Published paper in Radiology (First)


The thoracic duct (TD), as the largest lymphatic vessel, drains approximately three quarters of the body’s lymph into the venous system. In general, the thoracic duct derives from retroperitoneal cisterna chyli, ascends along the spine, and terminates in the venous angle, the region of the junction of the internal jugular vein and the subclavian vein. The cisterna chyli, as a conduit for the lipid products of digestion by receiving fatty chyle from the intestines, results from the convergence of two lumbar lymphatic trunks and the intestinal trunk. The cisterna chyli is a dilated sac located posterior to the abdominal aorta on the anterior aspect of the bodies of the first and second lumbar vertebrae.

In 1692 Nuck first identified the lymphatic system by injecting mercury into the lymphatic vessels, and then further characterized the lymphatic system by injecting colored dye into lymphatic vessels. Lymphangiography emerged in the 20th century, but it was an invasive, radioactive, and time-consuming examination that resulted in many complications. Subsequently, radionuclides were used to trace the lymphatic system, but could be a time-consuming operation and were unable to distinguish the anatomical structures between lymphatic vessels and lymph nodes. Since Hayashi et al. [1] succeeded in applying a 3D half-Fourier fast spin-echo sequence with late-diastolic electrocardiographic gating, in magnetic resonance imaging (MRI), to visualize the TD in 1999, subsequent studies have been demonstrating that MRI is a safe, non-invasive and reliable method for TD imaging.

The TD is filled with slow-moving lymphatic fluid that possesses a long T2 relaxation time, which is the basis of magnetic resonance thoracic ductography (MRTD). So far, sequences of MRTD include heavily T2-weighted turbo spin-echo (TSE) or fast spin-echo (FSE), single shot heavily T2-weighted TSE or FSE, and balanced turbo-field-echo (BTFE). Heavily T2- weighted sequences can enhance the still or slowly streaming fluids in liquid-containing structures and suppress the signal intensity in soft tissues and flowing blood, maximizing the difference in signal intensity between the still or slowly streaming fluids and soft tissues, and flowing blood [2]. TD, as a spindly duct containing slow-moving fluid, is brightly depicted, compared to the dim background, in heavily T2-weighted sequences and maximum intensity projection (MIP) postprocessing reconstruction images. Comparatively speaking, single shot heavily T2-weighted TSE or FSE is superior to heavily T2-weighted TSE or FSE, with the advantages of faster imaging speed, as well as higher signal noise ratio and contrast noise ratio, to conduct a thin slice scan and improve spatial resolution. However, it is difficult to use heavily T2-weighted sequences for evaluating the correlation between TD and surrounding structures, such as vessels and vertebrae, because of the suppression in signal intensity of surrounding tissues. A balanced turbo-field-echo (bTFE) sequence is initially applied to coronary artery magnetic resonance angiography. In 2011, Kato et al. used the bTFE sequence to image the TD. Compared with heavily T2- weighted sequences, bTFE can visualize not only the TD but also the structures surrounding the TD. In addition, bTFE can lessen the scanning time to a large extent, which makes patients more comfortable during the scan. With the advantages of shorter scan time and higher tissue contrast of the surrounding structures, BTFE is superior to single shot heavily T2-weighted TSE or FSE [3-5]. Moreover, three-dimensional (3-D) scanning is superior to two-dimensional (2-D) scanning. This is because 2-D scanning requires a slice thickness of at least 3 mm, in which the motion and respiration of patients are more likely to cause artifacts than 2-D scanning with a thinner slice thickness (less than 1 mm) and a higher spatial resolution. The TD is depicted clearest with respiratory gating in the supine position and at the same time the subject feels most comfortable [6].

The cisterna chyli is identified on MRTD as a dilated sac, of high signal intensity, with a variety of shapes in the retroperitoneum. Shapes of the cisterna chyli found on MRTD include tubular (or linear), plexiform, deltaic, beaded, and triangular, of which the tubular configuration is the most common [7]. Normally, the length of the cisterna chyli is 26.0 ± 10.5 mm, and ranges from 10.2 mm to 55.0 mm. The transverse diameter averages 5.0 ± 2.1 mm, ranging from 2.0 mm to 12.9 mm, and the anteroposterior diameter is 5.1 ± 2.1 mm, ranging from 2.0 mm to 10.9 mm. The TD is identified on MRTD as a tubular structure of high signal intensity ascending along the spine, occasionally with interruption of continuous high signal intensity, due to the spontaneous and rhythmic contraction of the TD. Normally, the transverse diameter of the TD is 3.7 ± 0.4 mm, ranging from 1.9 mm to 6.9 mm, and the anteroposterior diameter of the TD is 3.3 ± 0.4 mm, ranging from 1.4 mm to 6.7 mm [8]. In addition, a fatty meal can enhance lymphatic fluid production and enlarge the TD. Therefore, the TD can often be better delineated at 46 h after the intake of a fatty meal [9,10].

MRTD can be applied to identify TD configuration before the operation and plays a critical role in safer performance of thoracic surgery [11]. Preoperative MRTD is helpful to prevent TD injury, for example, in esophagectomies [12] and posterolateral thoracotomy of mediastinal cystic lymphangioma connected to the TD [13]. In 2009, Okuda et al. studied MRTD in 78 subjects. After excluding subjects with poor MRTD performance, the remaining 73 cases were categorized according to TD configuration. As a result, the results of the MRTD classification were no different from the anatomical classification [11]. In general, the incidence rate of postoperative chylothorax and chyle leakage after conducting esophagectomies is 0.4%2.6%. In 2011, Kato et al. demonstrated that preoperative imaging of TD was useful for preventing postoperative chylothorax and chyle leakage. Kato et al. introduced preoperative imaging of the TD, before conducting esophagectomies in more than 100 subjects, and no postoperative chylothorax or chyle leakage occurred [12]. In 2016, Kim et al. performed MRTD in 3 patients with lymphangioma and found that MRTD was helpful for confirmation of continuity of lymphangioma with the TD in cases of lymphangioma. Kim et al. found that lymphangioma displayed a fluid-filled cystic mass with high signal intensity on T2- weighted images, and low signal intensity on T1-weighted images. In addition, in 2 of the 3 patients, MRTD showed that the lymphangioma had a definite connection with the TD. On the basis of the preoperative MRTD findings, one of the patients underwent posterolateral thoracotomy, and at the same time had the TD ligated during the operation, resulting in resection of the mass without postoperative chylothorax or chyle leakage [13].

Ann Jose
Journal of Imaging & interventional radiology