Results
In this cohort of 21,336 unique patients aged 50-85 years who underwent CT including the chest (49.3% (n=10,508) were CT scans with intravenous contrast and 2.5% (n=540) were EKG-gated) for any clinical indication, arch anomalies were reported in 2.8% (n=603 patients). Bovine arch was the most common variant (n=354, 58.7% of all anomalies), followed by aberrant right subclavian artery (n=147, 24.4%), aberrant left vertebral artery (n=95, 15.8% ), aberrant left subclavian artery combined with right sided arch (n=12, 2%), double arch (n=1, 0.2%), pseudocoarctation (n=1, 0.2%) (Figure 1). Aberrant left subclavian artery arising from Kommerell’s diverticulum was noted in 5 patients and aberrant right subclavian artery arising from Kommerell’s diverticulum was reported in 3 patients. On bivariate analysis, patients with arch anomalies were more likely to be females (p<0.001), non-Caucasian (p<0.001), hypertensive (p<0.001), diabetic (p=0.012), had hyperlipidemia (p=0.037), and with significantly higher prevalence of aortic valve disease (p<0.001) and TAA (p<0.001). Patients with arch anomalies less commonly had history of smoking and COPD (Table1).
The prevalence of TAA disease was different according to the type of arch anomaly (Figure 3). Subjects with aberrant left subclavian artery combined with right sided arch had the highest prevalence of TAA (33% versus 4.4%; 4 out of 12 patients; P<0.001), followed by bovine arch (13% versus 4.4%,; 46 out of 354 patients with bovine aortic arch; P<0.001), and aberrant right subclavian artery (8% versus 4.4%; 12 out of 147 patients with aberrant right subclavian artery; P<0.001). Aberrant left vertebral artery was not associated with increased prevalence of TAA (3% versus 4.4%; 3 out of 95 patients with aberrant left vertebral artery; P=0.99). The regional distribution of TAA was not different between the arch anomaly group and the no anomaly group. TAA in both groups were more likely to affect the ascending aorta > descending aorta > aortic root > aortic arch (Figure 2). The aortic diameter in the respective region was also not different between the 2 groups. In the region of the root: mean aortic diameter was 4.2 ± 0.3 cm versus 4.3 ± 0.4 cm, P=0.68; in the ascending aorta: 4.3 ± 0.3 cm versus 4.3 ± 0.4 cm, P=0.92; in the arch: 4.6 ± 0.5 cm versus 4.6 ± 0.7 cm, P=0.96; and in the descending aorta: 3.6 ± 0.6 cm versus 3.7 ± 0.8 cm, P=0.60. In the validation sample, 49 cases were previously known to have TAA and were excluded from the analysis. In the remaining 151 scans, 32 patients were found to have bovine arch (prevalence=21%). Type 1 arch was present in 28 cases and type 2 was present in 4 cases. Also, the aberrant left vertebral artery was encountered in 10 patients (6.6%). Other anomalies were not encountered in the validation sample.
To define independent risk factors for TAA and dilatations, a multivariable logistic regression analysis was performed. The model included patient’s age, male sex, Caucasian race, BSA, smoking, hypertension, hyperlipidemia, COPD, aortic valve disease, and arch anomalies. Age (OR = 1.04 CI[1.03-1.05]), male sex (OR = 2.38 [2.01-2.80]), BSA (OR = 1.45 [1.13-1.86]), hypertension (OR = 1.47[1.25-1.73]), aortic valve disease (OR = 2.93 [2.31-3.73]), and arch anomalies (OR = 2.85 [2.16-3.75]) were independent risk factors for TAA (Figure 4).