Between August 2010 and September 2017, 82 patients with CoA from our hospital were retrospectively reviewed. The inclusion criteria were patients with CoA who were identified by DSCT and underwent TTE before surgery or percutaneous intervention. The exclusion criteria were patients with insufficient clinical information, other aortic diseases such as interrupted aortic arch, aortic dissection, atherosclerosis, arteritis and supravalvular aortic stenosis, malformations like Turner syndrome and Marfan syndrome, or history of cardiovascular catheter or surgical interventions. Finally, 53 patients remained (34 men and 19 women; mean age: 18.15 ± 18.48 years; range: 0.07–84 years). The institutional review board of our hospital approved our current study (No. 14–163) and granted to waive the patient consent based on the retrospective nature of the current study.

DSCT was performed using a related scanner (Somatom Definition; Siemens Medical Solutions, Forchheim, Germany), and a retrospectively ECG-gated protocol was used based on the following parameters: tube voltage of 80–120 kV, tube current of 100–220 mAs, gantry rotation time of 0.28–0.33 s, and pitch of 0.2–0.5. Patients younger than 6 years old were administered with a short-term sedative (chloral hydrate with 10% concentration, 0.5 mL/kg) before the cardiac DSCT examinations. Older patients were instructed to hold their breath when scanning. The scan was conducted in the craniocaudal direction, from the inlet of the thorax to 2 cm below the level of the diaphragm. Nonionic contrast agent (iopamidol, 370 mg/mL, Bracco, Italy) was injected to all patients via an antecubital vein, at a rate of 1.2–2.5 mL/s. All imaging data were analyzed on a workstation (Syngo; Siemens Medical System, Forchheim, Germany). The reconstructed images were based on a slice thickness of 0.75 mm and an increment of 0.7 mm.

All patients received a standard TTE examination based on a ultrasound system (iE33; Philips Medical Systems NA, Bothell, WA, USA). Following the recommendations of the American Society of Echocardiography Committee, the examination was performed with M-mode, two-dimensional, continuous wave and Doppler color flow imaging [9].

The aortic diameters and concomitant cardiac malformations including collateral circulation, bicuspid aortic valve (BAV), patent ductus arteriosus (PDA), ventricular septal defect (VSD), atrial septal defect (ASD), aortic regurgitation (AR) and aortic valve stenosis (AS) were recorded. The axial images, maximum intensity projection and volume rendering were used to analyze images by two experienced radiologists. Aortic diameters were measured at six levels: ascending aorta at its maximum diameter (ascending aorta), aorta just proximal to the origin of the brachiocephalic trunk (pre-coarctation aorta), aortic arch at the largest size (aortic arch), coarctation site at the narrowest size (coarctation site), the widest region of the descending aorta (post-coarctation aorta), descending aorta at the level of the diaphragm (Fig. 1) [10, 11]. The ratio of the aortic diameter at the coarctation site to that at the diaphragm (coarctation site–diaphragm ratio, CDR) was calculated to describe the degree of coarctation. The CoA was identified with a CDR less than 75% [2]. Patients with CoA were classified into two groups based on the severity of coarctation: mild group (CDR > 50%) and severe group (CDR < 50%) [2]. Given the growth-related changes, aortic diameter was standardized as z-score, the ratio of the aortic diameters over the square root of body surface area (BSA) [12]. Aortic dilation was considered if the z-score > 2 (Fig. 2) [12]. Age, BSA, gender, complexity of CoA, presence of BAV, PDA, VSD, AR, AS, history of hypertension (blood pressure > 140/90 mmHg) were assessed as aortic diameter associated factors. Complex CoA was defined as CoA with other cardiovascular abnormalities.

After DSCT examination, volume CT dose index and dose-length product were automatically shown on the CT console. The effective dose was calculated using conversion coefficients referring to the 2007 recommendations of the International Commission on Radiological Protection [13, 14].

This study included 26 (49.1%) mild and 27 (50.9%) severe coarctation patients. A total of 76 cardiovascular complications were confirmed, including 21 collateral circulations (Table 1). The patients in mild group presented smaller BSA (0.85 ± 0.50 vs. 1.26 ± 0.55 m², p < 0.05) and less collateral circulation (23.1% vs. 55.5%, p < 0.05) than those in the severe group (Table 1).

A total of 81 aortic dilations at different levels were developed in 40 (75.5%) patients and the most common site of dilation was the ascending aorta. Regarding the aorta, the z-scores of the ascending aorta and post-coarctation aorta were significantly higher in the severe group than in the mild group (2.10 ± 0.57 vs. 2.41 ± 0.39, and 1.68 ± 0.43 vs. 2.17 ± 0.48, respectively, both p < 0.05; Table 1). The severe group developed more dilation of the ascending aorta (85.2% vs. 53.8%, p < 0.05) and post-coarctation aorta (59.3% vs. 15.4%, p < 0.05) compared with the mild group.

As for ascending aorta, degree of coarctation was associated with the z-score of the ascending aorta (r = − 0.356, p < 0.05). Ascending aortic dilation displayed a negative correlation with degree of coarctation (r = − 0.375, p = 0.006; Additional file 1).

The collateral circulation was associated with z-score of the ascending aorta (r = 0.375, p < 0.05). The occurrence of ascending aortic dilation showed a mild association with age (r = 0.308, p = 0.025) and collateral circulation (r = 0.281, p = 0.042; Additional file 1). Age was related to the z-score of ascending aorta, pre-coarctation aorta and aortic arch(r = 0.37–0.54, all p < 0.05).

Regarding post-coarctation aorta, degree of coarctation was related to the z-score of post-coarctation aorta (r = − 0.414, p < 0.05; Table 2). Degree of coarctation was mildly associated with post-coarctation aortic dilation((r = − 0.354, p < 0.05). In addition, there were mild relations between post-coarctation aortic dilation and age, AR, and VSD (r = 0.345, r = 0.386, and r = − 0.316, respectively, all p < 0.05; Additional file 2).

In univariate analysis, the risk of ascending aortic dilatation was significantly increased by age, severe coarctation, and collateral circulation (Table 3). The presence of post-coarctation aortic dilatation was associated with age, VSD, AR, and severity of coarctation (Table 4). In two multivariate models, increased severity of coarctation was identified as the independent risk factor of ascending aortic dilation (Odds ratio 7.46; 95% CI 1.19–46.76; P < 0.05, Nagelkerke R² = 0.490, Table 3) and post-coarctation dilation (Odds ratio 8.42; 95% CI 1.84–38.56; P < 0.05, Nagelkerke R² = 0.503, Table 4). No statistical association was found between the dilation of all levels of aorta and risk factors including gender, BSA, history of hypertension, BAV, PDA and AS (all p > 0.05).

DSCT was conducted to avoid the excess exposure of radiation. Given that younger patients are sensitive to radiation, we calculated the estimated radiation dose within different age groups (Table 5). The mean effective dose was 1.69 ± 1.48 mSv in patients younger than 14 years old.