Comment
In this study, we examined our clinical CT scan database to define arch anomalies, its variants, and the possible association with TAA. We found the reported prevalence of arch anomalies on initial survey to be 2.8%, with bovine aortic arch being the most common anomaly (58.7% of all anomalies and 1.6% in this population), and aberrant left subclavian artery combined with right sided arch to have the highest association with TAA (33%, P<0.001) 12-14. The reported prevalence of 2.8% of arch anomalies was much lower than what we found in the validation sample (21% for bovine arch and 6.6% for aberrant left vertebral artery). The results of the validation sample correspond to previously published reports on the prevalence of bovine arch anomaly. 2,16 This deviation from the literature in the initial survey is due to under reporting in our clinical database, as the CT scans included in this study were done for indications other than aorta diseases (such as lung cancer screening, pulmonary diseases, infection/ sepsis, trauma, pulmonary embolism, esophageal diseases) and it is possible that the arch anatomy received less attention from the reader since most of these cases were report by radiologists without fellowship training in cardiovascular imaging. With this relatively high prevalence of bovine arch, the clinical significance of it should be considered. Morrehead et al reported 31.1% prevalence of bovine arch, including 14.9% type I bovine arch, and 16.2% type II bovine arch in patients with thoracic aortic pathology.15 Dumfarth et al reported bovine aortic arch to be the most common anomalous branch pattern in patients with thoracic aortic disease, with a prevalence of 24.6%. This was followed by isolated left vertebral artery in 6.3% of patients with thoracic aortic disease, and an aberrant right subclavian artery in 1.8%.18 Wanamaker et al., reported similar finding with bovine arch being the most common association with aortic dissection followed by isolated left vertebral artery.10
We also noticed that patients with arch anomalies had higher association with aortic valve disease and hypertension. It is difficult to know if one factor predisposes the occurrence of the other but we can hypothesize that it is also possible that fewer /anomalous branching of the aorta in this region leads to abnormal flow dynamics throughout the aortic wall in this region putting it under abnormal hemodynamic stress which when combined with other factors like hypertension or wall stress caused by abnormal aortic valve geometry could result in an aggressive phenotype leading to aortic dilatation and even dissection due to the added/synergistic effect of multiple wall stressors.19
Given the nearly 3 fold higher association with TAA in the arch anomaly subgroup compared to subjects without arch anomalies in this cohort and also from the forementioned reports, greater attention needs to be paid to carefully look and document for bovine arch anatomy in patients undergoing routine CT chest imaging, with follow-up over time for development of TAA. To some extent, the argument for this is similar to incidental diagnosis of bicuspid aortic valve congenital variant, which occurs in ~1-2% of the general population. These patients are carefully monitored over time for development of aortic valvulopathy and ascending aortopathy, although majority of subjects with bicuspid aortic valve will not need any surgical intervention.
The prevalence of other aortic arch variants noted in our study is in accordance to that reported in the literature, which is ~1-2% 17. Our study also shows differential influence of the type of arch anomaly on the size of the aorta. Right sided aortic arch combined with aberrant left subclavian had the highest association with TAA (33%), while aberrant left vertebral artery was not associated with increased risk of TAA (3%).
In this cohort, we also defined independent risk factors for TAA and found significant association of arch anomaly with incidental aortic aneurysm (OR 2.85). Other independent factors were: aortic valve disease (OR=2.95), age, male sex, and increased BSA, while diabetes was found to be protective (OR=0.75). Given the strong association of arch anomalies with TAA in this study and also in the literature, and the fact that TAA in this sub-population tends to dissect more often4,5, we suggest further investigation for inclusion of aortic arch anomalies in the guidelines for screening for TAA. Current guidelines for screening and follow-up of patients with known thoracic aortic disease does not take into consideration the presence of aortic arch anomalies 20, 21.