In our study, ECG-gated MDCT allowed an accurate diagnosis of a PDA. Furthermore, ductal size and morphologic type could be determined according to the angiographic classifications originally described by Miller et al. . In the present study, MPRs and 3-dimensional volume-rendering images were used to demonstrate the presence of PDA. In particular, oblique sagittal MPR images and 3-dimensional volume-rendering images were found to be best suitable for diagnosis of PDA.
Oblique sagittal, dorsal, and transverse MPRs were created for the measurement of minimal ductal size. Consequently, we were able to examine the 3-dimensional morphology of the PDA, which is generally not possible with angiographic or echocardiographic evaluation . However, morphologic assessment by angiocardiography and echocardiography were not performed. Furthermore, 3-dimensional echocardiography might be another imaging modality capable for defining the 3-dimensional anatomy of a PDA. Therefore, further studies are necessary to compare these image modalities with each other and to evaluate whether an MDCT examination may improve the accuracy of measurements and alter the treatment of affected animals.
Saunders et al. carried out transesophageal echocardiographic investigations and described the PDA ampulla in short-axis orientation as an elliptical rather than circular structure . This is in accordance with our findings. Besides the ampulla itself, the measurements of the minimal ductal orifice in all 3 perpendicular views showed small deviations, indicating that it might not always have a perfectly circular appearance.
A disadvantage of MDCT in animals is the necessity of anesthesia. In the case of PDA, affected dogs are potentially compromised, making additional anesthesia for MDCT examination undesirable. However, because MDCT is a fast image acquisition modality, surgical correction or transcatheter closure can be accomplished in the same anesthetic episode. Furthermore, in the case of transcatheter PDA closure, a previous cardiac MDCT evaluation might be suitable to reduce the required number of contrast medium injections and the duration of angiography because size and morphology of PDA can be evaluated before angiography is carried out. This will reduce radiation exposure to the surgeon and staff. Furthermore, previous MDCT examination will decrease angiography time, whereas the overall anesthesia time might increase by a few minutes. Bearing in mind that not every PDA is suitable for transcatheter device occlusion, angiocardiographic examinations of these patients can be avoided by cardiac MDCT examination. This makes a prolonged anesthesia reasonable.
Vascular rings are developmental anomalies of the main thoracic arteries and most commonly cause esophageal constriction. Several types of vascular ring anomalies have been reported in dogs and cats [2, 3]. In this study, we determined different types of vascular ring anomalies in dogs and cats by contrast-enhanced ECG-gated computed tomography and by MDCT with a thoracic standard protocol. A persistent right aortic arch was present in all examined cases, and an aberrant left subclavian artery was most commonly observed. In fact, only 1 cat showed no further coexisting vascular anomaly. An aberrant left subclavian artery in conjunction with a persistent right aortic arch has been previously reported, although this coexistence was less often reported than was observed in the present study . Further commonly reported anomalies coexisting with vascular ring anomalies are PDA and persistent left cranial vena cava. However, none of these anomalies were recognized in this study.
In veterinary medicine, some authors recommend ligation of an aberrant left subclavian artery because it may result in a dorsal constriction as the artery crosses the esophagus . In this study, only ligation and transection of the ligamentum arteriosum were performed because no remaining constriction of the esophagus was observed in any patient after ligamentum division. The fact that an aberrant subclavian artery did not lead to an esophageal compression may be explained by the distal origin of the aberrant subclavian artery in all cases. While the persistent right aortic arch lies at the right of the midline, the proximal part of the descending aorta already approaches the midline. In all cases, the left subclavian artery arose from the left side of the distal part of the aortic arch or the proximal part of the descending aorta. Therefore, its origin was quite dorsal and just beside the midline, leaving enough space for the esophagus after ligamentum division.
One main finding of this study was a diverticulum at the origin of the aberrant left subclavian artery in 2 cats. Similar findings are well described in human medicine and are defined as Kommerell's diverticulum. Kommerell's diverticulum is a dilation of the proximal portion of an aberrant subclavian artery near its origin from the aorta and is expected to be a remnant of the left fourth aortic arch . This diverticulum has been shown to compress the trachea and esophagus independently from the effects of the right aortic arch and left ligamentum arteriosum in children with vascular rings . Furthermore, it can cause recurrent symptoms in human patients if only division of the ligamentum arteriosusm is performed . Therefore, ligation of the aberrant left subclavian artery and resection of Kommerell's diverticulum or resection of Kommerell's diverticulum and reanastomosis of the aberrant left subclavian artery into the left carotid artery is recommended by some authors in human medicine . In addition, 4 dogs showed a mild dilation of the aberrant left subclavian artery at its aortic origin. Analogous conditions were described in a case report of 2 dogs in veterinary medicine . However, to the authors' knowledge, there have been no previous reports of Kommerell's diverticulum in cats. The clinical relevance of Kommerell's diverticulum in dogs and cats remains unclear. Complications comparable with those in human medicine are possible, but not yet investigated. Further examinations of vascular ring anomalies by MDCT might be helpful to identify Kommerell's diverticulum more often and to evaluate its clinical relevance.
In several reported cases, coexisting or rare vascular ring anomalies were not identified by presurgical angiography or even by surgical exploration [24, 33]. This is not unexpected, because the identification of rare or coexisting ring anomalies, such as an aberrant subclavian artery, requires accurate exploration and specific dissection of the region dorsal to the esophagus . Preoperative MDCT angiography can confirm the suspected diagnosis of coexisting vascular anomalies and can help in surgical treatment planning, making only minimal dissection necessary in truly affected animals.
In the case of suspected vascular ring anomaly, 2 different MDCT imaging modalities were performed. Whether a nongated thoracic MDCT protocol or an ECG-gated cardiac protocol was applied depended on the radiologist in charge. However, in 1 case, an ECG-gated MDCT was planned but could not be performed because the heart rate was greater than 130 beats per minute, which exceeded the limitations of ECG-gating in our CT scanner. Therefore, a nongated thoracic standard MDCT protocol was performed. Although coexisting cardiac anomalies are rare, 2 dogs in our study showed a ventricular septal defect. Diagnosis was made by echocardiography, and we were able to demonstrate the connection between the left and right ventricles with ECG-gated MDCT as well (Figure 5). However, the ventricular septal defect was not visible on the images taken by the nongated thoracic standard protocol. Because of the rapid heart movement, motion artifacts caused blurring of the ventricular cavities and interventricular septum and masked the connection between the left and right ventricles. Although ECG-gated images were able to demonstrate the ventricular septal defect, echocardiography appeared to be a more suitable imaging modality for the evaluation of simple lesions of the cardiac septum in our analysis. In particular, color Doppler flow echocardiography allowed the evaluation of cardiac flow across the ventricular septal defects.
Although transverse images are sufficient for the evaluation of vascular ring anomalies, multiplanar and 3-dimensional volume-rendered images facilitate the evaluation of cardiovascular anatomy. In particular, interactive rotation of 3-dimensional volume-rendered images was employed for the optimal understanding of spatial relationships. Although 3-dimensional image modalities are of great value, they remain time-consuming. Therefore, image acquisition and surgery should be performed on different days to avoid considerably prolonged anesthesia. Bearing in mind that vascular ring anomalies do not impair cardiac performance, we consider additional anesthesia for MDCT imaging reasonable.
Former major technical disadvantages of computed tomography, namely single slice scan acquisition and poor temporal resolution, have been greatly reduced by the introduction of MDCT combined with ECG-gating. However, some limitations of computed tomography remain and must be considered. MDCT is less available and more expensive than traditional imaging methods (radiography, angiography, and echocardiography). Furthermore, the ECG-gating technique is restricted to a maximal heart rate. However, the MDCT technique is especially advantageous in the diagnosis of congenital anomalies of the thoracic aorta. The large field of view and MPRs in any desired plane provide good visualization of the complex anatomy of these vascular anomalies and identify associated cardiac malformations. Three-dimensional images by MDCT are suitable to show the anatomy of aortic anomalies (PDA or persistent right aortic arch) and the spatial relationship of adjacent structures and can therefore provide valuable information for surgeons. Even without gating, motion artifacts were not relevant at the level of the aortic arch and its branching, and image quality was judged excellent in all but 1 case. Contrast enhancement of the aorta and great arteries in this cat compared with the other examinations was less, and delineation of the vascular system was harder to achieve. A possible reason for the impaired image quality is the low animal weight of 1.6 kg. However, images of another cat weighing 1.3 kg were acquired by the ECG-gated MDCT protocol and were of excellent image quality. A high dose of contrast medium was used in this study to achieve distinct contrast enhancement of the arterial system and cardiac chambers simultaneously. Therefore, application of the thoracic protocol, which leads to greater motion artifacts than do ECG-gated images, or timing of the contrast medium application is more likely to be of importance.
Limitations of this study include patient variability within a small sample size and the retrospective study design. Another drawback of this study is the lack of comparative data to a "gold-standard," such as catheter-based angiography, for direct comparison in the sizing of the ductus arteriosus.