Management of Proximal Gastrogastric Fistula Due to Staple Line Dehiscence

This Month’s Featured Expert: Alfons Pomp, MD, FACS, FRCSC
Dr. Pomp is a Leon C. Hirsch Professor, Vice Chairman, Department of Surgery, and Chief, Section of Laparoscopic and Bariatric Surgery, Weill Medical College of Cornell University, New York Presbyterian Hospital, New York, New York.

This Month’s Dilemma: Management of Proximal Gastrogastric Fistula Due to Staple Line Dehiscence

Bariatric Times. 2010;7(10)20

The Dilemma
A 50-year-old woman presented with recurrent episodes of stenosis and large marginal ulcerations post-nondivided gastric bypass for morbid obesity with a proximal gastrogastric fistula due to staple line dehiscence. The patient underwent a remnant gastrectomy complicated by a staple line disruption and a leak. Treatment with stents and drainage failed to control the fistula. Four months after her last surgery, she began parenteral nutrition and had a drain that produced approximately 80cc of saliva a day.
What would you recommend?

Expert Commentary
by Alfons Pomp, MD, FACS, FRCSC

This is a complex case and the surgical management would be technically difficult and underlines the surgeon’s reasoning to attempt to manage this nonoperatively for a prolonged period of time. Adjuncts to current management could include atropine to decrease sialorrhea. There is no evidence in either the literature or from my personal observation that octreotide is beneficial to decrease this type of fistula output.

The best treatment is prevention, and during the inital revisional surgery, every effort should be made to avoid crossing over staple lines during transection. Even when these areas are oversewn, they are a vulnerable site for necrosis and fistula formation.

When this case is brought to the operating room, it should be booked as the first case in a not overly busy schedule. The risk of conversion to an open procedure is significant. Antibiotic prophylaxis is mandatory and oropharyngeal pathogens must be covered. In my practice, we have found that there are many resistant strains to clindamycin and we do not hesitate to use vancomycin in patients who are allergic to cepahalosporin. Because of the certainty of dense adhesions to the liver, the usual self-retracting Nathanson retractors may not be useful and a solid, fan-type 10mm liver retractor is often helpful. Finally, it is best to position the patient with split legs, as the crux of this operation will be the dissection of the hiatus and positioning the surgeon between the legs offers the best access to this anatomical area.
The initial dissection is to identify the anatomy, and although this is not intuitive, it is often best to start along the lesser curve using the caudate lobe to find the right crus. Opening the phreno-esophageal ligament gives some added length to the pouch and frees the esophagus to move caudal. The dissection should finish on the greater curve side, and every attention must be paid to avoid damage to the spleen and tail of the pancreas, as this is likely the site of significant adhesions and inflammation from the fistula (and there is no protection from the remnant stomach, which was previously resected). Once it is decided there is enough pouch proximal to the fistula, the distal pouch/fistula and gastrojejunostomy should be resected. Even if there are only a few millimeters of stomach proximal, this should be conserved as the serosa will allow for a better anastomosis than the esophagus.

I have found that the circular “flip top” 25mm end-to-end anastomosis stapler gives the best chance for a technically sound anastomosis. A 21mm stapler would be much more prone to stricture in this setting. Hand sewn would be my second choice; however, if sutures are placed on the esophagus, meticulous attention must be observed to avoid excessive tension and shearing when tying the knots. If separated sutures are necessary, this is inordinately time consuming. Linear stapling is technically difficult in this area; the small diameter of a pouch this proximal and the angles required even with an articulating stapler are not optimal for the creation of the anastomosis. The anastomosis needs to be tested during the operation. I am most comfortable with methylene blue; however, if endoscopy with air insufflation is used, the anastomosis must be completely submerged, which is not always easy this high in the hiatus.
Although I rarely drain any anastomosis anymore, I would leave a drain here until the anastomosis integrity is verified by radiology. The drain is removed rapidly (<4 days), provided the patient’s condition is propitious.

A feeding tube, either in the small residual remnant (the best option) or in the jejunum, is mandatory to allow for nutritional support enterally during the perioperative period and should remain in place until oral intake is satisfactory and the patient is no longer in a catabolic mode.

Follow up from the treating surgeon on the case presented
In an attempt to revise the area of the leak, the surgeon performed a laparoscopy and found that the pigtail drain that was placed percutaneously and that was draining saliva was eroded into the proximal alimentary limb. The surgeon chose to remove the drain and repair the enterotomy. The case was aborted and the patient was started on physician’s orders diet 24 hours postoperatively after a swallow study could not demonstrate the leak any longer. The patient was discharged home 24 hours later tolerating diet. Follow up in the office 10 days postoperative was uneventful.

Category: Ask the Experts, Past Articles