Endoscopy 2003; 35(3): 246-247
DOI: 10.1055/s-2003-37261
Letter to the Editor

© Georg Thieme Verlag Stuttgart · New York

Conservative vs. Endoscopic Closure of Colocutaneous Fistulas after Percutaneous Endoscopic Gastrostomy Complications

T.  Gyökeres 1 , M.  Burai 1 , J.  Hamvas 1 , M.  Varsányi 1 , M.  Mácsai 2 , L.  Paput 2 , Á.  Köveskuti 3 , C.  Fekete 3 , Á.  Pap 1
  • 1Department of Gastroenterology, MÁV Hospital, Budapest, Hungary
  • 2Department of Otorhinolaryngology, MÁV Hospital, Budapest, Hungary
  • 3Department of Surgery, MÁV Hospital, Budapest, Hungary
Further Information

Publication History

Publication Date:
13 February 2003 (online)

We read with interest the paper by Kim et al. [1], presenting an ingenious idea for using a metallic clip to close a colocutaneous fistula after placement of a percutaneous endoscopic gastrostomy (PEG) tube. We would query whether it was really necessary to close the fistula interventionally. In the case reported, early dislodgment of the PEG tube occurred at 2 weeks. The time at which the colonoscopy was carried out is not clear from the paper. Evidently, a 10-day period with total parenteral nutrition and systemic antibiotic treatment elapsed after the patient had vomited the feculent material. It can be assumed that endoscopic removal of the PEG tube and endoscopic closure of the fistula were carried out after that period. Was it really necessary to use a sophisticated endoscopic technique after a 10-day delay, instead of the familiar conservative solution?

Last year, a 47-year-old man with an inoperable mesopharyngeal tumor was referred to us for PEG placement due to a locally advanced neoplasm causing swallowing difficulty. The PEG tube was placed by a relatively inexperienced team. After 6 weeks, the swallowing disturbances disappeared due to successful radiotherapy, and removal of the PEG tube was planned. A brown, malodorous fluid was observed flowing from the side of the PEG. Gastroscopy did not reveal the internal bumper in the stomach, but the former site of the gastrostomy was identified. Omnipaque was then administered via the PEG tube. The contrast material initially appeared in the transverse colon (Figure [1]), after which the gastric folds began to appear under fluoroscopy (Figure [2]). The outer part of the tube was then pulled gently and cut at the level of the skin. Mild finger pressure was placed on the former PEG site to push the inner part of the tube into the lumen. Passage of the bumper was not observed by the patient, but a subsequent barium examination did not detect it 4 days later. Although spontaneous closure of both the gastrocolic and cutaneous portions of the fistulous tract usually happens within a matter of hours [2], in some cases a cologastric fistula may persist for a long time [3]. In this patient, the cutaneous fistula healed within 1 week, but the cologastric fistula was demonstrated by barium enema as long as 5 weeks later, without any clinical consequences. The patient died due to massive bleeding from the locally advanced mesopharyngeal tumor 5 months after PEG placement, with no signs of an open fistula.

Figure 1 The contrast medium administered via the percutaneous endoscopic gastrostomy tube (arrow) first appeared in the transverse colon.

Figure 2 Afterwards, the gastric folds (arrow) began to appear.

In our opinion, Kim's method can be considered for primary use not for the closure of colocutaneous fistulas after dislodgement of a PEG tube, but for cases of colocutaneous fistulas of other long-lasting etiologies - e. g., colonic complications of necrotizing pancreatitis. On the other hand, persistent gastrocolic fistulas with relevant clinical symptoms after PEG placement can indeed be closed by metallic clipping.


T. Gyökeres, M.D.

Dept. of Gastroenterology · MÁV Hospital

Podmaniczky 111 · 1062 Budapest · Hungary

Fax: + 36-1-4752669

Email: [email protected]