The ductus arteriosus is a normal fetal anatomic structure that connects the systemic and pulmonary circulations. It usually closes shortly after birth; if it remains patent, the ductus arteriosus may cause heart failure that results from a large shunt into the lungs, the development of pulmonary hypertension, or endocarditis. In its isolated form, patients with patent ductus arteriosus (PDA) are frequently asymptomatic. PDA has been described in combination with virtually every other congenital heart disease, especially those that are characterized by cyanosis, in which this condition may be essential for survival. Patient age at diagnosis can vary from infancy to old age. Imaging results in patent ductus arteriosus are provided below.
The diagnosis is usually made clinically and confirmed echocardiographically, although magnetic resonance imaging (MRI) can also demonstrate a PDA. [1, 2, 3] The preferred imaging method for is 2-dimensional (2-D) echocardiography with color flow Doppler ultrasonographic study.  The primary limitation of echocardiography is the restriction imposed by limited acoustic windows.
In situations in which echocardiography is inadequate (eg, chest deformity, airway disease), magnetic resonance angiography (MRA) is a sensitive technique that can detect ductal flow in the left pulmonary artery, even when the PDA is too small to be visualized.
Although characteristic changes have been identified on chest radiographs, in many patients, chest radiographs are normal.The usefulness of chest radiographs is limited by a lack of specificity and sensitivity. Radiographic features of a shunt are nonspecific. Filling in of the aortopulmonary window is a good sign, but other causes of mediastinal masses or adenopathy can mimic this appearance. Demonstration of this finding also requires good-quality radiography. The most reliable, but least common, finding on chest radiographs is of a calcified PDA in the aortopulmonary window.
Conventional treatment has been with indomethacin or similar prostaglandin synthetase inhibitors in infants or with surgical ligation or transection. In older children and adults, an increasing number of patients are treated using percutaneous techniques. Surgical treatment is very safe in children and adults unless other defects are associated. 
Chest radiographs may demonstrate a large heart, depending on the size of the ductus shunt, with features of pulmonary plethora, heart failure (especially in neonates), or pulmonary hypertension. Filling in of the aortopulmonary window may be seen on good-quality radiographs (as demonstrated in the image below). In elderly patients, a calcified PDA may be demonstrated in this position. In patients with a significant shunt, the ascending aorta and the aortic arch are dilated, and the left atrium and left ventricle are enlarged.
Although a calcified PDA may be visible on lateral chest radiographs, a noncalcified PDA is not profiled and cannot be separated from other vascular structures in the mediastinum.
The degree of confidence for radiographs is moderate. The value of the chest radiograph is limited by a lack of specificity and sensitivity. Findings in a shunt are nonspecific, although filling in of the aortopulmonary window is a good sign. However, this is seldom appreciated prospectively and requires good-quality radiographs.
Other causes of mediastinal masses or adenopathy can mimic this appearance. However, identifying a calcified PDA in the aortopulmonary window is a reliable sign.
False-positive findings may occur as a result of other causes of left-to-right (acyanotic) shunts, including ventricular septal defect, atrial septal defect, endocardial cushion defect (arteriovenous canal), partial anomalous pulmonary venous return, aortopulmonary window, coronary artery fistula, and left ventricle-to-right atrium shunt (Gerbode defect).
Filling in of the aortopulmonary window may also occur as a result of mediastinal masses (eg, lymphoma, thymoma), lymphadenopathy, and mediastinal lipomatosis. In patients with a small patent PDA, chest radiographs may be normal. In rotated images or in patients who did not take a deep breath, the mediastinal contour may be sufficiently distorted that the aortopulmonary window is poorly evaluated.
Although a PDA may be visible on CT angiography, it requires the use of ionizing radiation and usually intravenous contrast agents. Occasionally, calcification of the PDA is demonstrated in a characteristic position. CT angiography may be used to image the aorta for possible aneurysm; this is when ductus arteriosus aneurysm may be detected. 
The degree of confidence is high in detecting complications of PDA, such as ductus arteriosus aneurysm. CT scanning is less applicable for detecting a PDA.
Magnetic Resonance Imaging
The value of MRI is limited in children, in whom echocardiography is almost always adequate to evaluate for PDA. However, in adults with restricted acoustic windows, echocardiography may not be possible. Larger PDA can be seen on spin-echo images, breath-held MRA, or cine MRA (all of which are seen in the images below). The flow disturbance produced by even small PDA in the pulmonary artery is visible as signal loss on cine MR. [7, 8] Flow disturbance is demonstrated best using sagittal cine MR through the distal aortic arch and left pulmonary artery.
Two particular studies have investigated the utility of phase contrast MRI in order to quantify the hemodynamic effect of a PDA. Given that there is ongoing clinical debate regarding the necessity of treatment in all infants with delayed closure of the ductus arteriosus, these 2 small studies suggested that phase contrast MRI may be a more reliable method of evaluating the hemodynamic effect of a PDA, rather than estimation using echocardiography. This may guide clinical decision making — either to determine which infants may be allowed time to close spontaneously or, as in the study by Kozak et al, to evaluate the hemodynamic effect of a PDA in infants with an obstructive left heart lesion, where a PDA is often essential for survival. [9, 10]
Although a PDA can be calcified, spin-echo MRI does not demonstrate calcification in this position. Soft tissue may be seen in the aortopulmonary window. In the first year of life, thymus tissue often obliterates the aortopulmonary window, which makes this an unhelpful sign. Ductus arteriosus aneurysm is a rare finding on echocardiography or any other imaging technique.
Although a PDA must be fairly large to be visualized on MRI or conventional MRA, detection of the flow void caused by turbulent flow passing through the ductus into the pulmonary artery is a sensitive and reliable method for diagnosis. This may not be visualized in patients with pulmonary hypertension in whom the pulmonary artery pressure equals the aortic pressure and no shunt is present.
False-positive and false-negative findings are unlikely with cine MR unless a suitable sequence (ie, one that is sensitive to the flow-dephasing effects of high-velocity turbulent flow) is used. Newer segmented breath-hold sequences are less sensitive to these effects and may not show a signal void due to the turbulent flow from the PDA. In addition, the use of an imaging section that is too thick or the use of an MRI contrast agent may reduce the signal loss due to the PDA.
Using cardiovascular magnetic resonance in 16 infants with PDA (corrected gestational age, 30+3[27+3-36+1] weeks), left ventricular dimension and output were found to be significantly increased, but there was no significant difference in ejection fraction and fractional thickening when compared with controls. There was a significant association between shunt volume and increased left ventricular mass correcting for postnatal age and corrected gestational age.
On M-mode echocardiography, the findings demonstrate normal-sized right heart chambers unless pulmonary hypertension is present. As a result of increased left heart output, the left atrium and left ventricle are dilated, with an increased stroke volume. The same findings are seen on 2-D echocardiography.
In suitable patients, especially the young, the PDA can be visualized directly between the distal arch aorta and pulmonary artery at the origin of the left pulmonary artery (see the image below). A left-to-right shunt is demonstrated using contrast echocardiography.
Continuous-wave or pulsed Doppler echocardiography usually demonstrates continuous flow at the origin of the left pulmonary artery. On color flow imaging, a continuous or diastolic jet of flow from the PDA is demonstrated. Continuous flow on continuous-wave Doppler imaging is a hallmark of PDA. [12, 13]
Echocardiographically directed indomethacin treatment can potentially minimize the number of drug doses needed for PDA closure, according to a study by Carmo et al. The authors investigated whether the duration of indomethacin administration in preterm infants with PDA could be reduced by using echocardiography to determine good treatment response. Following an initial dose of indomethacin in infants born at less than 30 weeks' gestational age, the patients were divided into either a standard treatment group (40 patients; 2 additional doses of indomethacin administered regardless of echocardiographic findings) or an echocardiographically directed group (34 patients; further indomethacin doses provided only if the PDA was >1.6 mm). The authors found that infants in the standard-treatment group received a median of 3 doses, while the echocardiography group received a median of 1 dose. 
Doppler echocardiography is an extremely reliable method for the diagnosis of PDA. Its sensitivity is high, and echocardiography may detect PDA in patients being evaluated for innocent murmurs in whom specific clinical features suggesting the condition are absent.
In one study, early color Doppler PDA diameter greater than 1.5 mm (N = 20) predicted development of symptomatic PDA (sensitivity 91%, specificity 100%), but symptoms resolved spontaneously without treatment in 30%. There was a significant linear correlation (P<0.001) with increasing early PDA diameter and the development of more persistent PDA symptoms and early neonatal mortality and morbidity. 
False-positive findings are rare and may result from misinterpretation of the Doppler signal from pulmonary regurgitation as diastolic flow from a PDA. False-negative findings are rare in children. In adults, false-negative findings may occur as a result of acoustic shadowing that obscures the pulmonary artery, preventing adequate imaging of the area of the PDA or of the main pulmonary artery to detect the characteristic flow abnormalities found on Doppler imaging.
The use of a peripheral intravascular ultrasound catheter in the evaluation of a PDA during percutaneous closure was found to have excellent correlation with CT angiography in the sizing of the PDA. Intravascular ultrasound may be considered in young patients to avoid unnecessary radiation exposure, as well as in patients with substantial renal impairment, allergy to contrast solution, or other contraindications to contrast administration. 
Three-dimensional echocardiography was compared with 2-D echocardiography and angiography in 25 patients with PDA, and there was complete agreement on location, size, morphology and surrounding structure of PDA between 2-D and 3-D echocardiography and angiography. Three-dimensional echocardiography determined type A and type E ductus arteriosus more accurately than 2-D echocardiography. 
In patients with reversal of flow through a PDA, a nuclear medicine shunt study shows early activity in the distal systemic circulation. In practice, this study is of little value, and echocardiography is the preferred examination.
Contrast angiography is seldom required for the diagnosis of PDA, but it may be needed before surgery or during percutaneous interventions. Angiography may also be required to evaluate any coexistent congenital heart lesions or to exclude a differential diagnosis, such as a coronary artery fistula or aortopulmonary window. 
The PDA is profiled best in a steep (60º) left anterior oblique orientation. After contrast material is injected into the descending aorta, the pulmonary artery fills rapidly. A true lateral aortogram may be the best view for sizing devices before percutaneous interventions because this view profiles the PDA well between the aorta and pulmonary artery. (See the images below.)
As demonstrated in the image below, the presence of a PDA can be confirmed with the passage of a catheter from the pulmonary artery (low pressure, low oxygen saturations) through the ductus to the aorta (high pressure, high oxygen saturations) below the diaphragm. This approach allows an aortogram to be obtained without risk of arterial puncture, an especially important consideration in younger patients.
At cardiac catheterization, changes occur in oxygen saturation and in the pulmonary artery and pulmonary artery pressures, which are dependent on the size of the PDA and the differences between pulmonary artery and aortic pressures. For a small PDA, the increase in oxygen saturation from the right ventricle to the pulmonary artery is small. The systolic pulmonary artery pressure is increased, but diastolic pressure remains low unless pulmonary vascular resistance is increased.
In some patients, pulmonary regurgitation may cause an increase in oxygen saturation in the right ventricle, necessitating excluding the diagnosis of a ventricular septal defect. Stretching of a patent foramen ovale by dilation of the left atrium may also cause an increase in oxygen saturation in the right atrium. Coexistent congenital heart lesions should be evaluated at the same time. (See the image below.)
In patients with pulmonary hypertension, pressure measurements should be repeated with administration of 100% oxygen to assess the lability of the increased pulmonary vascular resistance. In patients in whom intervention is planned, test occlusion of the PDA indicates which changes in pulmonary artery pressure may occur and how this may affect other cardiac lesions.
Angiography and cardiac catheterization findings are generally regarded as diagnostically accurate if PDA is under consideration in the differential diagnosis.
Few false-negative findings occur with good angiographic technique. If a double PDA exists, one of the two may be missed if not specifically looked for.