Development and Characterization of a Technique for Percutaneous Radiologic Gastrojejunostomy Tube Placement in the Dog
Rebekah M. Mack, DVM; Benjamin Staiger, DVM, DACVS; Daniel K. Langlois, DVM, DACVIM; Stephen J. Mehler, DVM, DACVS; Nathaniel Lam, DVM, DACVS; Trevor Moore, DVM; Andrew Brown, DVM, DACVECC and Matthew W. Beal, DVM, DACVECC
From Small Animal Clinical Sciences, Michigan State University, East Lansing, MI (Mack, Staiger, Langlois, Beal); Hope Veterinary Specialists, Malvern, PA (Mehler); VCA Family Animal Hospital, Pearl City, HI (Lam); College of Veterinary Medicine, Purdue University, West Lafayette, IN (Moore); and the College of Veterinary Medicine, University of Edinburgh, Edinburgh, Scotland (Brown).
Please address correspondence and offprint requests to: Dr. Matthew W. Beal, 736 Wilson Rd., Michigan State University, East Lansing, MI48824.
This investigation was funded by the Michigan State University’s MSU College of Veterinary Medicine Companion Animal Fund.
Dr. Matthew Beal was an advisory board member for Infiniti Medical LLC (Menlo Park, CA) at the time of this investigation. The authors declare no other conflicts of interest.
Presented in part as an abstract at the American College of Veterinary Internal Medicine Forum inMontréal, Quebec, Canada, 2009.
Running Title: Percutaneous Radiologic Gastrojejunostomy in Dogs
Abbreviations
EJesophagojejunostomy
ENenteral nutrition
GJTgastrojejunaltube
HGWhydrophilic guide wire
PRGJpercutaneous radiographic gastrojejunostomy tube
SDstandard deviation
Abstract
Objective: To develop and describe a technique for percutaneous radiologic gastrojejunostomy tube placement in the dog.
Design: Prospective technique development study
Setting: University teaching hospital
Animals:Six healthy adult male beagles
Interventions:Following anesthetic induction, fluoroscopic and ultrasound guidance were used to identify an appropriate gastropexy site on the left lateral abdomen. Gastropexy was performed using gastrointestinal suture anchors. An over-the-wire catheter technique using fluoroscopic guidancewas used to achievejejunal access. An 18F/8F, 58cm, dual-lumen gastrojejunal feeding tube was placed via serial over-the-wire dilation of the body wall using an 18F peel-away introducer kit. Tube location was determined radiographically immediately following placement andon days 2, 4,after emesis on day 4, and at time of gastrojejunal feeding tube removal (day 16 – 18).
Measurements and Main Results:Percutaneous radiologic gastrojejunostomy (PRGJ) tube placement was successful in all dogs. Median time to pyloric passage with the guide wirewas 23.5minutes (range, 9 – 93 minutes). Median total procedure time was 53minutes (range, 49 – 113 minutes). Significant tube migration was not observed at any point during the study. One dog developed linear foreign body obstruction secondary to the tube on day 5 that was relieved by release of the jejunal component. Other complications were minor and included mild to moderate peristomal inflammation in all dogs and removal of thefeeding tube on day 3by 1 dog. Feedings were welltolerated in all dogs.
Conclusions:PRGJ tube placement in the dog is a safe andminimally invasive technique that allows forjejunal feeding without surgeryor endoscopy. The high success rates, acceptable procedural times, and minimal complications are appealing for use in critically ill patients. Although additional evaluations are needed, PRGJ tube placement may be considered for dogs that require post-pyloric feeding with or without gastric decompression.
Keywords: enteral nutrition, fluoroscopy,gastrojejunostomy, post-pyloric feeding
Introduction
Recent studies in people and veterinary patients have reported associations between nutritional status and outcome.1-4Benefits of nutritional support extend beyond control of the catabolic state and include enhanced wound healing and improved immunologic function.1,3,5,6Historically, parenteral nutrition was usedin human and veterinaryhospital settings, but improved methods for enteral support have allowed safe and controlled means of nutritional provision with fewer complications and lower costs.1,3-6 Enteral nutrition (EN) is associated with shorter hospitalization times in both dogs and people, and reduction in catheter-related infections(compared to parental nutrition) has been reported in people.1-4,6,7 Furthermore, mounting evidence supports direct benefits for gastrointestinal health in veterinary patients, including reduced villous atrophy, increased enterocyte survival, improved mucosal barrier integrity, and decreased bacterial translocation.1,6,8,9
Enteral nutrition is frequently delivered via nasoesophageal, nasogastric, esophageal, and gastrostomy tubes in critically ill dogs.Whilethese feeding devices aresimple to place, patients withvomiting or regurgitationmay not consistently tolerate feedings delivered through them.10-12Inpeople, post-pyloric enteral feeding has become an accepted method to circumvent complications seen in vomiting patients with traditional feeding tubes.13-16
Numerous post-pyloric techniques, including: surgically placed jejunostomy and gastrojejunostomytubes(GJT), laparoscopically placed jejunostomy tubes, endoscopically placednasojejunostomy tubes,esophagojejunostomy (EJ) tubes,GJT, and fluoroscopically placednasojejunaltubes have been described in veterinary medicine in recent years.11,12,17-22Surgically placed jejunostomy tubes are associated with complications such as peristomal inflammation and infection, tube dislodgement, tube occlusion, and septic peritonitis.23Complications have been reported in 17.5 to 40% of dogs.18,20Surgically placed GJT have fewer major complications, but inflammation and infection are still reported frequently at peristomal sites.17,19In addition, tube migration, kinking, and dislodgement have been reported in 16 to 46% of cases.17,19 Although laparoscopically placed post-pyloric (jejunal) feeding tubes are less invasive, as reported,they extend procedural time approximately by twice that of other techniques.20 A minimally invasive, endoscopic assisted “PEG-J” (percutaneous, endoscopically placed jejunostomy-through-gastrostomy tube) placement has also been described in healthy dogs.17 However, fluoroscopic placement verification is still requiredand complications were observed in 40% of dogs including tube removal, retrograde tube migration, and peristomal inflammation.17Additionally, a recent case series of endoscopically assisted EJ tube placement in clinical patients was published.22 Placement was successful in all patients, but similar complications of tube obstruction (40%), vomiting (80%), and peristomal inflammation (20%) were noted. Procedural length was not described. Fluoroscopic nasojejunaltube placementhas been retrospectively described in dogs.11 Although successful placement occurred in greater than 90% of dogs, orad tube migration occurred in 25% of cases.
Given the limitations of current post-pyloric feeding methods, development of alternate techniques is needed. Ideally, new methodologies would allow minimally invasive and consistently reliable tube placement. The purpose of this study was to characterize PRGJ placement in clinically healthy dogs. Procedural time, methodology,and complicationswere assessed. Furthermore, the effects of emesis on tube migration and placement were also evaluated.
Materials and Methods
Animals:Six young adult male intactbeagles were used for study completion. The dogs were housed in freely accessible runs and fed a balanced commercial diet. The median age was 11 months (range, 10 – 12 months) and the median weight was 9.5kg (range, 7 – 12kg). All dogs were considered to be healthy based on normal physical examination, CBC, and serum biochemistry panel. This study was approved by the Institutional Animal Care and Use Committee at Michigan State University.
Experimental Protocol: All 6 dogs underwent PRGJ tube placement. On day 1, GJT placement was performed under general anesthesia using the technique described below. Dog dependent, intermittent feeding with gradual escalationwas utilized to reach resting energy requirements. Radiographic assessment of tube placement was performed on days 1,2, 4, and at GJT removal (day 16 – 18). Four daysfollowingPRGJ placement, emesis was induced using conjunctival apomorphine.a The location of the distal tip of the jejunal component of the tube was assessed radiographically immediately before and after emesis. Clinically relevant tube migration was considered aborad movement of the distal tip of the jejunal componentinto the duodenum or stomach. The GJT remained in place for an additional 12 – 14 days after emesis induction. On day 16 – 18 the GJT was removed and the stoma was allowed to heal by second intention.
Anesthetic Protocol: Dogs were premedicated with acepromazineb (0.05mg/kg) IM 20 minutes prior to anesthetic induction. A 20-Ga over the needle catheter was placed into a cephalic vein. Anesthesia was induced using repeated 1mg/kg IV propofolc boluses administered slowly to effect. Anesthesia was maintained via endotracheal intubation with isoflurane in 100% oxygen. Dogs were monitored with continuous pulse oximetry and ECG, and serial Doppler blood pressure measurement throughout the anesthetic period. Analgesia was provided withbuprenorphined(0.01mg/kg). The initial dose was given IVafter successful PRGJ placement and before extubation.The same dose was administeredIV every 6-8 hours, as needed for pain, for 48 hours after PRGJ placement.
Technique: Following anesthetic induction, dogs were positioned in right lateral recumbency. An orogastric tube was placed to allow for gastric insufflation, and maintained throughout the procedure. Fluoroscopy was used to identify a location caudal to the left 13th rib where the gastric wall was apposed to the abdominal wall. This area would pose minimal risk to the patient and also serve as an appropriate anatomical location for an ostomy. Brief ultrasound examination ensured that the spleen was not interposed between the stomach and the body wall. This site was widely clipped and aseptically prepared and draped. Air was insufflated through the orogastric tube until the stomach extended beyond the 13th rib. The location was verified again with
fluoroscopy, and 3 gastrointestinal suture anchorsewere deployed through the body
wall and stomach in a triangular pattern, creating a gastropexy (Figure 1). An18-Ga,5 – 6cm introducer/puncture needlef was advanced through the center of the triangle created by the gastrointestinal suture anchors. A 0.035in, 150cm, standard stiffness, hydrophilic guide wire (HGW)g with a straight flexible tip was passed through the puncture needle and into the stomach. The 18-Ga needle was removed. A lubricated 65cm, 5-FrBerensteincatheterhwas then passed over the HGW and used to direct the wire across the pylorus and into the small intestine. If pyloric visualization was difficult because of positioning, the patient was rotated upward into a more sternal position to allow the pylorus to fill with air. The Berenstein catheter and HGW (always leading with the wire) were advanced through the duodenum and into the jejunum, just beyond the duodenojejunal flexure. This is a location where the bowel turns dorsally following the ascending duodenum. Three to 5 mL of iodinated contrastiwere infused through the Berenstein catheter to confirm the location of the catheter tip. The Berenstein catheter was then removed, leaving the HGW in place. Serial over-the-wire dilations of the body wall were accomplished using a coaxial dilator system and an 18-Fr peel-away introducer kit.j An 18-Fr/8-Fr, 58cm, dual-lumen gastrojejunal feeding tube (GJT)k was advanced over the HGW and into the jejunum, past the duodenojejunal flexure(Figure 2). Again, 3 – 5 mL of iodinated contrast medium were used to ensure appropriate placement. The GJT was secured against the body wall using the integrated balloon (Silicone Internal Retention Balloon)l and external bolster (SECUR-LOK External Retention Ring®).m The balloon was insufflated with 10 – 12 mL sterile 0.9% NaCl (Figure 3). The GJT was left in place for 16 – 18 days to ensure complete stoma formation.
Post-Procedural Monitoring: Dogs were monitored every 8 hours following PRGJ placement for 5 days, then every 12 hours for the remainder of the study period(through day 16 – 18). Major complications were defined as perforation of the duodenum or jejunum during placement, septic peritonitis, subcutaneous abscess, gastrointestinal bleeding, frequent vomiting (> 3 times daily), small intestinal obstruction, or death. Minor complications were considered tube dislodgement, external leakage of gastric contents, peristomal inflammation or infection, tube migration, and intermittent (≤ 3 times daily) vomiting. The tube location, appearance of the peristomal site, abdominal discomfort, evidence of vomiting, and basic vital signs were recorded. The stoma site was evaluated every 6 hours for inflammation and infection using the local qualifiers erythema (present or absent), induration (present or absent), and discharge. Discharge, if present, was classified as serous, serosanguinous, mucoid, or purulent in nature based on gross inspection. Infection was considered to be present if both erythema and induration were present and also accompanied mucoid or purulent discharge, and discomfort at the stoma site. The presence of lethargy, fever, abdominal pain, or frequent vomiting prompted abdominal radiographic and / or ultrasonographic evaluation.
Statistical Analysis
Tube migration (distance of tube tip from the duodenojejunal flexure as compared to original placement) was evaluated in each dog with a paired t-test with Bonferroni correction for multiple comparisons (all time points) using commercially available software.nFor all analyses, P < 0.05 was considered significant. Descriptive statistics were used to describe the mean, median, and standard deviation (SD) for all time points considered when evaluating tube migration.
Results
Successful placement was achieved in 100% of cases (6/6). The median time to traverse the pylorus was 23.5 minutes (range, 9 – 93 min). Median total procedural time was 53minutes (range, 49 – 113 minutes; Table 1).
On day 1,there was a meandistance of 343.5 ± SD 50.9mm from the end of the jejunal tube to the duodenojejunal flexure. Day 2and day 4 mean distances were 325.8 ± SD 54.3 mm, and 332.6 ±SD 33.3mm, respectively. Post-emesis distance on day 4 was a mean of 293 ±SD 53.9mm. On day 16 – 18 at GJT removal, the mean distance was 314.8 ± SD 39.8mm. Migration occurred in every dog directionally towards and away from the duodenojejunal flexure (Table 2), but these differences were not significant. Furthermore, clinically relevant tube migration, as defined by movement of the tip of the jejunal component of the tube into either the duodenum or stomach, was not observed in any dog.
The single major complication was a severe enteroplication (linear foreign body) over the jejunal segment of the GJT on day 5 in 1 dog. This dog developed intractable vomiting, abdominal distention, and abdominal pain consistent with gastrointestinal obstruction.Radiographs revealed small intestinal plication, which was confirmed with ultrasound examination.Theplicationwas relieved by releasing the jejunal component of the PRGJ, which then passed in the feces.Releasing the jejunal component was accomplished by deflating the gastric retention balloon and withdrawing the gastric component of the GJT over a hydrophilic guide wire. Once the length of the gastric component of the GJT was withdrawn, the jejunal tubing was exposed and was cut, thus freeing the remaining tubing to pass. In order to maintain the ostomy, a10-Fr Foley cathetero was placed over the wire and secured using a purse string and finger trap suture pattern in the skin.This dog was included in monitoring analysis for days 1 through 4, but excluded from the extended monitoring analysis. The dog recovered without additional intervention and completed the study period.
Minor complications included occasional vomiting≤ 3 times daily (2/6 dogs),removal of the PRGJon day 3 (1/6 dogs), and peristomal inflammation (6/6 dogs) and infection (1/6 dogs). The dog who removed the PRGJ on day 3 was excluded from all post- procedural analysis. In order to maintain the ostomy, a 10-Fr Foley cathetero was placed over the wire and secured using a purse string and finger trap suture pattern in the skin. The dog recovered uneventfully. Vomiting and ptylism in 2 dogs occurred immediately upon anesthetic recovery; no further episodes occurred. All dogs experienced mild ostomy complications. This included mild inflammation (1/6),moderate inflammation(5/6), and infection (1/6). The dog with peristomal infection recovered uneventfully after treatment with cephalexinp (22mg/kg PO q 12 h for10 days) and local peristomal cleaning.
Discussion
This study is the first to characterize a technique for minimally invasivePRGJ tube placement in dogs. The success rate and overall procedural time compare favorably to other reported post-pyloric feeding methods.10-12,17,20,21 The use of fluoroscopic guidance alone to place the PRGJ tube is aunique benefit because other techniques require concurrent laparotomy, laparoscopy, or endoscopy in addition to fluoroscopy.17-21Recent reports in people have documented advantages of percutaneous radiologic placement of gastrojejunostomy tubes as performed in our canine study, including a very high success rate of appropriate GJT placement, easier placement of the jejunal component, and reduced procedural time.13,14,24
Dogs with severe, ongoing vomiting or regurgitation, gastric dysmotility, pancreatitis, oral or esophageal disease, or concurrent aspiration pneumonia are good candidates for a GJT. The PRGJ tube placement technique bypasses the proximal gastrointestinal tract and uses a distal post-pyloric feeding technique, which has been shown to reduce gastric volumes and minimize aspiration16.
Inability to successfully place the jejunal component of the tube has been reported in up to 32% of people undergoing endoscopic GJT placement.24In contrast, failedjejunal placement using radiographic guidance as used in this report is as low as 2 – 7%.14 When placed under radiologic guidance, the needle can be angled more directly toward the duodenum, resulting more consistent placement of the jejunal component.13,14,24Although evaluation was limited to 6 dogs in this report, the consistently successful placement of both the gastric and jejunal components are in accord with these human reports.
Although not critically evaluated in this study, suture anchors are designed to improve apposition of body wall and stomach during and immediately after tube placement. Their use may reduce the risk of infection and enhance stoma tract formation. There were no complications associated with the suture anchors. The suture anchors typically remain in place for approximately 3 days. They are designed to assist in initial gastrostomy site stability rather thanlong-term ostomy maintenance. They can be manually removed after several days, or they may be naturally released when the 3-0 suture between the intraluminal “T” bar and external button dissolves or breaks. Serial dilation by peel-away introducer placement applies significant inward pressure on the body wall and gastric wall. Albeit short-lived, the gastrointestinal suture anchors appeared effective in re-enforcing the stoma site initially since no dogs appeared to developpneumoperitoneum or signs of focal or diffuse peritonitis. In addition, had the gastrointestinal suture anchors failed, placement of the peel-away introducer would have been impossible. The percutaneous serial coaxial dilator system for placement of the peel-away sheath introducer was used successfully in all cases to establish the ostomy site. Traditionally, endoscopic-assisted gastrostomy tubes are placed by a “pull” technique that involves passage of the tube through the oral cavity during the procedure.25Human studies suggestlowerperistomal infection rates when the “introducer” technique is used.26-28 This is speculated to be due to the complete avoidance of the oral cavity.27,28Peristomal infection was diagnosed in only 1 dog in our report; however, mild to moderate peristomal inflammation was observed in all dogs. In general, peristomal inflammation is the most common complication associated with percutaneous feeding tubes and is considered a minor, often expected, complication in both dogs and people.17,26,29,30 The inflammation observed in dogs in this report may be attributable to the technique itself, as a self-limiting local inflammatory response has been well described in people.26,27