Management of Intra-operative Chyle Leak during Neck Dissection: Recognition and Control

Chyle fi stula is a rare but serious complication following neck dissection with an incidence ranging from 1% to 2.5% [1]. This uncontrolled extravasation of chyle arise from damage to the thoracic duct, which transports triglycerides and cholesterol from intestinal lymphatics into the venous system [2]. Beginning at the cisterna chyli, the thoracic duct ascends though the aortic hiatus of the diaphragm. It enters the neck to form a loop that courses between the internal jugular vein and anterior scalene muscle [3], where it terminates into the venous system. Due to the integral role of this lymphatic structure in fl uid balance, metabolism, and immunity, unresolved chyle leaks can cause extreme morbidity secondary to the loss of fl uids, electrolytes, and other proteinaceous nutrients. These losses can lead to severe dehydration, electrolyte disturbances, and lymphopenia ultimately delaying wound healing, causing skin-fl ap necrosis and infection, and substantially prolonging the duration of hospitalization [3].


Introduction
Chyle fi stula is a rare but serious complication following neck dissection with an incidence ranging from 1% to 2.5% [1].
This uncontrolled extravasation of chyle arise from damage to the thoracic duct, which transports triglycerides and cholesterol from intestinal lymphatics into the venous system [2]. Beginning at the cisterna chyli, the thoracic duct ascends though the aortic hiatus of the diaphragm. It enters the neck to form a loop that courses between the internal jugular vein and anterior scalene muscle [3], where it terminates into the venous system. Due to the integral role of this lymphatic structure in fl uid balance, metabolism, and immunity, unresolved chyle leaks can cause extreme morbidity secondary to the loss of fl uids, electrolytes, and other proteinaceous nutrients. These losses can lead to severe dehydration, electrolyte disturbances, and lymphopenia ultimately delaying wound healing, causing skin-fl ap necrosis and infection, and substantially prolonging the duration of hospitalization [3].
The management of chyle fi stula is multi-faceted and contingent upon timing of the recognition of thoracic duct injury (TDI). If identifi ed during surgery, operative management by over-sewing and ligating the TDI is indicated.
However, if unrecognized during the index operation, a chyle leak will usually manifest after resuming enteral feeding through the appearance of milky fl uid from neck drainage contents. Findings on physical examination include a ballotable subcutaneous mass the medial neck or supraclavicular fossa, and associated induration, edema, and erythema of the overlying skin [4,5]. Management of a chyle leak includes dietary modifi cation consisting of total parenteral nutrition (TPN) or the implementation of a medium-chain triglyceride (MTC) diet as to bypass the remnant thoracic duct and prevent intestinal peristalsis and lymph fl ow. Adjunctive treatments include the administration of octreotide to decrease triglyceride absorption and inhibit splanchnic circulation [6], elevation of the head of the bed, and application of compressive dressings [7][8][9][10]. Delayed surgical management is indicated for cases in which medical treatment has failed to decrease the amount of chyle drainage by half, or for high-output fi stulas (>1000 ml/day for 5-7 days), or when there are serious complications such as chylothorax with respiratory insuffi ciency and severe malnutrition/electrolyte disturbances [11]. In these instances, the neck is re-explored and thoracic duct is ligated.

Abstract
Chyle fi stula (CF) is a rare complication of neck dissection. The extravasation of chyle can result in potentially devastating metabolic, nutritional and immunologic sequellae.
We report the effi cacy a protocol for treatment of intraoperative (CF).
Hospital length of stay, time to oral alimentation, and type of diet were analyzed.
There were 19 patients with thoracic duct injury development following neck dissection (0.08%) The mean age was 62 years and the majority were male with squamous cell carcinoma of the oral cavity. The TDI's were identifi ed on the left side in 16 patients and on the right side in 3 patients. In all cases TDI were identifi ed intraoperatively, packed with micro-fi brillar collagen and oversewn with monofi lamemnt nylon. In one patient required re-exploration and placement of a muscle fl ap. The mean number of days NPO was 2.5 (range 1 to 13 days, SD ± 2.8). The mean LOS was 4 days (range 2 to 14 days, SD ± 2.7). Only patient number 5 and 12 required MCT administration for 14 and 12 days and no patients required parenteral nutrition. Herein we describe an approach to the treatment of thoracic duct injuries, focusing on early intraoperative recognition and immediate repair. The aim of this study is to demonstrate the effi cacy of Avitene, a microfi brillar collagen hemostat, and post-operative adherence to a low fat, low carbohydrate diet as an adjunct to the standard intraoperative management of a chyle fi stula.

Materials and Methods
Between 1995 and 2015, 1736 patients underwent 2381 neck dissections or node biopsies by the senior author (GJP). After eliminating sentinel node biopsies (96) and Zone 1,2 3 selective neck dissections (1100), there were 1185 necks at risk for TDI having undergone either comprehensive neck (161), modifi ed neck (995), zones 4 node dissection (1) or Zone 6-para-tracheal node (28) dissections. Of the 1185 at risk necks, a total of 19 thoracic duct injuries were identifi ed intraoperatively ( Figure  1). Data was collected from a password protected database of deidentifi ed patient surgical records maintained by the senior author.
In all cases, the TDI was identifi ed intraoperatively, oversewn with 4-0 mono-fi lament non-absorbable suture, packed with Avitene microfi brillar collagen hemostatic material and a medially directed suction drain was placed. Patients remained nil per os for 24 to 48 hours then began a high protein, low-fat, low-carbohydrate diet.
Demographic information was collected regarding patient age, gender, primary cancer, primary tumor site, and TDI sidedness. The primary outcome was the presence or absence of a delayed chyle leak, and secondary outcomes evaluated were length of hospital stay, days NPO, and need for MCT and/ or TPN. Statistical analysis of data was performed using the SPSS statistical package XXX. Chi-squared and Fisher exact tests were used to compare percentages.

Results
There were 19 patients with thoracic duct injury development following neck dissection with an overall incidence of 0.80%.
Among the 19 patients, the mean age was 62 years, and the majority was male with a diagnosis of squamous cell carcinoma of the oral cavity (Table 1). In the patient population who suffered TDIs, 14 underwent modifi ed radical neck dissections, 4 underwent selective neck dissections, and 1 underwent a radical neck dissection. The procedures were performed on the left side in 16 patients and on the right side in 3 patients (Table  2). Specifi cally, the incidence of TDI following a right-sided neck dissection was 0.24% and the incidence following a leftsided neck dissection was higher at 1.42%.
In one patient intraoperative management was not successful and he required a return to the operating room for re-exploration and placement of a sternocleidomastoid muscle fl ap. The mean number of days NPO was 2.5, ranging from 1 to 13 days (SD ± 2.8). The mean duration of hospitalization was    4 days with a range from 2 to 14 days (SD ± 2.7). Only patient number 5 and 12 required MCT administration for 14 and 12 days, respectively, and no patients required TPN (Table 2).

Discussion
The best method for prevention of a chyle fi stula is prompt recognition of the index thoracic duct injury. In addition to standard practice of direct suture repair and drain placement, we advocate for the use of intraoperative Avitene and postoperative compliance with a low-carbohydrate low-fat diet. As a result, we encountered no delayed chyle leaks. Avitene is a microfi brillar collagen hemostat (MCH) that accelerates platelet and protein aggregation, resulting in a fi brin plug. It is naturally derived, lacks antigenicity, and is insoluble in water. These properties have made it a popular choice for hemostasis in neurosurgical, urological, and endoscopic surgery, and recent literature has shown that MCH's structure can increase surface area for cell adhesion as well as provide a lasting scaffolding for new cartilage synthesis [12]. Santaolalla also described the use of fi brin, polyglactin, or collagen application to intra-operatively detected TDIs, and reported an incidence of 1.31% (4/304) post-operative chylous fi stula [2]. Gregor et al., also utilized fi brin glue for management of chylous fi stula, both early and delayed, and his study cited an incidence of 5.8% [5]. It is possible that our lack of post-operative chyle fi stula was due not only to our use of a more suitable product, but also because of our pre-emptive application of the MCH at the index operation.
Also unique to our study was that all patients adhered to a low-fat low-carbohydrate diet post-operatively. This nutrition regimen seeks to minimize fat digestion and subsequent lymphatic transit, which ultimately slows the fl ow of lymph through the thoracic duct. Additionally, this diet promotes ketosis, which new research has shown to inhibit the NLRP3 infl ammasome from activating IL-1b and IL-18 generation, thereby dampening the infl ammatory process [13]. Of our 19 patients, 17 responded to this diet while the remaining 2 patients required supplementation with MCT. Both of these patients had left-sided MRNDs for squamous cell carcinoma, were older than the average age of our cohort, and had longer lengths of hospitalization.
Nevertheless, we still sustained injuries to the thoracic duct during neck dissections in this series. We describe an overall incidence of 0.80% for injury, which is much lower than the cited literature values. There are numerous anatomic variations of the duct that may contribute to the occurrence of ductal injury following surgery, such as the existence of multiple tributaries as opposed to one duct or termination into the subclavian vein, innominate vein, and the external jugular vein [10]. Consequently, chyle fi stula occur more commonly following radical neck dissections and are typically left-sided, specifi cally at the base of the neck lateral to the carotid sheath [3]. Not surprisingly, the most common procedure in our series associated with a TDI was a left-sided, modifi ed radical neck dissection.
Unfortunately, there are no exact objective signs of chyle fi stula to aid in immediate diagnosis during the postoperative period. A leak of up to 1 L/day can be tolerated for 1 to 2 days before causing any electrolyte abnormalities [14]. Once identifi ed, medical management is the fi rst line of treatment for delayed chyle fi stulas and is aimed at measures that promote closure by reducing chyle fl ow [15,16]. If these attempts fail, a minimally invasive surgical repair can be undertaken by percutaneous lymphangiography-guided cannulation with embolization of the thoracic duct [17] or through thoracoscopic ligation. Finally, re-exploration of the neck with local muscle fl aps and adjuvant fi brin glue or cyanoacrylate tissue glue are recommended as a last resort14 due to the delicacy of the already damaged thoracic ductal system. Thus, it behooves one to meticulously inspect the wound at the index operation, especially after hyperinfl ating the lungs to increase intrathoracic pressure in order to identify the source of the leak [10]. Once localized, prompt suture repair, placement of hemostatic material (Avitene) and a suction drain, as well as post-operative adherence to a diet that decreases lymph fl ow and infl ammation are instrumental to success.

Conclusion
The management of chylous fi stula is multi-faceted and depends on the timing of identifi cation. We believe that microfi brillar collagen hemostat (Avitene) effectively seals a thoracic duct injury and prevents post-operative chyle fi stula.
Adjuvant diet modifi cation to low-fat low-carbohydrate enhances the effi cacy of the repair.