Aberrant right subclavian artery: the importance of distinguishing between isolated and non-isolated cases in prenatal diagnosis and clinical management

Introduction

In normal circumstances, the aortic arch gives rise to three supra-aortic trunks: the brachiocephalic trunk, which subsequently divides into the right subclavian artery and the right common carotid artery; the left common carotid artery; and the left subclavian artery. When the right subclavian artery arises directly from the aortic arch as an additional artery, distal to the left subclavian artery, it is known as the aberrant right subclavian artery (ARSA). The ARSA courses towards the right arm, crossing the middle line of the body and typically passing behind the trachea and esophagus. This configuration results in the formation of an incomplete C-shaped vascular ring that partially surrounds these structures [1]. While ARSA typically does not cause any apparent clinical symptoms, in rare cases it may exert compression on the aforementioned organs, primarily the esophagus, resulting in dysphagia. Prior publications have also reported respiratory distress and stridor [2], 3]. The causal relationship or coincidence between this symptomatology and the presence of ARSA, in particular postnatally, remains uncertain.

The ARSA is the most common embryological anomaly of the aortic arch occurring in approximately 0.5–1.4 % of the healthy population [4]. However, it can also be part of a complex cardiac malformation, with a reported prevalence of 4.5 % in congenital heart diseases [5], or may be associated with a genetic syndrome [6]. Multiple studies have investigated the prevalence of ARSA in fetuses with Down syndrome, revealing an average prenatal prevalence of 24 % [7]. In addition, ARSA has been reported in individuals with the 22q11 deletion syndrome, typically alongside other cono-truncal cardiac defects, but also in isolation [8], 9].

As a result, ARSA has emerged as a potential prenatal marker for identifying not only cardiac malformations but also genetic anomalies. However, variations in the prevalence of Down syndrome have been observed depending on whether ARSA is detected in isolation or in combination with other anomalies [4], [10], [11], [12]. Consequently, recent studies have indicated reduced predictive efficacy in detecting Down syndrome when ARSA is isolated [7], 11], 13]. Further research is needed to fully understand the significance of ARSA as an isolated finding.

The aim of this study was to evaluate the association of ARSA with genetic abnormalities and postnatal symptomatology by comparing isolated and non-isolated ARSA cases. Thus, facilitating more accurate prenatal counselling and postpartum management in cases of isolated ARSA.

Subjects and methods

This was a retrospective, descriptive and comparative study of all fetuses diagnosed with ARSA in a tertiary referral university hospital in Barcelona from January 2007 to December 2023.

Fetuses presenting with ARSA were identified during routine fetal ultrasound scans, including a mid-trimester scan (19–22 weeks) and third-trimester scans (at 28, 36 and 40 weeks), using color Doppler according to the technique previously described by Chaoui et al. [14]. At the level of the transverse three-vessel and trachea view, ARSA was detected as an additional vessel arising from the junction of the aortic arch and the ductus arteriosus, following its course behind the trachea towards the right arm. Pulsed-wave Doppler can help to differentiate between ARSA and the azygos vein, which also courses to the right of the trachea (Figure 1A and B). Detailed fetal anatomical scanning and echocardiography were systematically conducted in all patients diagnosed with ARSA to identify any potential supplementary abnormalities. The fetal echocardiography was performed according to the international fetal echocardiography guidelines, including the recommended three vessels and trachea view described by Yagel et al. [15] prior 2013, and subsequently following the guidelines established by Carvalho et al. [16]. The cases of ARSA were divided into two groups: isolated ARSA, characterized by ARSA being the only ultrasound finding in either the second or third trimester, and non-isolated ARSA, those with concomitant sonographic findings such as cardiac or extracardiac abnormalities and/or soft markers. Fetal growth restriction (FGR) was included as an extracardiac abnormality. Examinations were conducted transabdominally using high-resolution equipment (Voluson E8 and E10, General Electric Health Care^®, USA) by obstetricians and pediatric cardiologist with expertise in fetal anatomic surveys and echocardiography.

Figure 1:

Ultrasound images of the three vessels and trachea view with color Doppler highlightening the presence of an ARSA. (A) Color Doppler showing the ARSA arising from the junction of the aortic arch and the ductus arteriosus, coursing behind the trachea towards the right arm. (B) Pulsed-wave Doppler over the vessel demonstrating an arterial waveform. PA, pulmonary artery; Ao, aorta; SVC, superior vena cava; T, trachea.

Following the approach of Agathokleous et al. [17], the risk of aneuploidy was recalculated after the visualization of ARSA. Specifically, the initial risk from the first-trimester combined screening was multiplied by the likelihood ratio (LR 3.94) associated with ARSA, resulting in a subsequent classification into low, intermediate, or high risk. Additional soft markers – including intracardiac echogenic focus, ventriculomegaly, increased nuchal fold, echogenic bowel, mild hydronephrosis, short humerus, short femur, and absent or hypoplastic nasal bone – were also considered when present, each carrying specific LRs to further refine the risk assessment. If no soft markers were identified, the initial risk determined from the first-trimester combined screening remained unchanged.

Genetic counselling was provided based on the results. In cases classified as low or intermediate risk and/or featuring isolated ARSA, the recommendation was for non-invasive prenatal test (NIPT) in maternal blood. Conversely, for cases categorized as high risk or non-isolated ARSA, the primary recommendation was an invasive procedure for karyotype analysis and chromosomal microarray analysis (CMA). For patients who declined prenatal genetic testing, a normal karyotype was assumed if neonatal evaluation findings were normal.

All live newborns underwent examination by a pediatrician prior to hospital discharge, following established protocols. Subsequently, ongoing monitoring included at least one assessment by a pediatric cardiologist in an outpatient setting.

For each case, a combination of data was compiled from fetal ultrasound examinations, echocardiograms, non-invasive prenatal test or genetic studies, along with postnatal pediatric records obtained through clinical questionnaires. In cases where clinical questionnaires were unavailable, information was obtained through direct communication with the parents. Cases were excluded if they were unable to complete follow-up or if there was an incomplete dataset. Data were collected using Research Electronic Data Capture software (REDCap). REDCap is a secure web-based software platform designed to support data capture for research studies. Continuous data were presented as mean ± deviation standard, and categorical data were described by percentages and number of cases. For bivariate analysis was used chi squared test or Fisher’s exact test. All tests were two-tailed, and p<0.05 was considered statistically significant. All analyses were performed using IBM^© SPSS Statistics V29. This study received prior approval from COMITÉ ÉTICO DE INVESTIGACIÓN con Medicamentos (CEIm) GRUPO HOSPITALARIO QUIRÓNSALUD-CATALUNYA, Institutional Review Board for Human Investigation and the Ethics Committee (07/02/2023, Reference No. 03/2023).

Results

During the study period, a total of 53,655 pregnancies were attended at our institution. We identified 154 fetuses presenting with ARSA through routine ultrasound examinations, and this diagnosis was subsequently corroborated through fetal echocardiography. At the time of diagnosis, the mean gestational age was 23.8 ± 4.4 weeks, and the mean maternal age was 35.9 ± 4.7 years. Of the 154 fetuses, 75.3 % (116) exhibited the condition as an isolated finding, while the remaining 24.7 % (38) of cases were non-isolated with other structural abnormalities and/or soft markers concomitantly.

In the non-isolated ARSA group, cardiac anomalies were detected in 23.7 % (9 out of 38) of cases, extracardiac malformations in 84.2 % (32 out of 38), and soft markers in 28.9 % (11 out of 38) of cases. Notably, certain fetuses had multiple concurrent abnormalities, such as both cardiac and/or extracardiac anomalies and soft markers in the same individual. The list of sonographic findings observed in fetuses with non-isolated ARSA is shown in Table 1.

Concerning the genetic study, 27.3 % (42) of cases opted for NIPT. Of these, 86 % (36) of cases pertained to the isolated ARSA group, and six cases to the non-isolated ARSA group. All results were low risk, with the exception of a single case in the non-isolated ARSA group, where screening for trisomy 21 was positive and later confirmed postnatally. In 15.6 % (24) of cases invasive procedures were conducted, comprising 22 amniocentesis and two chorionic villus samplings. Among these, 79 % (19) of cases underwent invasive procedures due to belonging to the non-isolated ARSA group, two due to presenting a high risk in the first trimester combined screening, one due to an intermediate risk, and the remaining two were performed at the specific request of the patients. All of them were analyzed for chromosomal alterations. For those with a normal karyotype, testing for 22q11 deletion was performed either by CMA (14 patients) or by fluorescence in situ hybridization (one patient). Eight patients did not undergo further testing. Genetic abnormalities were detected in three fetuses: one presenting trisomy 21 and two with 22q11 deletion. All the affected fetuses had associated structural defects. Details of cases with a genetic abnormality are shown in Tables 2 and 3.

Of the 154 ARSA fetuses, 151 survived after birth, with 116 belonging to the group with isolated ARSA and 35 to the non-isolated ARSA group. One pregnancy was terminated at 29 weeks because of an intracranial hemorrhage in a fetus diagnosed with right unilateral nephromegaly with idiopathic renal thrombosis, and another case was terminated at 22 weeks due to the associated congenital heart disease. One case with IUGR was stillbirth at 33 weeks. The mean gestational age at the time of delivery was 39.0 ± 2.5 weeks, with 68.9 % of deliveries occurring vaginally and 31.1 % by caesarean section. The Apgar score was 9 (range 9–10) at 1 min and 9 (range 8–10) at 5 min, and the average newborn weight was 3,171.4 ± 498.3 g.

Known outcomes were available for all 151 live cases. All neonates underwent postnatal examinations, and since no additional abnormal findings were detected, the karyotype was deemed normal in cases that had not undergone antenatal genetic testing. Symptoms at birth were observed in 8.6 % (3 out of 35) of newborns with non-isolated ARSA. Details of cases with compression symptoms are shown in Table 4. Notably, no cases of isolated ARSA with symptoms at birth were reported.

Of the three neonates with symptoms at birth, only one – diagnosed with tracheoesophageal fistula, esophageal atresia and stenosis – has experienced recurrent bronchitis and dysphagia. This child underwent an uncomplicated Nissen fundoplication and esophageal dilation and is currently 9 years old, showing significant clinical improvement since last year. The other two cases have remained symptom-free since birth and are now 6 years old.

Discussion

This retrospective study aimed to compare isolated and non-isolated ARSA cases, evaluating potential differences in their correlation with genetic abnormalities. Our findings reveal that while no genetic abnormalities were detected in the isolated ARSA group, a notable association between non-isolated ARSA and genetic abnormalities became apparent. This is consistent with prior studies, although the majority were conducted with smaller sample size [4], 8], [11], [12], [13, 18], 19]. A meta-analysis by De León et al. [11] reported an LR+ of 0 for isolated ARSA, and LR+ of 199 for non-isolated ARSA. A more recent meta-analysis by Scala et al. [7] suggested that ARSA could be an independent and clinically useful marker of trisomy 21, but not sufficient to recommend invasive testing in cases with isolated ARSA. Along the same lines, Pico et al. [20] concluded that in patients where the combined risk of chromosomal abnormalities was evaluated, the presence of an isolated ARSA was rarely associated with a chromosomal abnormality.

Regarding the possible association between ARSA and 22q11 deletion syndrome, we identified two fetuses with this genetic condition, both presenting non-isolated ARSA. Previous studies have evaluated the prevalence of 22q11 deletion syndrome in fetuses with ARSA and found that isolated ARSA does not confer an increased risk for pathogenic copy number variations (CNVs) when CMA is performed [8], 21]. Svirsky et al. [8] observed that the incidence of pathogenic CNVs in fetuses with ARSA was similar to that reported in fetuses who underwent CMA without any medical indication, suggesting that ARSA is likely a normal variant or a soft marker rather than a risk factor for pathogenic CNVs. More recently, Xue et al. [13] reported significant differences in the detection rate of clinically significant CMA anomalies when comparing isolated ARSA (2.5 %), ARSA with soft markers (14.3 %), and ARSA with structural anomalies (16.4 %).

Therefore, we support the notion that an isolated ARSA may not be a strong independent indicator of chromosomal abnormalities. However, based on our findings and those of previous studies, we advocate for a meticulous assessment of fetal anatomy, including fetal echocardiography, to detect the presence of other anomalies or soft markers when ARSA is identified. If such additional findings are present, then invasive testing should be performed for fetal microarray analysis. Conversely, isolated ARSA does not inherently increase the risk for genetic disorders, and consideration of NIPT could be warranted in these cases, as proposed by Sagi-Dain et al. [22]. The conflicting evidence in the literature regarding the association of ARSA and chromosomal anomalies may be attributed to some earlier studies [1], 18], 23] that did not differentiate between isolated and non-isolated ARSA. Indeed, in most of the previous studies [14], 17], 18], 24], the populations were composed of high-risk pregnancies with a higher incidence of Down syndrome than the expected incidence.

Among the fetuses in this case series, only three experienced respiratory distress in the postnatal period, all of whom were in the non-isolated ARSA group. One fetus was diagnosed with fetal growth restriction, weighing 2,790 g at birth at 38 weeks; another had tracheoesophageal fistula with esophageal atresia and stenosis; and the third had a right ductal arch forming a complete vascular ring around the trachea. Concerning the potential symptomatology of ARSA, the literature published to date is not elucidating. Patients are commonly asymptomatic in 60–93 % of cases. Given this low frequency of symptoms compared to its incidence, the debate about whether ARSA is merely an incidental finding or if it plays a potential causal role for dysphagia remains unresolved [2], 23]. Secondly, the majority of studies agree that the symptoms typically arise when the strangulating effect becomes critical with age, often due to an aneurysm or arteriosclerosis [2], 25]. However, Puri et al. [26] reported that symptoms could also manifest in childhood. Van Son et al. [27] found that respiratory symptoms were the most commonly associated, attributed to the greater compressibility of the trachea in infants, in contrast to adults whose tracheas are more rigid, with respiratory symptoms being rare. Regarding the neonatal period, one study reported neonatal respiratory symptoms in cases with concomitant right aortic arch and right ductal arch or ALSA, but no cases with ARSA [28]. Another more recent study assessed 50 fetuses with ARSA, none of whom had any complication due to airway compression at birth [29]. In 2023, Li et al. [30] reported that two cases out of 132 with isolated ARSA were born with mild symptoms of dysphagia.

Thus, although it has been hypothesized that tracheoesophageal compression can occur with the presence of ARSA, there is still no consistent data to confirm it. Based on our findings and the existing literature, we attribute the cause of the symptomatology in our cases more to the associated pathology of each individual case rather than to the ARSA itself. Therefore, we agree with Morlando et al. [29] who suggested that, in cases with adequate prenatal assessment performed by experienced clinicians, delivery can confidently occur without the necessity of referral to a tertiary-level hospital. Postnatal assessment and subsequent pediatric outpatient monitoring may be advantageous for those cases with other associated anomalies.

We recognize the inherent limitations of our current study, stemming from its retrospective design and the constraints of a relatively modest sample size. Secondly, despite our initial recommendation, some fetuses with non-isolated ARSA in our series were not tested by CMA. Despite these limitations, this study still has several strengths. It was conducted in a tertiary referral university hospital where all the scans were performed by experienced obstetricians and confirmed by a pediatric cardiologist. Furthermore, every fetus diagnosed with ARSA underwent thorough monitoring through multiple scans in utero, and continued follow-up after birth to validate the accuracy of the antenatal diagnosis and monitor possible symptomatology.

In summary, we suggest that truly isolated ARSA does not confer an increased risk for genetic abnormalities or postnatal compressive symptoms compared to non-isolated ARSA. We advocate that isolated ARSA may be considered a normal vascular variation. While definitive conclusions cannot be drawn from this study due to its modest sample size, we strongly recommend conducting a larger-scale study or a meta-analysis to validate these findings.