Low-Dose Dual-Source Computed Tomography for Evaluating Persistent Truncus Arteriosus Associated with Cardiovascular Anomalies: Comparison with Transthoracic Echocardiography

Background To assess the morphological features of persistent truncus arteriosus (PTA) on low-dose dual-source computed tomography (DSCT) and compare its diagnostic value for associated cardiovascular anomalies with that of transthoracic echocardiography (TTE). Methods: Twenty-four PTA patients were enrolled in this retrospective study. The types of PTA, diameters of the truncus artery (TA), main pulmonary artery (MPA), right pulmonary artery (RPA), left pulmonary artery (LPA), and ventricular septal defect (VSD) on DSCT were recorded. Besides, all associated cardiovascular abnormalities were assessed. The diagnostic performance of DSCT and TTE for associated anomalies were compared. The effective doses of DSCT were calculated. Results: of VSD on patch diameter of DSCT. clinicians. transfer MPA: LPA


Introduction
Persistent truncus arteriosus (PTA) is a rare congenital cardiac malformation, accounting for 0.7-1.4 % of all congenital heart abnormalities [1]. It is characterized by a single arterial trunk originating from the heart's ventricles and supplying systemic, coronary, and pulmonary circulations 2 . Early surgical intervention is the primary treatment for PTA patients. About 90% of PTA patients would die in the rst year of life without surgery, many during the early infancy 3 . The optimal timing and strategies for surgical repair are based on the morphological features. Thus, the accurate preoperative diagnosis and description of anatomy are crucial. Moreover, the associated anomalies, especially the coronary and pulmonary artery abnormalities, aortopulmonary collateral arteries, are associated with the surgical process and mortality [4][5][6].
Transthoracic echocardiography(TTE)is the rst imaging modality for congenital heart disease. However, its disadvantages, such as operator-dependent essence, small acoustic window, make it inadequate to exhibit the extracardiac structures [7]. Low-dose dualsource computed tomography (DSCT), with low radiation dose, excellent image quality, and powerful post-processing techniques, can evaluate intracardiac and extracardiac malformations. It has been widely applied to cardiovascular examinations in recent years [8][9][10][11][12][13]. To the best of our knowledge, original researches with a certain number of patients focusing on the preoperative evaluation on DSCT were rare [14][15][16]. Therefore, the purpose of this study is to comprehensively assess the morphological features of PTA on lowdose DSCT and compare its diagnostic accuracy for associated cardiovascular anomalies with that of TTE.

Study population
From January 2010 to September 2020, a total of 32 patients with PTA in our hospital were retrospectively enrolled. They were con rmed by surgery or conventional cardiac angiography (CCA). The exclusion criteria included:(a) unavailable data for the DSCT or TTE (n = 5), (b) incomplete clinical data (n = 3). Finally,24 patients remained, including 12 females and 12males (mean age: 9.78 ± 10.12 years, ranging from 1 month to 32 years).
This study was performed in line with the principles of the Declaration of Helsinki. Approval was granted by Biomedical Research Ethics Committee of our hospital, Sichuan University (Chengdu, Sichuan, China; No.14-163), and informed consent was waived off.
The sensitive information of the participants was protected with complete con dentiality and only used for this research.

Dual-source computed tomography
All scans were performed with a DSCT scanner (Somatom De nition; Siemens Medical Solutions, Forchheim, Germany). Patients younger than six years were administered short-term sedative (chloral hydrate at a concentration of 10%, 0.5 ml/kg) before the examination. Older patients were trained to hold their breath during scanning with full cooperation. The scans were performed from the thorax inlet to 2 cm below the diaphragm in the craniocaudal direction. All patients received the nonionic contrast agent (iopamidol, 370 mg/ml, Bracco, Italy) via an antecubital vein, at a rate of 1.2-2.5ml/s, followed by 20 ml of saline solution at the same ow velocity. The total injected volume was determined by body weight (1.5 ml/kg). A retrospective electrocardiography (ECG)-gated protocol was applied. The acquisition parameters were as follows: tube voltage, 80-120 kV; tube current, automatic adjustment; gantry rotation time, 0.28 s and pitch, 0.2-0.5 (adapted to heart rate; higher heart rate used a higher pitch). The ECG-pulsing window was set to auto. Bolus tracking technique was applied in the region of interest (ROI) in the descending aorta, with a prede ned threshold of 100 HU. When the ROI attenuation threshold reached 100 HU, image acquisition was triggered following a delay of 5 seconds. All acquired data were processed on an external workstation (Syngo; Siemens Medical System, Forchheim, Germany). Images were reconstructed with a slice thickness of 0.75 mm and an increment of 0.7 mm. The multiplanar reformation, maximum intensity projection, and volume rendering were used to interpret the images.

Image analysis
The PTA classi cation system suggested by Van Praagh and Van Praagh was used as our standard of classi cation [17]. Type A and B represent the presence and absence of ventricular septal defect, respectively. Type A was classi ed into four types, namely A1, A2, A3, A4. CT images were analyzed blindly by two experienced cardiac radiologists with more than ve years' experience. Types of PTA, origins of the truncus artery (TA), associated intracardiac and extracardiac malformations, the diameter of TA, main pulmonary artery (MPA), right pulmonary artery (RPA), left pulmonary artery (LPA), and ventricular septal defect (VSD) was measured. When the two radiologists disagreed on the ndings, they discussed and reached a consensus. One radiologist's DSCT measurement was reanalyzed by the same radiologist two weeks later to determine the intra-observer variability. To determine the inter-observer variability, the other radiologist repeated the measurement blindly.
The associated cardiovascular malformations contained the intracardiac and extracardiac structures. To compare the diagnostic accuracy of DSCT with TTE, we categorized the con rmed malformations into six groups, namely intracardiac anomalies, anomalies of the aortic arch and its branches, the pulmonary vessels, aortopulmonary vessels, the systemic veins, and coronary artery anomalies.

Assessment of image quality
The overall image quality of DSCT was analyzed by the two radiologists mentioned above, using a four-point scale system. 4 represents excellent (excellent image quality, an incredible display of anatomic details). 3 illustrates good (good image quality, clear display of anatomic details). 2 represents fair (fair image quality, insu cient for a complete evaluation of anatomic details).1 illustrates poor (poor image quality, severe artifacts that make anatomic details incomplete or useless). A score of 3 or 4 was thought to be diagnostic.

Radiation dose estimation
Two radiation dose parameters, volume CT dose index (CTDIvol) and dose-length product (DLP), were automatically recorded during the examinations. The effective dose (ED) was derived from the DLP multiplied by age-speci c DLP conversion coe cients, based on the 2007 recommendation of the International Commission on Radiological Protection [18]. These conversion coe cients were as follows:0.039 for patients aged younger than 4months,0.026 for patients aged between 4 months and 1 year, 0.018 for patients aged between 1 and 6 years, 0.012 for patients aged between 6 and 10 years, and 0.014 for those aged older than 10 years.

Trans-thoracic echocardiography
Standard TTE examinations were performed in all subjects using a Philips Sonos 7500 ultrasound system (Philips Medical Systems, Bothell, WA) based on the American Society of Echocardiography Committee [19]. A skilled echocardiographer with more than ve years' experience blindly interpreted the TTE images.

Statistical analysis
The data were analyzed using SPSS software for Windows (version 24.0; IBM Corp., Armonk, NY, USA). Categorical variables were expressed as numbers and percentages. Continuous variables were presented as means ± standard deviations. The diagnostic performance of DSCT and TTE in associated cardiovascular malformations was expressed as accuracy, sensitivity, speci city, positive predictive value, and negative predictive value. Agreement of image quality was evaluated using the kappa value. Kappa values from 0.61 to 0.80, from 0.81 to 1.00 were considered as good and excellent agreement. The intraclass correlation coe cients (ICC) were calculated to evaluate the inter-observer and intra-observer variability.

Baseline characteristics
The baseline characteristics of the patients were presented in Table 1. The most common clinical signs and symptoms were heart murmurs, post-exercising tachypnea, cyanosis, and easy cold.

Morphological features and measurement on DSCT
Type A1 to A4 were found in our patients, and type B patient was not detected. The most and least common type was type A1 and A4, respectively ( Figs. 1-4). These results were entirely consistent with the results con rmed by surgery or CCA. The average diameter of VSD TA MPA RPA and LPA measured on DSCT was summarized in Table 2.
The summary of ndings obtained by DSCT and TTE was presented in Table 3.    Table 4.

Intra-observer and inter-observer variability
The ICCs of intra-observer variability were 0.91-0.98, and the ICCs of inter-observer variability were 0.87-0.97.

Discussion
PTA is an uncommon congenital cardiac abnormality that occurs due to conotruncal septation failure during fetal development and is correlated with chromosome 22q11 deletion and DiGeorge syndrome [20]. It is always accompanied by VSD and other cardiovascular abnormalities. An early surgical repair is an ideal option in its management. The accurate delineation of morphological characteristics and associated cardiovascular anomalies are essential for the surgical strategies and outcomes. Thus, the exact preoperative diagnosis of these anomalies is imperative.
DSCT, with high temporal and spatial resolution, low radiation, and powerful post-processing techniques, could accurately visualize cardiovascular anomalies. Our study showed 13 patients of typeA1 and only one patient of type A4, which were the most and least common types, respectively. No type B patient was found. These results were entirely consistent with the results con rmed by surgery or CCA. The preoperative measurement of VSD on DSCT was valuable for VSD patch closure must be performed during the operation. The VSDs were always large, for the average diameter in our research was 1.46 ± 0.56cm. Also, the average diameter of TA,MPA RPA and LPA can be measured on DSCT. These data could supply detailed quantitative information for the clinicians.
DSCT was superior to TTE in detecting associated cardiac vascular anomalies. It con rmed more accurately than TTE in all extracardiac groups in the present study. The main reasons could be the small eld of view examined via suprasternal approach, the pneumatic lung, and the short neck of pediatric patients. Additionally, these structures in the neck and chest were shielded partly by the cervical and thoracic bones, which could also be responsible for the unsatisfactory diagnostic performance for TTE. Our data revealed that DSCT was inferior to TTE for the detection of intracardiac anomalies. The results could be partially attributed to the truth that DSCT is an imaging modality requiring a workstation to transfer digital information to gray-scale images [21].
Accurate preoperative diagnosis of coronary artery anomalies is critical. Firstly, during the surgical repair of PTA, the right ventricle's infundibulum needs to be incised before restoration of right ventricle-pulmonary artery continuity [22]. Therefore, any malformed coronary artery that goes across the right ventricle's infundibulum could be accidentally damaged during the right ventriculotomy. Secondly, coronary artery anomalies are signi cant risk factors for death after operation [23]. Our results showed that DSCT could detect this abnormality with 100% sensitivity and 100% speci city. Besides, in the PTA patients, the coronary arteries and pulmonary arteries originate from the truncus artery. It is crucial to identify the ostia number and location of coronary arteries because the coronary ostia or proximal coronary segments could be distorted during the excision of pulmonary arteries from the truncus artery. They could be showed and evaluated accurately by CT, While TTE could miss them [24].
The concomitant pulmonary artery abnormalities such as crossed pulmonary artery and pulmonary artery stenosis are essential in performing the surgery [6,14]. In addition, during PTA surgical procedure, detachment of the pulmonary arteries from the arterial trunk was necessary to form a right ventricle to pulmonary artery conduit. So, a noninvasive assessment of the origins and route of pulmonary arteries was crucial. DSCT could provide detailed and comprehensive information about them through powerful postprocessing techniques. With regards to aortopulmonary collateral arteries, accurate preoperative evaluation of these collateral arteries could be helpful in surgical planning [5,25]. DSCT could easily identify the number, origin, branching pattern, and supplied lung lobes.
However, only large collateral vessels could be demonstrated on TTE [26].
The radiation dose is a signi cant concern undergoing cardiac CT. In our series, several steps were made to reduce radiation dose.
Firstly, tube voltage was one of the crucial factors affecting radiation exposure. The tube voltage was set at a low level of 80 kV in our study, which not only decreased the radiation exposure but also made higher cardiovascular enhancement without the loss of contrastto-noise ratio [27]. Secondly, An ECG-gated sequential technique was used which reduced the radiation dose dramatically. Finally, the heart rhythm adaptive pitch was applied, which adapted higher pitch to t higher heart rhythm. It reduced both radiation exposure and scanning time. In summary, the estimated mean ED was < 1mSv (0.98 ± 0.37mSv), which was under the ALARA (as low as reasonably achievable) principle.
Our study has several limitations. Firstly, because PTA is rare congenital heart disease, our single-center research had a relatively small number of patients. Further multi-center research is needed for larger sample enrollment. Secondly, radiation exposure is inevitable in CT scanning. Thus, we adopted several effective measures to minimize the radiation dose, and the estimated mean ED was 0.98 ± 0.34mSv, which was under the ALARA principle. Finally, as this is a retrospective study, the selection bias is inevitable.

Conclusions
In general, low-dose DSCT could accurately con rm the morphological features of PTA. Furthermore, it has advantages in diagnosing the associated cardiovascular malformations, compared with TTE. Nevertheless, some intracardiac anomalies could be missed by lowdose DSCT. Therefore, it would be bene cial to combine low-dose DSCT with TTE to provide more accurate information for clinical interventions. Availability of data and materials The datasets used during the current study are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests.

Funding
This work was nancially supported by Sichuan Science and Technology Program (2020YJ0229) and 1·3·5 project for disciplines of excellence, West China Hospital, Sichuan University (ZYGD18013).