular structures or an air-filled lung.
Various techniques have been used to improve the sensitivity of MRI. These include a rotating, gated MRI technique and gradient recalled-echo MRI. Some investigators have suggested that phase-contrast cine sequences are most accurate in detecting PAVM. Magnetic resonance angiography (MRA) has been used to define the vascular anatomy of a PAVM.
Degree of confidence
In research investigating the role of MRI in the diagnosis of PAVM and examining the modality's false-positive rates, a relatively small number of patients were studied. The sensitivity and specificity of magnetic resonance images are lower than those of CT scans. Furthermore, MRI techniques often yield conflicting results.
Although a combination of MRI techniques may be useful in differentiating PAVMs from other lesions, more studies are required before the routine use of MRI can be recommended.
Contrast-enhanced echocardiography is used to confirm a right-to-left intrapulmonary shunt. Cardiac imaging with 2-dimensional echocardiography is performed while 5-10 mL of agitated sodium chloride solution is injected into a peripheral vein. In a healthy person, microbubbles are rapidly visualized in the right atrium before they gradually dissipate. When an intracardiac shunt is present, the contrast material is visualized in the left heart chambers within 1 cardiac cycle after it appears in the right atrium. The visualization of contrast material in the left atrium after a delay of 3-8 cardiac cycles confirms the presence of an intrapulmonary shunt (ie, PAVM).
Degree of confidence
Contrast-enhanced echocardiography has high sensitivity in detecting right-to-left intrapulmonary shunts, such as PAVMs. This finding should be confirmed by using other tests. The finding of an intrapulmonary shunt during contrast-enhanced echocardiography still warrants further anatomic evaluation with contrast-enhanced CT or pulmonary angiography.
Contrast-enhanced echocardiography is generally not used as a first-line screening test because of its cost, limited availability, and possible overdetection of clinically insignificant PAVMs.
The sensitivity of contrast-enhanced echocardiography approaches 100%; therefore, false-negative test results are unlikely. No specific data address false-positive results; however, contrast-enhanced echocardiography may depict microscopic or clinically insignificant PAVMs.
Radionuclide perfusion lung scanning also is used in the diagnosis of PAVMs, and the findings usually confirm a right-to-left shunt. This test is particularly useful if contrast-enhanced echocardiography or the 100% oxygen method is not available. The test involves the peripheral, intravenous injection of macroaggregated albumin labeled with technetium-99m (99m Tc). In healthy persons, these particles are filtered by pulmonary capillaries. However, in the presence of true right-to-left shunts, radiolabeled particles pass through the lungs and are trapped in the brain and kidneys. The shunt fraction may be calculated by quantifying the renal uptake as a percentage of the total dose given or lung uptake.
Although radionuclide scanning is comparable to the 100% oxygen method in detecting PAVMs, its expense and limited availability prohibit its widespread use. Nonetheless, although the radionuclide method of shunt calculation is not routinely available at most hospitals, it has the following advantages over the 100% oxygen method: (1) radionuclide scanning does not require arterial blood gas sampling, (2) the 100% oxygen method may cause overestimation of the intrapulmonary shunt, and (3) the radionuclide method is more suitable for shunt measurement during exercise.
Radionuclide perfusion scanning depicts only right-to-left shunts and is not useful in determining a shunt's intracardiac or intrapulmonary location.
Contrast-enhanced pulmonary angiography should be used in the treatment of PAVM using embolization or prior to surgical treatment. This test is usually necessary to study the detailed anatomy of PAVM, particularly if the resection or obliterative therapy is planned. Furthermore, angiography may depict other unsuspected PAVMs, as well as unsuspected intrathoracic or extrathoracic vascular communications.
Digital subtraction angiography has replaced conventional angiography in the diagnostic imaging of PAVM. To the authors' knowledge, however, no systematic comparison of the 2 techniques has been reported in the literature. Because nonselective pulmonary angiograms may have difficulty depicting the anatomy of a PAVM, superselective pulmonary angiography is recommended. (See the image below.)
Degree of confidence
Pulmonary angiography has been the criterion standard for defining the anatomy of PAVMs. However, spiral CT scanning is now considered more appropriate for diagnosis.
Pulmonary angiography has high specificity, but it has a sensitivity lower than that of contrast-enhanced ultrafast CT scanning or 3D spiral CT scanning.
Gill SS, Roddie ME, Shovlin CL, Jackson JE. Pulmonary arteriovenous malformations and their mimics. Clin Radiol. 2015 Jan. 70 (1):96-110. [Medline].
Cartin-Ceba R, Swanson KL, Krowka MJ. Pulmonary arteriovenous malformations. Chest. 2013 Sep. 144 (3):1033-44. [Medline].
Kjeldsen AD, Oxhøj H, Andersen PE, et al. Pulmonary arteriovenous malformations: screening procedures and pulmonary angiography in patients with hereditary hemorrhagic telangiectasia. Chest. 1999 Aug. 116(2):432-9. [Medline]. [Full Text].
Lacombe P, Lacout A, Marcy PY, Binsse S, Sellier J, Bensalah M, et al. Diagnosis and treatment of pulmonary arteriovenous malformations in hereditary hemorrhagic telangiectasia: An overview. Diagn Interv Imaging. 2013 Sep. 94 (9):835-48. [Medline].
Vaišnyte B, Vajauskas D, Palionis D, Misonis N, Kurminas M, Nevidomskyte D, et al. Diagnostic methods, treatment modalities, and follow-up of extracranial arteriovenous malformations. Medicina (Kaunas). 2012. 48(8):388-98. [Medline].
Parra JA, Bueno J, Zarauza J, Fariñas-Alvarez MC, Cuesta JM, Ortiz P, et al. Graded contrast echocardiography in pulmonary arteriovenous malformations. Eur Respir J. 2009 Dec 8. [Medline].
Al-Saleh S, Mei-Zahav M, Faughnan ME, MacLusky IB, Carpenter S, Letarte M, et al. Screening for pulmonary and cerebral arteriovenous malformations in children with hereditary haemorrhagic telangiectasia. Eur Respir J. 2009 Oct. 34(4):875-81. [Medline].
White RI Jr, Pollak JS. Pulmonary arteriovenous malformations: diagnosis with three-dimensional helical CT--a breakthrough without contrast media. Radiology. 1994 Jun. 191(3):613-4. [Medline]. [Full Text].
Prokop M. General principles of MDCT. Eur J Radiol. 2003 Mar. 45 Suppl 1:S4-S10. [Medline].
Rubin GD. 3-D imaging with MDCT. Eur J Radiol. 2003 Mar. 45 Suppl 1:S37-41. [Medline].
Rubin GD. Techniques for performing multidetector-row computed tomographic angiography. Tech Vasc Interv Radiol. 2001 Mar. 4(1):2-14. [Medline].
Nawaz A, Litt HI, Stavropoulos SW, Charagundla SR, Shlansky-Goldberg RD, Freiman DB, et al. Digital subtraction pulmonary arteriography versus multidetector CT in the detection of pulmonary arteriovenous malformations. J Vasc Interv Radiol. 2008 Nov. 19(11):1582-8. [Medline].
Zukotynski K, Chan RP, Chow CM, et al. Contrast echocardiography grading predicts pulmonary arteriovenous malformations on CT. Chest. 2007 Jul. 132(1):18-23. [Medline].