Anastomotic leaks and strictures after esophagectomy are associated with high morbidity
and mortality [1 ]
[2 ]. We present here a patient who had a subtotal esophagectomy with gastric interposition
because of esophageal adenocarcinoma. Anastomotic leakage occurred and was treated
by endoscopic applications of fibrin glue. One month after discharge she presented
with a large (2.5 cm × 5 cm) iatrogenic esophagotracheal fistula. Because of the extensive
defect in the pars membranacea and the inflammation we decided against surgical intervention
and treated the defect endoscopically using a combination of a Vicryl plug and fibrin
glue, as described previously [3 ]. The tracheal fistula was treated with a covered self-expanding tracheal stent [4 ].
Despite repeated endoscopic dilations, the stenosis recurred due to ischemia. Additional
stenting of the anastomotic stenosis was impossible because there was only a thin
tissue bridge between the pars membranacea and the (neo-)esophagus and because a stent-to-stent
arrangement would most probably have led to further necrosis. The presence of the
nasojejunal tube caused considerable psychological stress and impaired the patient’s
quality of life. To enable her to recommence oral feeding, ascending colonic bypass
surgery with collar side-to-side esophagocolostomy and intra-abdominal side-to-side
colojejunostomy was performed 11 months after the initial operation, leaving the gastric
bypass in situ. The endoscopic control showed a wide anastomosis of the esophagocolostomy,
but a siphon-like reservoir at the stenotic entrance of the gastric interposition
led to regurgitation ([Fig. 1 ], [2 ]). We therefore performed an endoscopic closure of the esophagogastric anatomosis,
with de-epithelialization of the stenotic gastric tube and sealing with bucrylate
and histoacrylate. Control radiography showed adequate efflux without filling of the
sealed gastric tube ([Fig. 3 ]). She was able to resume normal oral feeding and her body weight stabilized. This
endoscopic approach has not been described in the published literature before. It
represents a useful alternative for the treatment of this serious clinical situation.
Fig. 1 A water-soluble contrast study and a schematic illustration key showing the gastric
tube without filling of the colonic bypass (A, tracheal stent; B, esophagus proximal
to the stenotic collar anastomosis; C, the gastric tube; D, location of the colonic
bypass).
Fig. 2 a Upper gastrointestinal endoscopic image showing a sufficiently wide collar anastomosis
of the side-to-side esophagocolostomy. b A stenotic collar anastomosis with a siphon-like anatomical configuration at the
entrance to the gastric interposition.
Fig. 3 A water-soluble contrast study and a schematic illustration key showing the appearance
after endoscopic de-epithelialization of the stenotic gastric tube and sealing with
bucrylate and histoacrylate. An adequate efflux of the contrast medium without filling
of the sealed gastric tube was achieved (A, tracheal stent; B, esophagus proximal
to the stenotic collar anastomosis; D, the colonic bypass).
Endoscopy_UCTN_Code_TTT_1AO_2AI