Insights from fetal cardiac magnetic resonance imaging in double aortic arch

Abstract

Double aortic arch (DAA) is a congenital vascular anomaly identified in approximately 1 in 6640 pregnancies1, in which the ascending aorta divides into two arches that encircle the trachea and esophagus and create the potential for compression of these structures. DAA may present with respiratory insufficiency in infancy and cases that have undergone surgery may have residual respiratory morbidity; therefore, early diagnosis is beneficial2. We present a case suspected to have DAA prenatally and the impact of fetal cardiovascular magnetic resonance imaging (CMR) using motion correction techniques in understanding the evolution of findings3. A woman was referred at 20 weeks' gestation with suspicion of fetal right aortic arch (RAA) for specialized cardiac assessment. Detailed fetal echocardiography (Toshiba Aplio 500 ultrasound system, Canon Medical Systems, Crawley, UK) demonstrated a RAA, left-sided arterial duct and normal intracardiac anatomy. However, the first arterial branch from the RAA was prominent and crossed in front of the trachea, creating a Z-like configuration, which we call the Z-sign, suggestive of a left aortic arch (LAA) (Figure 1). In the standard planes of specialized two-dimensional and color Doppler fetal echocardiography, the first arterial branch could not be traced beyond the origin of the left subclavian artery to assess the presence of left-sided aortic isthmus or a connection to the RAA behind the trachea. Therefore, there was uncertainty as to whether this was a true LAA or simply the brachiocephalic artery. Fetal CMR examination was performed at 31 weeks' gestation using multiple black blood sequences (standard T2-weighted single-shot fast spin echo without intravenous contrast) acquired using a 1.5-Tesla Ingenia magnetic resonance imaging (MRI) system (Philips, Best, The Netherlands). The two-dimensional data were subsequently processed using a motion-correction algorithm to produce a three-dimensional (3D) volume of the fetal thorax3. The corresponding transverse section on fetal CMR replicated the echocardiographic findings showing an arterial branch from the RAA crossing to the left of the trachea (Figure 2a). This appearance is similar to that seen in RAA with mirror image arrangement of the head and neck arteries (Figure 2b). However, alternative views on fetal CMR confirmed the presence of a smaller transverse LAA which gave rise to the left common carotid and left subclavian arteries (Figure 2c,d). The left lateral view, a novel imaging plane that can be visualized on 3D-motion-corrected fetal CMR, confirmed the presence of the distal region of the LAA beyond the left subclavian artery. The left-sided aortic isthmus turned abruptly in the caudal direction and appeared to insert into the arterial duct. Thus, the presence of all segments of a complete LAA was confirmed. The baby was born at term without respiratory difficulty. Given the prenatal suspicion of DAA, contrast computed tomography (CT) and bronchoscopy were performed according to our institutional practice. The CT findings appeared to be consistent with RAA with mirror image branching and showed a closed arterial duct (Figure 3) and significant compression of the airway with > 75% loss of the cross-sectional area of the distal trachea. Following multidisciplinary review of the vascular and airway findings, surgery was recommended. At surgery, presence of DAA with atresia of the aortic isthmus was confirmed (Figure 4). The diagnosis of DAA on prenatal ultrasound is made when the ascending aorta appears to divide into two branches coursing on either side of the trachea. The LAA is often smaller than the RAA, such that visualization of the diminutive LAA is poor and can be overlooked on prenatal imaging4. In order to facilitate visualization of the LAA, a cranial tilt with angulation to open up the area above the left arterial duct from the standard three-vessels-and-trachea view is suggested. In the presented case, the aortic isthmus of the LAA was not visualized on standard two-dimensional fetal echocardiography due to its small size and unexpected position. Fetal CMR allowed visualization of both the right and left arches and identification of the left-sided aortic isthmus with excellent clarity. The aortic isthmus became atretic after birth, but remained as a ligamentous connection. This variant was reported by Edwards in his initial description of aortic arch morphology5 and has been suspected before in the prenatal setting6. This evolution from prenatal patency to acquired atresia of the distal LAA following closure of the arterial duct is identified increasingly in our practice. On postnatal CT angiography, the atretic region of the LAA did not fill with contrast agent since it did not have a lumen, thus, it appeared as a gap between the brachiocephalic artery and the descending aorta. Therefore, this pattern may be misdiagnosed as a RAA with mirror image branching if postnatal imaging is reviewed in isolation. In many pediatric units, the diagnosis of RAA with mirror image branching pattern may not meet the threshold for ongoing monitoring or treatment, which may lead to symptoms of tracheoesophageal compression being overlooked or attributed to laryngomalacia in infancy or asthma in later childhood. MRI is increasingly performed prenatally when extracardiac anomalies are suspected, and recent technical advances, such as motion correction, can extend its role to demonstrate vascular anatomy before birth. As demonstrated in the present case, this technique may permit visualization of vessels which subsequently close after birth and become undetectable on postnatal contrast-enhanced imaging. Fetal MRI is considered a safe examination7 and does not require sedation or intravenous contrast. Currently, most motion-corrected fetal CMR examinations in our center are undertaken in the early third trimester3, but this may change with improvements in technology. In conclusion, fetal CMR is a promising 3D imaging method which can provide additional insights into the cardiovascular anatomy and help establish a definitive diagnosis. When a ‘Z-sign’ is seen on the three-vessels-and-trachea view on fetal echocardiography and fetal CMR is not available, there should be a low threshold for suspecting DAA, even when the left-sided aortic isthmus cannot be visualized on fetal echocardiography. Prenatal imaging provides additional information to postnatal imaging and should be reviewed when considering surgery for a vascular ring. The expectant mother provided informed consent for research participation which included research on fetal cardiac magnetic resonance imaging as part of the Intelligent Fetal Imaging and Diagnosis Project-2 (Research Ethics Committee no:14/LO/1806). This work was supported by the Wellcome/Engineering and Physical Sciences Research Council (EPSRC) Centre for Medical Engineering [203148/Z/16/Z), Wellcome Trust Innovative Engineering for Health Award (102431). The research was funded by the National Institute for Health Research (NIHR) Biomedical Research Centre based at Guy's and St Thomas' NHS Foundation Trust and King's College London and supported by the NIHR Clinical Research Facility (CRF) at Guy's and St Thomas'. The views expressed are those of the author(s) and not necessarily those of the NHS, the NIHR or the Department of Health.

Topics & Concepts