Use of Different Enteral Feeds to Impact on Chyle Leaks in Oesophagectomy

What is the purpose of the study? Oesophagectomy (surgery to remove a cancerous portion of the oesophagus or gullet) is the cornerstone of treating oesophageal cancer. In recent years, minimally invasive techniques, including robotic assisted oesophagectomy have been introduced. These techniques reduce stress on patients, reduce pain, reduce the length of stay in hospital after their operation, without compromising cancer outcomes (and in some cases improving cancer outcomes). Any surgery carries the risk of complications. One complication that may arise with oesophagectomy is an increase in chyle leaks. Chyle is a fluid produced by the body that helps transport nutrients from the bowel to the bloodstream to allow them to be absorbed and processed. One of the channels that transports chyle, the thoracic duct, is divided as part of an oesophagectomy. Although it is clipped to reduce the risk of chyle leak, this may still occur, in up to 25% of operations. If a chyle leak occurs, a drainage tube needs to remain in the chest for a number of days, there may be alterations in the use of feeding techniques, and in a small portion of cases, there may need to be an operation to stop a leak, or a procedure in the radiology department. The goal of this study is to see whether use of a different type of post-operative feed (medium chain triglyceride or MCT feeds) can reduce the rate of chyle leak. This is already used to treat chyle leaks, and the question is whether using this as the routine post-operative feed can reduce rates of chyle leakage.

What will happen during the study? The oesophagectomy will proceed as planned. As part of all minimally invasive oesophagectomies, a feeding tube will be placed into the small bowel (feeding jejunostomy). This will be used to support feeding and nutrition in the post-operative period. As standard, the investigators also place a drainage tube in the right chest to drain fluid post-operatively. If the patient is recovering well, the investigators will begin feeding on the first day post-operative day. This will either be with the currently used feed (Nutrison Protein Plus) or with the MCT feed (Nutrison Peptisorb). If progressing well, oral intake will be introduced slowly from the third post-operative day. The investigators will measure the volume of fluid draining from the post-operative drain and analyse it in the lab to determine if there is a post-operative chyle leak. Should a chyle leak occur, it will be treated as per our usual protocol, which may include using alternative jejunostomy feeds, intravenous feeding (total parenteral nutrition), a procedure in radiology, or a repeat operation. All patients will be transitioned back to standard feeds for discharge.

Currently, unless contra-indicated (eg a feeding jejunostomy has not been placed), patients post MIO/RAMIO will commence percutaneous enteral feeds on the first post-operative day using a standardized enteral feeding formulation (Nutrison Protein Plus 1.25kcal/ml at 20ml/hr for 24 hours). This is then sequentially progressed until the patient is at their goal rate to meet their caloric and nutritional requirements. If there are no contraindications, PO intake commences on the 3rd post-operative day with restricted PO fluids, followed by progressive introduction of PO intake, as per our standard post-operative protocols.

Our hypothesis is, that by switching the enteral feed from standard feed to an MCT feed, this will reduce the production of chyle, allowing lymphatics to seal, and reduce the incidence of chylothorax post MIO/RAMIO.

This is a prospective randomized controlled trial. The sample size is 148 patients (72 each in intervention and control groups). The principal investigators (JB and WR) will be blinded to which feed patients are receiving. Due to the practicalities of administering jejunostomy feeds, the clinical dietetics service, nursing service and the non-consultant medical staff will not be blinded to the interventions. Due to the need to instigate training for home feeding, patients will not be blinded to the intervention.

Randomization will be conducted using block randomization, under the supervision of the clinical leads (from clinical dietetics and NCHD leads). Patients will be randomized in blocks of 5. This will facilitate streamlining of interventions and reduce the risk of error in dispensing the correct feeds.

The surgical approach is standardized, both for MIO and RAMIO. For Ivor Lewis (2-stage) oesophagectomy, the gastric conduit is mobilized laparoscopically or robotically, fashioning a 4cm wide gastric conduit. A standard lymphadenectomy is performed, and transhiatal dissection is undertaken, with care to widely dissect the oesophagus from the crura, pericardium and from the aorta. Dissection is undertaken with an energy device as per operating surgeon (Harmonic Scalpel, Ligasure or Vessel Sealer). A standard feeding jejunostomy tube is placed, approximately 30-60cm from the duodenal-jejunal flexure. As the pleural is resected, a wide bore (minimum 24Fr) chest drain is placed in the left pleural space. The thoracoscopic phase is undertaken in standard fashion. Peri-oesophageal lymph nodes and subcarinal nodes are resected as standard. The thoracic duct is identified and clipped or ligated proximally and distally, and a portion is resected en-block with the oesophagus. Dissection is undertaken with electrocautery or an energy device (Harmonic Scalpel, Ligasure or Vessel Sealer) as per operating surgeon. A stapled oesophagogastric anastomosis is undertaken in standard fashion, at or above the level of the azygous vein. A wide bore chest drain (at least 24Fr) is placed in the right pleural space and connected to an underwater seal to allow drainage.

The approach for a McKeown operation is similar, although the thoracic phase is the first portion of the operation, and the anastomosis is performed in the neck (usually hand-sewn).

The primary outcome is the incidence of proven chylothorax. To this end, a sample of chest drain effluent from the right chest drain will be sent for cytological and biochemical analysis on days 1, 3 and 5 post-operatively. A chyle leak will be defined as per the Esophageal Complications Consensus Group as:

• A milky effusion >200ml in 24 hours upon initiation of enteral feeds AND/OR pleural fluid analysis demonstrating triglyceride level >100mg/dL AND/OR chylomicrons in pleural fluid (at least 2 elements present to confirm chyle leak)
• A final definition of the severity of chyle leak is based on definitive treatment:
• Type 1: dietary modification
• Type 2: TPN
• Type 3: Interventional radiology or surgery
• Each is subdivided into A) or B) based on volume of effusion at commencement of treatment (A is <1L/24h, B is >1L/24h)

If a chyle leak is established, treatment will be at the discretion of the responsible surgeon and clinical team.

For secondary outcomes

• The volume of chest drain effluent is recorded on a daily basis
• Re-intervention (IR or surgical) will be recorded
• LOS will be recorded
• All other complications will be recorded according to ECCG definitions and will be graded using the Clavien Dindo system.

Following recruitment, patients will be randomized to receive standard jejunostomy feeds, or MCT feeds. Standard feeds will be with Nutrison Protein Plus 1.25kcal/ml. MCT feeds will be initiated using Nutrison Peptisorb 1kcal/ml. Peptisorb has lower fat content, with a higher proportion of MCTs. The feeding protocol for both formulae will be the same, with equivalent rates of enteral feeding:

• Day 1: if no clinical contra-indication, commence feeding either with standard or MCT feeds at 20ml/hr for 24 hours. Feeds to commence at 8am.
• Day 2: if tolerating feeds continue feeds at 20ml/hr for 12 hours, then increase to 40ml/hr for 12 hours
• Day 3: if tolerating feeds increase feeds to 60ml/hr and await dietician review regarding target rate and duration.

By POD6, if there is no chyle leak, all patients will be given standard feeds for discharge home.

For oral feeding, if patients are progressing clinically, they will have restricted fluids on POD3, with introduction of diet in each group from day 4 onwards. This allows a direct comparison of chyle leak rates for 72h of regular vs MCT enteral feeding.