Is blind naso-gastric tube insertion safe after sleeve gastrectomy? A porcine model

Thomas Fabian, MD, Lauren Naile, BA, Mara McErlean, MD. Albany Medical College.

INTRODUCTION: Sleeve gastrectomy is an effective treatment to help morbidly obese patients lose weight. The major technical risk related to this procedure is staple line dehiscence. Naso-gastric tubes (NGT) are placed postoperatively reduce the risk of distension and usually are removed without complication. However, some patients who initially do well later decompensate and need a NGT replaced. Some surgeons are reluctant to place the NGT blindly due to the perceived risk of damage to the staple line from blind instrumentation of the stomach. We sought to determine whether such concern was warranted.

METHODS: A thawed, frozen porcine tissue model (Animal Technologies, Inc, Tyler, TX) composed of esophagus, stomach, respiratory tract, heart and a small portion of liver was used. We performed a sleeve gastrectomy on the greater curvature of the stomach from pylorus to fundus using a flexible gastroscope as a guide for the Endo GIA staples (Covidien, New Haven, CT) in an identical fashion as we use in our patients. The suture line was checked for leaks by immersing the specimen in water while insufflating with a gastroscope. The specimen was then placed in a plastic model of the thorax to mimic anatomical constraint of the body cavities and to ensure proper alignment of the specimen (VATS Trainers, LLC. Lansing, MI). The total length of the specimen was determined by passing a 16 gauge NGT to the pylorus (55 cm) and the pylorus was then cross-clamped. The NGT then was blindly advanced to 55cm for a total of 50 passes. Endoscopy with air distension was performed to evaluate for mucosal injury following the 5th, 15th, 25th, 35th, and 50th attempts. The specimen again was submerged to check for leaks after the 25th and 50th attempts. The NGT was then advanced to 75cm to simulate excessive advancement in the same tissue model. This was done another 50 times. The specimen was examined for injury with endoscopy and a leak test was performed in the same manner as before.

RESULTS: A single 3mm injury to the gastric mucosa was observed by endoscopy after the first 5 attempts to a length of 55cm. No additional injuries were observed after the 45 remaining 55 cm passes nor were leaks or perforations observed after 50 passages. Following 50 passes to 75 cm, multiple small, superficial petechiae were observed on the gastric mucosa, similar to but smaller than the 3 mm injury identified earlier. None were full thickness or penetrated the mucosa. Although some were located near the staple line, the staple line itself showed no evidence of trauma. The model was re-tested with underwater submersion and no leak was found. A subsequent gastrotomy demonstrated a pristine staple line.

CONCLUSION: In this porcine model, blind NGT placement was not associated with significant mucosal injury or damage to the sleeve gastrectomy staple line.

View Poster

« Return to SAGES 2014 abstract archive