Endoscopic closure of a tracheal gastric fistula using an amflatzer septal occluding device. A 71-year-old woman with a history of esophageal cancer status post esophagectomy with gastric pull-up 6 years prior presents with wheezing and coughing after swallowing. A barium swallow demonstrated the presence of a tracheal gastric fistula at the level of the anastomosis with barium freely passing into the trachea and producing a barium bronchogram. Here is the esophagram showing contrast going through the fistula into the bronchus. Given the size and location of the fistula, it was thought that endoscopic closure or stenting would be ineffective and poorly tolerated. We planned instead the placement of an atrial septal defect occlusion device with the assistance of our interventional radiologists. The fully covered amflatzer device has a catheter delivery system with a sheath. The inner coil attaches to the end of this device, which is shaped like a flattened lumen-opposing metal stent. The stent is withdrawn into the sheath and then the sheath is advanced through the atrial septal defect under fluoroscopic guidance when used in the heart. The distal aspect of the device is then deployed and the deployed flange is withdrawn up to the wall of the defect and then the inner flange is released. Our plan was to place this device across the esophageal fistula into the trachea, deploy one side under fluoroscopic and endoscopic guidance, and then withdraw the distal flange against the trachea and deploy the rest of it into the esophagus. We chose a 14mm wide device with a very short space between the flanges. The procedure was performed under general anesthesia and we first used a 5mm gastroscope through the endotracheal tube to interrogate the bronchus and aspirate pus. We next used the standard gastroscope to examine the stomach looking for a fistula. We saw the gastropharyngeal anastomosis nicely, but was not able to see an opening. We even performed retroflexion, but this was to no avail. We then passed a duodenoscope and we identified what looked like a diverticular outpatching. As we probed further, the orifice was more apparent and we used a guide wire through a catheter to enter the trachea. As the catheter manipulated the orifice, it became apparent that the hole was quite large. We left a guide wire in place and then fed the amplator delivery device over the wire as you can see here. We deployed the distal end into the trachea and used endoscopic as well as fluoroscopic control to confirm placement. Here you see the tracheal side of the occlusion device which is quite bulbous. We then utilized an endoscope to assist with the deployment of the esophageal flange which you can see here. Because the tracheal side was so plump, we used the thin gastroscope to pat it down within the trachea. This was successful as you can see on the final view, it looks quite nice. We did CAT scans following day to show that the tracheal device was in place. After the procedure, the patient was monitored in the ICU and had no pain or dysphagia. There were no further episodes of aspiration. A modified barium swallow subsequently demonstrated no evidence of leakage into the trachea. Her pneumonia symptoms resolved and the patient was discharged after 4 days. Here is the barium swallow demonstrating contrast going across the region of the gastric anastomosis without leakage. In summary, tracheogastric fistula is a rare complication following esophagectomy for esophageal cancer. Standard endoscopic closure methods may be ineffective if the fistula is near the pharyngeal gastric anastomosis due to its proximal location. Amplature cardiac atrial septal occlusion devices offer a safe method of endoscopic closure without the need for open surgery. Thank you.

tracheogastric fistula

Amplatzer septal occlusion

esophagectomy

endoscopic closure

barium swallow