Single-Center Experience With Dynamic Contrast-Enhanced Magnetic Resonance Lymphangiography for Diagnosing Lymphatic Disorders and Guiding Percutaneous Embolization

Abstract

Background: The pragmatic role of dynamic contrast-enhanced magnetic resonance lymphangiography (DCMRL) needs to be evaluated and compared across distinct lymphatic disorders. We aimed to evaluate the performance of DCMRL for identifying the underlying causes of lymphatic disorders and to define the potential benefit of DCMRL for planning lymphatic interventions.

Methods: Patients who underwent DCMRL between August 2017 and July 2022 were included in this retrospective analysis. DCMRL was performed with intranodal injection of a gadolinium-based contrast medium through inguinal lymph nodes under local anesthesia. Technical success of DCMRL and feasibility of percutaneous embolization were assessed based on the lymphatic anatomy visualized by DCMRL. Based on the underlying causes, clinical outcomes were evaluated and compared.

Results: Seventy consecutive patients were included. The indications were traumatic chylothorax (n = 42), traumatic chylous ascites (n = 11), and nontraumatic lymphatic leak (n = 17). The technical success rate of DCMRL was the highest in association with nontraumatic lymphatic disorders (94.1% [16/17]), followed by traumatic chylothorax (92.9% [39/42]) and traumatic chylous ascites (81.8% [9/11]). Thirty-one (47.7%) patients among 65 patients who underwent technically successful DCMRL had feasible anatomy for intervention. Clinical success was achieved in 90.3% (28/31) of patients with feasible anatomy for radiologic intervention, while 62.5% (10/16) of patients with anatomical challenges showed improvement. Most patients with traumatic chylothorax showed improvement (92.9% [39/42]), whereas only 23.5% (4/17) of patients with nontraumatic lymphatic disorders showed clinical improvement.

Conclusion: DCMRL can help identify the underlying causes of lymphatic disorders. The performance of DCMRL and clinical outcomes vary based on the underlying cause. The feasibility of lymphatic intervention can be determined using DCMRL, which can help in predicting clinical outcomes.

Keywords: Chylothorax; Lymphangiography; Magnetic Resonance Imaging; Thoracic Duct.

Fig. 1. Flow chart of patient categorizations by underlying causes, treatment strategies, and clinical outcomes based on DCMRL results.

DCMRL = dynamic contrast-enhanced magnetic resonance lymphangiography.

Fig. 2. A 58-year-old man developed bilateral chylothorax 4 days after an Ivor-Lewis operation for esophageal cancer. The chylothorax, refractory to diet modulation, produced a drainage of 1,100 mL/day. (A) Chest radiograph on the same day showed left pleural effusion (asterisk). (B) DCMRL on postoperative day 12: coronal 3-dimensional T1-weighted image with thin minimal intensity projection reconstruction 8 minutes after intranodal contrast injection revealed retroperitoneal lymphatics and cisterna chyli (arrowhead) at the T12 vertebral level. (C) Contrast leakage from the thoracic duct at the T8 vertebral level (arrow) 7 minutes post-injection. (D) Pre-intervention DCMRL indicated feasible anatomy for radiologic intervention with an accessible central lymphatic route; thoracic duct embolization was performed 2 days later via direct cisterna chyli puncture (arrowhead). The leakage, consistent with DCMRL findings, was located at T8/9 (arrow) and managed with one microcoil and N-butyl cyanoacrylate. (E) Post-embolization plain abdominal radiograph showing embolic materials (arrowheads) along retroperitoneal lymphatics and the thoracic duct. (F) Chest radiograph 3 days post-embolization showed reduced left pleural effusion and persistent embolic materials along the thoracic duct (arrowheads).

DCMRL = dynamic contrast-enhanced magnetic resonance lymphangiography.

Fig. 3. A 73-year-old man presented with refractory chylous ascites 3 months after right nephrectomy for renal cell carcinoma. (A) Abdominal computed tomography illustrated perihepatic ascites (asterisk) with pneumoperitoneum due to therapeutic thoracentesis. (B) DCMRL performed after 2.5 postoperative years: coronal 3-dimensional T1-weighted images at 4, 6, 7, 10, and 16 minutes after intranodal contrast injection showed progressive contrast leakage (arrows) at the right inferior vena cava at L3. The axial image at 16 minutes post-injection showed contrast media leakage (white arrow) from retroperitoneal lymphatics. (C) Owing to the complex retroperitoneal lymphatics, the anatomy was considered technically unsuitable for intervention. The following day, conventional lymphangiography hinted at a faint leak (arrow), prompting direct puncture and N-butyl cyanoacrylate embolization. (D) One day post-intervention, plain abdominal radiography revealed ethiodized oil along retroperitoneal lymphatics and the leakage site. Despite this and three subsequent embolizations, the ascites did not resolve.

DCMRL = dynamic contrast-enhanced magnetic resonance lymphangiography.

Fig. 4. A 34-year-old woman diagnosed with lymphangioleiomyomatosis. (A) Chest CT displayed numerous small cysts throughout both lungs, consistent with the underlying diagnosis. (B) CT scan 2 weeks prior to DCMRL revealed a newly developed right pleural effusion (white asterisk), confirmed as chylothorax by thoracic fluid analysis. (C) DCMRL was conducted with coronal 3-dimensional T1-weighted images using thin, minimal intensity projection reconstruction, showing retroperitoneal lymphatics (white arrowheads) 4 minutes after intranodal contrast injection (left), as well as tortuous, engorged lymphatic enhancement along the distal thoracic duct and right paravertebral area (white arrowheads) 9 minutes post-injection (right). (D) Following contrast medium injection, minimal intensity projection reconstruction at 20 minutes also showed an engorged lymphatic structure in the right paravertebral area (white arrowhead) with adjacent contrast leakage (black asterisk) and diffuse enhancement along the pleura (white arrows), indicating leakage from the anomalous lymphatic structure. The absence of a cisterna chyli or any other retroperitoneal lymphatic structure of sufficient caliber for guidewire navigation deemed this anatomy technically infeasible for intervention. (E) Consequently, conventional lymphangiography was performed 1 week post-DCMRL via a retrograde approach from the left subclavian vein, corroborating DCMRL findings of leakage from the anomalous structure (black arrowheads). Embolization used a microcoil and N-butyl cyanoacrylate. (F) Chest radiograph immediately post-embolization showed embolic materials in the right paravertebral area (black arrow) and draining right pleural effusion. (G) Over 3 months post-intervention, the right pleural effusion gradually resolved.

CT = computed tomography, DCMRL = dynamic contrast-enhanced magnetic resonance lymphangiography.

Fig. 5. An 18-year-old woman presented with bilateral leg edema for 10 years. She underwent lymphovenous and lymph node-to-vein anastomosis in bilateral inguinal, pretibial, and foot areas. (A) Despite these operations, chest computed tomography 1 month before DCMRL showed bilateral pleural effusion (asterisks), confirmed as chylothorax. (B) DCMRL indicated that contrast media did not travel through the central retroperitoneal and abdominal lymphatics but through abnormally dilated lymphatics in the right pelvic wall. Coronal (left) and axial (right) 3D T1-weighted images with thin maximal intensity projection at the pelvic level demonstrated a tortuous, engorged lymphatic structure along the right pelvic wall (white arrow). (C) Coronal 3D T1-weighted images with maximal intensity projection before and 7, 8, and 10 minutes after contrast medium injection revealed progressive enhancement of abnormal tortuous lymphatic structures along the right thoracoabdominal wall (white arrowheads), draining into the right subclavian vein (black arrow). The retroperitoneal lymphatics and thoracic duct were not discernible. Based on DCMRL findings, the patient was diagnosed with a central lymphatic flow disorder. As no accessible route for embolization to address the chylothorax was present, the anatomy was considered technically infeasible for intervention. (D) Conventional lymphangiography 3 weeks post-DCMRL mirrored the DCMRL findings, displaying prominent lymphatics along the abdominal and chest walls (left) and the absence of normal central lymphatics (right). Due to persistent chylothorax and the lack of feasible anatomy for intervention, the patient ultimately underwent right pleurodesis 3 months after DCMRL, which partially ameliorated the chylothorax.

DCMRL = dynamic contrast-enhanced magnetic resonance lymphangiography, 3D = 3-dimensional.