To the best of our knowledge, this is the largest multicenter retrospective cohort study investigating the natural history of DAA, prenatal ultrasound and postnatal CTA characteristics, associated abnormalities, and clinical outcomes. These cases were all thoroughly followed up and had postnatal confirmation. From an anatomical point of view, we report a new branching variant of DAA for the first time. In contrast to the usual pattern wherein each aortic arch gives rise to two branches, in this case, only one branch originated from the left aortic arch, namely, the LCCA, but three branches originated from the right aortic arch, namely, from proximal to distal, the RCCA, RSA, and LSA. However, additional similar cases are needed to confirm this discovery.
DAA is a mostly isolated abnormality but can also be associated with other abnormalities. We found that 31% of cases had DAA with intracardiac malformations, in contrast to 16.6% of cases in another report [7], with ventricular septal defect, double outlet right ventricle, and persistent left superior vena cava being the most common. Extracardiac malformations were rarely mentioned previously [7], but 5 cases (14%) were found in this study, highlighting that a careful and thorough fetal examination outside of the cardiovascular system is also needed. The prevalence of chromosome 22q11.2 microdeletions (DiGeorge syndrome) in fetuses with RAA has been reported to be between 6.1% and 10% [14-16]. Compared to RAA, DAA was associated with a smaller proportion of 22q11 deletion (6%), which is especially rare in isolated DAA, and the two cases of 22q11.2 chromosome deletion in this study were both accompanied by intracardiac malformations and extracardiac malformations (including thymus dysplasia). Ultrasound assessment of dysplasia or the absence of a fetal thymus is useful for predicting 22q11.2 microdeletion [17]. Therefore, for fetuses diagnosed with DAA by ultrasound, thymus size assessment should be performed routinely. In this study, 9 women had terminations of pregnancy, among which 7 were associated with intracardiac or extracardiac malformations or 22q11.2 microdeletion, indicating that whether DAA is associated with other malformations or chromosomal abnormalities has an important impact on pregnancy outcomes and maternal decision-making.
Based on the abovementioned sonographic characteristics of DAA, most DAAs can be diagnosed by prenatal echocardiography. The three-vessel tracheal view is the most characteristic view for DAA diagnosis, but it alone is not enough, especially when an arch shows atresia, in which a vascular ring is not always typical as a result of an interruption in the blood flow of the atretic segment and then differential diagnosis with a right aortic arch is challenging. At this time, multiple views should be performed to better observe the origin, course, and branches of the two arches. The following steps and key points for systematic examination are recommended. (1) Whether the right aortic arch is present on the three-vessel tracheal view should be confirmed. (2) A search for bifurcation of the ascending aorta in the left ventricular outflow tract view is needed; it is necessary to trace the ascending aorta far enough into the arch as the bifurcation is usually not at the origin of the ascending aorta. (3) The three-vessel tracheal view and sagittal view of the aortic arch should be used to confirm that the left and right arches both arise from the ascending aorta and are connected to the descending aorta. (4) The three-vessel trachea view should be used to find the left and right arches surrounding the trachea and esophagus to form a complete O-shaped vascular ring; because the two arches may not be at the same level, the probe will need to be tilted slightly. (5) To determine the type of DAA, the inner diameter of the left and right arches need to be accurately measured. (6) The branching patterns of the left and right aortic arches on the sagittal views and coronal views of the aortic arch need to be confirmed. (7) Careful examination of the heart and other systems, including the thymus, should be performed to determine the existence of associated intracardiac or extracardiac malformations. (8) Chromosome and gene tests should be performed when necessary. Besides, adjusting the parameters of the color Doppler mode is necessary to reduce the scale of blood flow velocity to display a small nondominant arch fully. It is difficult to identify the fibrous cord formed by partial atresia of the nondominant arch by ultrasound examination, so the vascular ring needs to be diagnosed indirectly through the formation of a blind end or diverticulum of the arch. When the left arch is small and difficult to display, the only visible right arch should not be mistaken for the right pulmonary artery, which would also be going in the right direction.
Regarding cases before birth, very few reports have suggested that tracheal compression by the complete vascular ring of the DAA could lead to CHAOS (congenital high airway obstruction syndrome), which can cause intrauterine fetal respiratory distress or stillbirth [10, 18]. However, the sonographic characteristics for predicting perinatal complications of DAA, especially for tracheal compression requiring airway ex utero intrapartum therapy (EXIT), are not clear. In our study, none of the fetuses showed significant airway obstruction before delivery, consistent with most reports.
After birth, some DAA patients have compression due to the vascular ring surrounding the trachea and esophagus, resulting in wheezing, dyspnea, dysphagia and other symptoms, most of which occur within the first year. Currently, it is believed that surgical treatment is necessary for patients with respiratory or digestive symptoms [8, 19]. The proportion of children with symptomatic DAA is reported to be approximately 72.4% [7]. However, during the follow-up after birth in this study, 16 cases (59%) were asymptomatic, and 11 symptomatic cases (41%) underwent surgical treatment.. From our study, DAA seemed to have a more aggressive clinical presentation (41% vs 5.6%-25.2%) and usually required surgical intervention (41% vs 5.6%-17.1%), as opposed to other forms of RAA [2, 14, 20].
Echocardiography is considered to be the first-line postnatal imaging method for DAA. However, rings with special anatomical features, such as a fibrous cord and tracheal compression, cannot be recognized by ultrasound, and pulmonary air easily interferes with image quality. Therefore, MRI or CT is considered to be the gold standard for identifying such variations. However, MRA may require prolonged sedation of pediatric patients [21]. Moreover, image reconstruction, density, and time resolution are worse with MRA than with CT. Compared to MRA, MDCT is a superior imaging modality as it requires less time for a child to calm down and provides more detailed information, including vascular structures and spatial relationships with adjacent organs, especially the airways and esophagus [22]. MDCT combined with various postprocessing options, such as VR, MIP, MinIP, and MR, can display clear details of the compressed trachea, esophagus, and vascular ring, even in cases in which the ring comprises the atretic aortic arch and arterial ligament. Generally, evidence of inferior and posterior convexity of the initial course of the LSA and a descending aortic diverticulum suggests the presence of an imperforate vessel or fibrous cord connecting the structures of the atretic aortic arch [23].
The two cases of misdiagnosis in this study should be noted. Except for the familiar RAA-LPDA-ALSA forming a U-shaped vascular ring, DAA, especially the dominant right arch type, should also be identified with MRAA-LPDA-DAO. When the left arch is small and tortuously curved, it is easily mistaken for the left innominate artery, and the O-shaped vascular ring is not obvious and typical. Repeated multiple cross-sectional examinations show that MRAA-LPDA-DAO fails to demonstrate the connection between the left branch of the aortic arch (the left innominate artery) and the ductus arteriosus or descending aorta. However, in DAA, the left arch is usually connected with the left ductus arteriosus and descending aorta.
Regarding prenatal counseling, we offer some suggestions. First, although DAA was mostly isolated, a certain proportion of intracardiac and extracardiac abnormalities may be associated with it. Therefore, systematic and comprehensive anatomical ultrasonic screening for DAA fetuses is essential. DAA fetuses with intracardiac or extracardiac malformations should be evaluated in detail to assess the severity of the combined malformations and inform the mothers of the prognosis. Secondly, invasive prenatal genetic diagnosis is recommended for DAA fetuses with other intracardiac or extracardiac abnormalities. However, for isolated DAA, genetic testing is not all required and can be discussed with patients. Termination of pregnancy is recommended for fetuses with chromosomal abnormalities (mainly 22q11.2 microdeletion), which often lead to a severe syndrome. Finally, fetal retention is recommended for DAA without severe associated abnormalities and chromosomal abnormalities as the clinical outcomes are favorable, even if surgery is performed due to compression symptoms. However, timely CTA examination for a definite diagnosis and accurate postnatal evaluation is needed.
The main strengths of the study lie first in its large sample size, reliable postnatal confirmation diagnosis and an adequate follow-up duration. In addition, a detailed prenatal ultrasound and postnatal CTA assessment were performed. For the first time, we systematically propose the steps and key points for prenatal ultrasound diagnosis of DAA to improve prenatal diagnosis. And through analyzing the features and advantages of postnatal CTA images, we recommend timely application of CTA examination for accurate postnatal evaluation. Finally, our study provides useful information for prenatal counselling of DAA fetuses.
However, several limitations of this research are worthy of note. First, the incidence of DAA in our study (0.01%), was higher than that in the unselected population according to the literature (0.005%-0.007%) [1, 2]. This discrepancy may be because the research centers in this study were provincial or municipal tertiary referral centers, where many women with high-risk pregnancies or who were suspected of a fetal heart anomaly were referred and selectively examined. Moreover, since we included only fetuses with prenatal diagnoses of DAA, false-negative diagnoses could not be derived and analyzed. Finally, we tried to find useful ultrasonographic manifestations that can predict respiratory symptoms after birth but failed, so further prospective studies are needed.