Thoracic Duct Embolization in Post-Neck Dissection Chylous Leakage: A Case Series of 6 Patients and Review of the Literature

Abstract

Chylous leak is a serious complication of neck dissection. Patients are often managed conservatively, and failure of conservative management necessitates surgical management, which adds to their morbidity. We present a case series demonstrating the value of thoracic duct embolization (TDE) in management of patients with chylous leaks following neck dissection who have failed conservative management, and to review the literature on this specific application of TDE. Between 2011 and 2019, six patients underwent a total of seven TDE procedures. Lymphatic leak was identified, and clinical success was achieved in all patients, with one patient requiring repeat TDE. No minor or major complications were reported. In conclusion, TDE is a safe and effective tool in management of chylous leaks following neck dissection.

Introduction

Chylous leak is a serious complication of head and neck surgery ^1–4. It is reported to occur following 0.5 – 1.4% of thyroidectomies, with incidence rising up to 8% following neck dissections ^1,2,5. Accumulation of chylous fluid in the neck leads to dermal irritation and erythema, and may lead to delayed wound healing, wound dehiscence and electrolyte disturbances with prolonged hospitalization, malnutrition and immunosuppression ^1,2. They occur due to inadvertent injury of the thoracic duct or one of its tributaries in the lower neck, and that risk is amplified by the variable course and branching pattern of the thoracic duct in the neck ^1,2. Patients may present with a palpable fluid collection in the surgical bed or increased output from the surgical drain (particularly following resuming feeding) with creamy fluid character or chylothorax ^1,4. Confirmation of the diagnosis requires lab assessment of the fluid for triglycerides, with a level more than 110 mg/dL confirming its chylous nature ^1,4,6.

Management for post-operative chylous leaks starts with conservative measures ^1,4. In patients with persistent high output, or when conservative measures are ineffective, other options include surgical exploration of the wound to address the leaking lymphatic duct, thoracoscopic thoracic duct ligation, which increase the length of hospital stay and add to the patient’s morbidity, and percutaneous embolization ^1,2,4,5,7,8. Lymphangiography with or without embolization of thoracic duct or the site of leak directly through a percutaneous approach are safe, effective procedures that have been sporadically described in management of chylous leaks after neck dissection ^6–9. Herein, experience from 6 patients with chylous leaks following head and neck surgery managed by Thoracic Duct Embolization (TDE) is presented.

Case Series

Institutional Review Board approved this retrospective study. Medical records review from 2011 to 2019 captured 6 patients who underwent intra-nodal lymphangiography (INL) followed by TDE for management of chylous leaks following neck surgery (Table 1).

Table 1.

Patient characteristics and output of the chylous leak at time of presentation.

Patient Number	Age	Gender	Primary Cancer	Description	Time between Surgery and Lymphatic Intervention in Days	Output from drain one day prior to TDE	Presentation	Presentation Side
1	59	M	Papillary Thyroid Carcinoma	Total thyroidectomy with central dissection and left neck dissection	8	520	Neck leak	Left
2	37	F	Papillary Thyroid Carcinoma	Left subtotal thyroidectomy and level 3, 4 and 6 left lymph node dissection	3	800	Neck Leak and Chylothorax	Left
3	60	M	Papillary Thyroid Carcinoma	Total Thyroidectomy with central and lateral bilateral neck dissection with extended caudal dissection	5	1200	Neck Leak	Left
4	30	F	Papillary Thyroid Carcinoma	Total Thyroidectomy with central and lateral bilateral neck dissection	3	600	Neck leak	Left
5	61	F	Recurrent Papillary Thyroid Carcinoma	Total laryngectomy, pharyngectomy, bilateral central neck dissection and level 2–4 lymph nodes	21	Open wound, output quantified as dressing changes	Neck Leak	Left
6	76	F	Hurthle Cell Thyroid Carcinoma	Total Thyroidectomy with modified radical neck dissection and Thoracic lymphadenectomy	13	250	Neck leak and Chylothorax	Bilateral

Technique

Technical aspects of INL and trans-abdominal TDE have been previously reported (Fig. 1) ^10,11. Once INL is completed with visualization of the central lymphatics, specifically the cisterna chyli (CC), as well as the site of the lymphatic leak, catheterization of the CC is attempted via transabdominal approach. Under fluoroscopic guidance, a 22-gauge Chiba needle (Cook Inc, USA) is used to access the CC. Adding a gentle curve to the distal 1 cm of the Chiba needle improves steerability.

Fig. 1:

Figure (a) shows the opacified cisterna chyli (thick arrow) after completion of the intra-nodal lymphangiogram. Cannulation is done using a Chiba needle with a gentle curve added to its distal 1 cm (arrow-head) under fluoroscopic guidance with no contrast injection so that the access is not compromised. Figure (b) shows the 0.018” inch wire (thin arrow) within the thoracic duct after successful cannulation of the cisterna chyli

Figure (c) shows a fluoroscopic image of the opacified thoracic duct (thick arrow) and the chylous leak (asterisk) with a small lymphatic tributary (thin arrow) that is likely contributing to the leak

Figure (d, e and f) show fluoroscopic images of the pull-back technique used to embolize the thoracic duct. The microcatheter tip (thin arrow) is pulled back while simultaneously injecting the n-BCA - lipiodol mixture (ratio 1 to 1). The n-BCA-lipiodol mixture opacifies the thoracic duct as well as small lymphatic tributaries (thick arrow) that are contributing to the leak.

Figure (g) shows a fluoroscopic image of the entire length of the thoracic duct (thick arrow) opacified by the NBCA-lipiodol mixture with coils (thin arrow) seen within the distal thoracic duct.

After confirmation of accessing the CC using oblique fluoroscopic views and contrast displacement within the CC on movement of the Chiba needle, a 0.018 inch V-18 guidewire (Boston Scientific, USA) is used to cannulate the thoracic duct (TD), with the tip of the Chiba needle aimed cranially to facilitate access. Confirmation of access into the TD relies on tactile feel of the wire within the lumen and visualization of the wire within the course of the TD with minimal resistance.

The Chiba needle is then exchanged over wire for a 2.4 French Progreat microcatheter (Terumo, Japan) which is navigated over the guidewire into the thoracic duct as cephalad as possible. Omnipaque 300 (General Electric Healthcare, USA) is then injected through the microcatheter to confirm the site of leakage and evaluate the anatomy of the distal thoracic duct. Then, TDE is initiated by deploying two or three Tornado embolization microcoils (Cook Inc, USA) in the most cephalad portion of the thoracic duct. After flushing the microcatheter with 5% dextrose solution,1:1 mixture of lipiodol (Guerbet, Villepinte, France) and n-butyl-2-cyanoacrylate (n-BCA) (TRUFILL, Codman Neuro, Raynham, Massachusetts, USA) is slowly injected through the microcatheter from the level of the deployed coils while pulling it backwards to avoid forward propagation of the glue-mixture into the venous system, to the entry site into the CC, and the catheter is promptly removed. If no leakage was identified, TDE was not done. The procedure was deemed technically successful if TDE was accomplished. The procedure was deemed clinically successful if the surgical drain was removed within 7 days without the need for surgical intervention.

Outcomes

Six patients underwent seven INL and seven TDE procedures. One patient received two INLs, the first TDE attempt was unsuccessful (failure of catheterization of the CC), however, TDE was successfully performed 15 days later. One patient underwent one lymphangiography and 2 TDEs as the first TDE (using n-BCA-lipiodol mixture with no microcoils) failed to stop the leak and it was repeated 5 days later (using microcoils and n-BCA lipiodol mixture) without the need for repeat lymphangiography. Technical success for lymphangiography was 100% (7/7) and for TDE was 86% (6/7). Lipiodol extravasation in relation with a lymphatic leak was identified on live fluoroscopic images in all patients in the left neck (n=5) and TD tributary at T4–5 (n=1). Clinical success was achieved in all patients (100%) and either the cutaneous leak stopped (n=1) or drains were removed 3–5 days after TDE (n=5) (Fig. 2). Follow up after the procedure ranged from 5 months to 36 months with no minor or major complications reported.

Fig. 2:

A graph showing the decrease in surgical drain output (in ml) one day prior to, and two days following thoracic duct embolization. One patient did not have a surgical drain and therefore was not included in this graph.

Discussion

Initially described in 1998 by Constantin Cope, percutaneous transabdominal catheterization of the CC provides access to the TD and allows identification and management of sites of lymphatic leakage that was previously not possible without surgical exploration ¹². The initial description required a pedal lymphogram to opacify the lymphatic system ¹². Pedal lymphogram is a time-consuming and technically challenging procedure that has largely been replaced by percutaneous, ultrasound guided INL ^10,13. TDE has proven efficacy in treatment of chylous leaks caused by injury to the thoracic duct ¹¹. With a reported success rate of around 70% and no added morbidity to the patient, TDE is an alternative to surgical re-operation, saving the latter as a last resort ⁴.

Chylous leaks following neck dissection for thyroid cancer at specialized centers have been reported in about 1% of patients ¹⁴. The rapid accumulation of chyle in a confined space in the neck can lead to swelling which can compromise respiration if not identified and drained in a timely manner ¹⁵. Slow leaks can lead to formation of chylous lymphoceles, or “chylomas”, weeks or months later, which can be managed by percutaneous drainage and sclerotherapy or surgical resection ^16,17. Prolonged chylous leaks can lead to nutritional deficiency, hypovolemia and immunosuppression ^1,2,18. When identified intra-operatively, surgeons advocate for immediate management to avoid these complications ¹⁸. When identified post-operatively, early surgical management following a short trial of conservative management is advised, as the prolonged leak adds to complexity of the surgery ¹⁸. In the rare occurrence of failed conservative and surgical management, mortality can be as high as 50% ¹⁸.

Lymphangiography with or without embolization of the TD or the site of leak have been reported in management of chylous leaks from neck surgery (Table 2). One study described the procedure as time consuming, likely due to the use of pedal lymphograms to opacify the lymphatic system, and painful, indicating that local anesthesia and sedation were not enough ⁷. The use of TDE and Thoracic Duct Disruption (TDD) in patients with chylous neck leaks has been reported in a total of six patients within larger groups of patients with mixed etiologies ^11,19. One case reported that lymphangiogram alone can have a therapeutic effect, which has been reported to occur in up to 60% of post-surgery chylous leaks, more so in minor leaks ^9,20. Another case reported the use of Cone-Beam Computed Tomography to percutaneously access and embolize the site of leakage directly ²¹. Only one case report of a chylous leak following neck dissection treated by TDE with the use of INL is available in the literature ⁸.

Table 2.

Review of the cases reported in the literature.

Author	Year	Patients	Original Surgery	Surgery to address Chylous Leak	Lymphangiography	Intervention	Results
Van Goor et al 7	2007	2	Total Thyroidectomy and left selective neck dissection (levels II - V) in one patient and bilateral in the second patient	One patient underwent thoracic duct ligation and pectoralis major muscle flap on Post-operative day 16, the second patient did not get surgical intervention	Pedal	Thoracic duct embolization	One patient had to receive repeat TDE for persistent high output, with ultimate control of leak in both patients
Patel et al 6	2008	1	Total Laryngectomy with bilateral selective neck dissection	None	Pedal	Thoracic duct embolization (Duplicated ducts visualized; both embolized)	Decreased output from surgical drain with recovery
Itkin el al 11	2010	2	Neck dissection for cancer and lipoma	Not reported individually	Pedal	Thoracic Duct Embolization	Not reported individually
Pamarthi et al 14	2012	4	Neck dissection, thyroidectomy and thyroid mass resection	Not reported individually	Pedal	Thoracic Duct Embolization and disruption	Not reported individually
Chen et al 8	2015	1	Left radical neck dissection	Leakage site repair with fibrin glue on Post-operative day 10	Intra-nodal	None	Surgical drain removed 7 days later
Ierardi et al 16	2016	1	Total Thyroidectomy and radical neck dissection	None	Intra-nodal	Direct percutaneous embolization of the site of leak	Surgical drain removed 4 days later
Kluijfhout et al 9	2016	1	Left lateral neck dissection	Surgical exploration identified fistula, but localization and management of leak was unsuccessful	Intra-nodal	Thoracic Duct Embolization	Surgical drain removed 1 day later

In this case series, the source of the leak was identified and controlled in all patients. Similar to previous reports, the main obstacle remains cannulation of the CC, but this report falls in line with previous reports highlighting the importance of operator experience ¹¹. The reported cure rate for TDE in management of traumatic chylous leaks of all etiologies is 90%, which is similar to this series ¹¹. Due to improved technique, patients did not complain of intraprocedural pain as reported previously ⁷. This may be attributed to the use of a non-coaxial 22G Chiba needle for cannulation of the CC, as opposed to a larger 18G coaxial system, and may also be because a 4 French introducer was not used after cannulation, but instead the microcatheter was used directly ⁷. There were no acute complications in this series which is in concordance with the literature. The reported rate of complications for TDE in other studies is as high as 14% and included chronic leg edema, chronic diarrhea and asymptomatic embolization of lungs with glue ^11,22.

Management of post-operative chylous leaks typically begins with conservative measures, which include bed rest with mild elevation of the head of the bed and dietary transition to a non-fat/low fat diet or a medium-chain fatty acid diet. Other conservative measures include subcutaneous octreotide, the use of suction drainage and/or pressure dressings, and the injection of sclerosing agents through the drain ^1,3. While low output chylous leaks (<1000 ml/day) often respond well to these conservative measures, they often do after several days, and high output chylous leaks (>1000 ml/day) rarely respond well ³. Management options then include thoracoscopic thoracic duct ligation, surgical exploration and TDE ^1,4. Thoracoscopic thoracic duct ligation is an attractive alternative to surgical exploration, but still carries significant morbidity ⁷. Surgical exploration is usually resorted to when all else fails, but the complexity of the surgery is often amplified by local inflammation caused by the extravasated chyle and the malnourished and dehydrated state of the patient ¹.

Conclusion

TDE is a safe and effective tool in management of chylous leaks elsewhere in the body, and this case series shows its potential value in chylous leaks following neck dissection, which was previously only reported in sporadic case reports. Owing to its very low complication rate, rapid success rate, and ability to repeat the procedure in case of inadequate embolization, it should be pursued prior to surgical exploration in this patient population. Better understanding of its value can be achieved from a prospective study.