Endoscopically Assisted Nasojejunal Feeding Tube Placement: Technique and Results in Five Dogs

Introduction

Malnutrition is a common complication in the critically ill patient.^1–4 Supplemental nutrition is indicated in animals intolerant of oral feeding. Enteral nutrition is preferred to parenteral nutrition due to the preservation of gut function, improved immune function, reduced rate of infection, reduced duration of hospitalization, and reduced cost to the client.^5–9

Routes of providing enteral nutrition include the use of nasoenteral (nasoesophageal, nasogastric, nasoduodenal, nasojejunal) and enterostomy feeding devices (pharyngostomy, percutaneous endoscopic gastrostomy, percutaneous endoscopic gastrostomy–jejunostomy, gastrostomy, jejunostomy).

The placement of postpyloric feeding tubes is indicated in dogs in which gastric feeding is contraindicated due to vomiting, regurgitation, pancreatitis, hepatobilitary surgery, hepatic lipidosis, and gastroparesis.^10–13 Interest in alternative methods for the placement of postpyloric feeding tubes led to the publication of several articles describing techniques for the nonsurgical placement of enteral feeding tubes in companion animals.^14–22

A nonsurgical technique for the endoscopic placement of a nasojejunal feeding tube described in humans was modified and adapted for use in five dogs.²³ The purpose of this article was to describe the technique and present complications associated with the feeding tube placement in five dogs.

Materials and Methods

Inclusion criteria for this study involved an attempt to endoscopically place a nasojejunal feeding tube between January 2006 and September 2007 and the availability of a medical record for review. The medical records of all dogs that underwent an attempt at endoscopically assisted nasojejunal feeding tube placement were reviewed and the following data retrieved: age, gender, breed, weight, relevant history, relevant histopathalogic findings, anesthesia technique, type of nasojejunal feeding tube used, duration of the proceedure, duration patients were fed via the feeding tube, and presence of complications.

Results

Five cases met the inclusion criteria for this study. Ages of the dogs ranged from 5 to 14 yr (median, 8 yr). Signalment, presenting complaint, and diagnosis are provided in Table 1. Three dogs presented with a history of vomiting and were diagnosed with pancreatitis (dogs 1, 2, and 5). One of these dogs (dog 2) was taken to surgery for debridement and omentalization of a pancreatic abcess.

Table 1 Signalment, Indication, and Clinical Diagnosis in Four Dogs that Underwent Endoscopically Assisted Nasojejunal Feeding Tube Placement

FS, female spayed; MN, male neutered

View Fullscreen

One dog presented with a history of hypoglycemia and seizures (dog 3). Exploratory laparotomy confirmed the presence of a pancreatic mass that was removed and subsequently determined to be a pancreatic carcinoma. Two days after exploratory surgery, the patient developed pancreatitis and a nasojejunal feeding tube was placed endoscopically. One dog (dog 4) presented for regurgitation and was diagnosed with megaesophagus secondary to myasthenia gravis. The clinical data for all five dogs are summarized in (Table 2).

Table 2 Clinical Data from Four Dogs that Underwent Endoscopically Assisted Nasojejunal Feeding Tube Placement

View Fullscreen

Feeding tubes were attempted and successfully placed in all five dogs. The median length of time for the placement of the feeding tube was 35 min (range, 30–45 min). Feeding tubes were advanced at the level of the pylorus in all dogs. An additional stylet was preplaced in one dog in an attempt to stiffen the tube to allow for a more consistent advancement of the feeding device. The two stylets became entangled, resulting in the procedure being repeated. The median length of time dogs were fed through the tube was 7 days (range, 5–18 days). Minor complications such as facial irritation (5/5), intermittent sneezing (5/5), diarrhea (3/5), and fractured sutures associated with the Chinese finger trap (4/5) were observed. The majority of dogs (4/5) demonstrated upper gastrointestinal symptoms (vomiting and regurgitation) while the feeding tubes were in place. Mild epistaxis was observed in one dog. Clogging of the feeding tube with food (2/5) and crimping of the Entriflex and Dobbhoff catheters at the alar fold was observed (3/5) in dogs. No feeding tube was removed prematurely for persistent upper gastrointestinal symptoms, facial irritation, sneezing, epistaxis, clogging, or crimping. One dog prematurely removed the nasojejunal feeding tube after an episode of vomiting.

Discussion

Veterinary literature describes the surgical (laparotomy and laparoscopically), percutaneous, fluoroscopic, and endoscopic placement of enteral feeding tubes.^11–22,24 Surgically and percutaneously placed enteral feeding tubes are not indicated in all patients requiring postpyloric nutrition. Surgical placement of enteric feeding tubes subjects a patient to significant stress from the procedure alone. Surgical and percutaneously placed enteral tubes can create abdominal pain, local physical irritation, and bile-induced dermatitis at the stoma site in dogs.^13,24,25 The presence of surgically or percutaneously placed feeding tubes commits a patient to a feeding device for 10–14 days while an adhesion forms between the gastrointestinal tract and the body wall. Inadvertent tube removal, which is common to all techniques, poses a greater risk for major complications, such as local peritonitis, diffuse peritonitis, and death.^11,25

Multiple human and veterinary studies demonstrated the role of endoscopy and fluoroscopy as viable alternatives for the correct and rapid positioning of nasojejunal feeding tubes.^{16,17,19,21–23,26–30}

Techniques for the fluoroscopic and endoscopic placement of nasojejunal feeding tubes can be classified into three general categories: pull along technique, Seldinger technique, and a push-along technique.³¹ In the endoscopically assisted pull along technique, a suture is used to attach the tip of the feeding tube to the tip of the endoscope, which are simultaneously advanced through the pylorus and into the intestines. The feeding tube is released and the endoscope withdrawn, leaving the feeding tube positioned within the intestine.²⁹ In the Seldinger technique, endoscopy or fluoroscopy is used to preplace a guidewire within the intestines. Guidewires are pushed across the pharynx, stomach, and into the pylorus, or are deployed through the biopsy channel of an endoscope that has been positioned within the intestines. Depending on the technique used, an oronasal transfer may be necessary, after which the feeding tube is advanced over the guidewire into the intestines.³⁰ The technique used in this study was a variation of the push-along method of deploying an enteral tube. Using endoscopic or fluoroscopic assistance, a feeding tube was manually pushed across the pylorus and into the intestines.^17,19,21,23 Regardless of the technique used, tube placement must be evaluated with radiography or fluoroscopy.

Three fluoroscopic studies for the placement of jejunal feeding tubes were described in veterinary literature, utilizing both the Seldinger and push-along techniques.^17,19,21 Fluoroscopic assistance in the placement of nasojejunal feeding tubes was highly accurate in achieving the correct tube placement; fluoroscopic assistance was rapid and, in some instances, might be performed with only heavy sedation.¹⁷ The primary disadvantage of fluoroscopic guidance was the limited availability of veterinary fluoroscopic units, minimizing the application of these techniques to large referral practices and institutions. Furthermore, difficult transpyloric passage and duodenal placement were reported.¹⁷ An article describing the endoscopic placement of a nasojejunal feeding tube using a Seldinger technique was recently published.²² Feeding tubes were quickly and accurately deployed in all dogs in which they were attempted. The use of an oronasal transfer of the feeding tube did not appear to be as cumbersome as described in human studies.³⁰ The evaluation of the tube placement with fluoroscopy was unnecessary and could be replaced by radiography. Finally, the use of a limited number of healthy dogs could have influenced the speed and accuracy of this technique.

Advantages of the endoscopically assisted push-along technique for the placement of a nasojejunal feeding tube used in the present study included its simplicity, minimal risk of complications, high degree of successful tube placement, wide availability of endoscopy, and limited need for specialized equipment (i.e., fluoroscopy).

The authors considered the push-along technique at the level of the pylorus technically challenging. Handling the endoscope, grasping forceps, and feeding tube simultaneously took a coordinated effort. Difficulty gripping the shaft of the feeding tube with the grasping forceps was observed in all patients. This was attributed to the angle with which the grasping forceps approached the feeding tube. The more parallel the grasping forceps approached the feeding tube, the more difficult it was to grip and hold on to the shaft of the tube. Once grasped, all tubes were positioned through the pylorus easily. All nasojejunal feeding tubes attempted were successfully placed, which was compatible with human literature but might have been influenced by the small sample size.^23,26–30 The median time for the placement of feeding tubes in this study was 35 min. This was compatible with other nonsurgical techniques for the placement of jejunal feeding tubes described in veterinary literature but was considerably longer than what was reported in human literature.^17,19,21,22 The average times reported for the endoscopically assisted enteral feeding tube placement in humans were typically less than 15 min.^25–30 The longer procedure times observed in this study were attributed specifically to the difficulty in grasping the shaft of the feeding tube with the grasping forceps.

Nasoenteric feeding devices are considered appropriate when supplemental feeding requirements are expected to be less than 30 days.³¹ Tolerance to enteral feeding is multifactorial. Anatomic location of the tube influences patient comfort and thus patient tolerance. Nasoenteric feeding tubes induce oropharyngeal and nasal stimulation, which could limit the duration of time with which tubes are tolerated. Jergens et al. reported a percutaenous technique in healthy dogs and cats, where a jejunal feeding tube placed through a percutaneous endoscopically placed gastrotomy tube demonstrated an average feeding time of 14 days, with one dog being fed via the tube for 30 days.¹⁴ In contrast, studies by Wohl et al. and Beal et al., which placed nasojejunal feeding devices, reported median durations of tube feeding to be 3 and 5.25 days respectively.^17,21 The median duration of time dogs were tube fed in this study was 7 days. Small numbers of patients, lack of case variability, and personal clinical bias as to the duration of time with which patients were tube fed likely influenced this value. The location of the feeding tube as it relates to the duration of time with which dogs tolerate enteral feeding is unclear at this time and needs further study. The technique discussed in this study is no longer being used and has been modified to minimize oropharyngeal stimulation in an attempt to improve patient comfort and maximize the duration with which patients can be fed through the device.

The placement of a nasoenteric feeding tube in a dog with myasthenia gravis was performed in this study and could be considered controversial. Increasing nasal and oropharyngeal stimulation in a dog regurgitating from esophageal disease should be performed with caution. Techniques such as the percutaneous endoscopically assisted gastrojejunostomy tube could also be considered a viable alternative to provide enteral nutrition in such patients.

Mechanical complications associated with the nasoenteric tubes included clogging of the tube with food, crimping of the tube at the alar fold, and removal of the luer lock connector from the Fredrick Miller feeding tube line. Clogging of nasojejunal feeding tubes in this study was likely due to the combination of pushing oral medications through the feeding tube, small tube size, and precipitation of protein components of the diet in the lumen of the tube. Clogging of the feeding tubes might have been minimized by flushing the tube with water every 6 hr regardless of whether a constant rate infusion of food was being administered. The Enteriflex and Dobbhoff catheters crimped as tubes were passed around the alar fold across the nasal planum and between the eyes. The Enteriflex catheters are made of a soft collapsible polyurethane designed to reduce mechanical irritation. Running the tube along the upper lip (labium superius) for 2.5–5 cm before positioning the tube on the nasal planum corrected the crimping in both cases. The removal of the luer lock connector was observed on both Fredrick Miller feeding tubes. This was managed by placing a Chinese finger trap connecting the feeding tube to the luer lock connector.

Minor complications observed during nasojejunal feeding included vomiting, regurgitation, diarrhea, facial irritation, and minor epistaxis. Upper gastrointestinal symptoms were present in all patients before tube feeding and persisted in four patients during a portion of the feeding. It is unknown whether the placement of the feeding tube aggravated the upper gastrointestinal symptoms. It is unknown whether the persistence of the gastrointestinal symptoms observed in all patients was the result of tube migration as radiographs were not taken at predetermined intervals while the feeding tubes were in place or at the time of tube removal. Oral migration of nasojejunal feeding tubes is a known complication.^19,21,22 Documenting the location of feeding tubes during their use as well as at the time they were removed would have provided additional evidence regarding the utility of the procedure and added insight into the significance of the complications (i.e., vomiting and regurgitation) observed. One patient (dog 5) was able to remove the nasojejunal feeding after an episode of vomiting. The patient was ultimately euthanized due to acute necrotizing pancreatitis. Diarrhea is a common complication that develops in the majority of patients in human critical care wards fed enteral diets.³² The authors observed variations in stool consistency in all patients. Changes in stool consistency could have been the result of multiple causes that included but were not limited to concurrent antibiotic therapy, fat malabsorption, and rapid food administration. Dogs in this case series were fed a commercially available polymeric enteral diet. Polymeric enteral diets are considered well tolerated by humans, producing equivocal results to element diets when delivered through enteral feeding tubes placed in people with normal or near normal intestinal function. Facial irritation and sneezing were observed in all dogs as a result of mechanical irritation of the nasal cavity with the feeding tube. Minor epistaxis was observed in one patient with a polyvinyl chloride nasojejunal feeding tube. Polyvinyl chloride is a stiff material, which likely traumatized the nasal mucosa. The feeding tube in this patient was in place for a prolonged period of time (18 days), increasing the likelihood for chronic irritation. The epistaxis was not deemed severe enough to warrant the removal of the feeding tube.

Conclusion

In summary, the results of this study confirmed that endoscopy could be used to position nasojejunal feeding tubes with a high degree of accuracy in clinically ill patients. Despite the cumbersome technique, feeding tubes were deployed within similar time frames as reported in other studies. Minor complications were common and premature tube removal was possible.