This ASG video tip is brought to you by an educational grant from Braintree, a part of Cibela Pharmaceuticals, makers of SUTAB. My name is Dr. Hu, and I will be presenting our video titled, Holy Moly, Comprehensive Management of an Eroded Aortic Prosthetic Graft. These are our disclosures. The patient is a 45-year-old male with a history of intravenous drug abuse who presented with a ruptured mycotic aortic aneurysm and immediately underwent thoracic and vascular aortic repair. About two months later, he presented to the emergency department with fevers. CT angiography showed fluid and several air locules around his aortic graft. However, there was no visible extraluminal extravasation to suggest a leak on his swallow study. He was taken to the operating room by the vascular surgery team for planned explant of his aortic graft with the cardiac surgery team on standby. He underwent upper endoscopy by the thoracic surgery team and was found to have a defect in the esophagus. This was located 31 centimeters from the incisor on the posterior lateral wall of the esophagus and measured 5 millimeters wide. As you can see, the graft is visible through this defect. Purulent fluid could be seen emanating out of this defect on suction. There were also portions of necrotic mediastinal fat entering the lumen. The planned explant of the aortic graft was aborted, and surgical endoscopy was consulted intraoperatively at this point for management of this defect. A multidisciplinary discussion on next steps was held with a total of four surgical services. The defect was small enough to potentially be closed endoscopically, but this would orphan a substantial amount of purulence around the aortic graft, which would have nowhere to go. Given the heavy amount of mediastinal contamination, we elected to proceed with serial endoluminal vacuum therapy and reassess for progress over the course of time. The patient underwent a total of five rounds of endoscopic debridement of the abscess cavity and endoluminal vacuum dressing changes. These served to temporize his infection by clearing the gross contamination and allowed us to design an acceptable staged management plan. During this time, he had a percutaneous gastrosomy tube placed for distal feeding access. The decision was made to take the patient back to the operating room for a two-stage procedure to address this esophageal defect. The first stage would involve endoscopic-assisted, CT-guided placement of a percutaneous drain into the infected aortic aneurysm sac, between the sac and the actual graft itself. This would allow external drainage posteriorly and closure of the esophageal defect on the medial aspect. The second stage would involve endoscopic esophageal defect closure. The first stage was conducted in the CT scanner for simultaneous CT imaging and endoscopy using protocols already established in the literature. A scope was placed through the oropharynx into the esophagus, and we could see that the defect was now smaller in size. Our initial views of the CT scan scalp film found minimal to no air around the aortic sac around the graft, and there was no target for our interventional radiologist to place the drain. The diagnostic scope was then removed and an ultraslim endoscope was advanced into the aorta. The aortic sac around the graft was insufflated, which allowed visual inspection of the aorta. There was minimal contamination of the defect itself and no large debris, suggesting that endoluminal vacuum therapy had been successful in removing the gross debris from this area. Working in this position through a series of CT fluoroscopic images, our interventional radiologist was able to navigate a wire into the space between the graft and the aorta, avoiding the intervening lung and pleural space, and avoiding damage to the aortic graft. Our interventional radiologist was then able to place an 8 French drain into the space with no bleeding in the aneurysm sac. Fluid was irrigated into the drain and was seen draining into the lumen, suggesting that the drain was in an adequate position. The drain itself was not able to be visualized by the endoscope, as the endoscope was on the contralateral side of the aorta. Two days later, he was taken to the operating room for closure of the esophageal defect. The decision was made to place a helical TAC suturing system in a figure-of-eight fashion. We started distal anterior, then placed a TAC distal posterior, then placed a proximal anterior, and then placed a TAC proximal superior, creating a figure-of-eight suture. Slack was taken out of these sutures in between each bite. We then used a suture cinch tension on the suture and aggressive suction to close the defect. The cinch was fired. Lumen patency was checked with a scope and was confirmed to be patent and not narrowed. The defect was no longer endoscopically visible. The patient had no immediate postoperative complications, and at his three-month postoperative visit, he was clinically doing well with no symptoms or complaints. He underwent a routine postoperative swallow study, which demonstrated an esophageal wall defect that appeared to abut his aortic graft. There was no free contrast leaking around the aortic graft, but the findings were concerning for a possible ongoing aortoesophageal fistula. The decision was made to evaluate him endoscopically with the possibility of another endoscopic closure of this defect. On endoscopy, the old helical TAC suturing system was identified and was still affixed to the mucosa. These TACs were in the way of our over-the-scope clip placement, so we removed them. Utilizing a 3D scope hemo clip and our novel removal technique, each of the four TACs were unscrewed from the mucosa. The entire TAC system and suture cinch were removed, and the system appeared to be intact. A small punctate opening was identified in the lateral esophageal wall and was injected with contrast, utilizing an ERCP sphincteratome. There was no undrained cavity, and most of the contrast flowed back into the esophageal lumen. The decision was made to endoscopically close the esophageal wall defect using a slim upper scope with a small over-the-scope clip. We felt safe using the clip because there was more soft tissue to grasp. Fluoroscopic images weren't used to assure that the clip was not located near the aortic graft, and the clip was fired. It appeared to be in good position endoscopically. Again, there were no immediate post-operative complications. A routine post-operative swallow study showed a possible small but persistent aorto-esophageal fistula. The patient clinically was doing well and had no concerns or complaints. On his most recent endoscopy, there was no evidence of ongoing communication between the esophageal lumen and the lumen of the native aorta, and there was no defect in the mucosa. There was no evidence of exposed graft endoscopically or fluoroscopically. As you can see, the clip rests on a very long stalk of mucosa and was felt to be too difficult to remove. The decision was made to place a loop ligature around the mucosal stalk to allow it to autonecrose and pass spontaneously. This case demonstrates a few key points. The standard of care in this patient's case is a graft explant, however, this was not a good option for this patient at that point per his primary surgical team. His operation would have been extensive, and he would have required an extra anatomic bypass as well as an esophageal repair. As such, surgical endoscopy was consulted intraoperatively and asked to attempt endoscopic management of this infection and esophageal defect. His aortic graft currently remains in place, which does put him at risk of chronic graft infection problems. Clinically, however, he has done very well, and his aortoesophageal fistula defect has closed with endoscopic management. Our comprehensive endoscopic management, in addition to the help from our interventional radiology colleagues, was safe, effective at controlling his infection, and bought us time to develop a staged approach for his infection and esophageal defect. Another key point is the importance of being familiar with a wide variety of endoscopic defect closure tools. In this case, we first used a helical taxutrine system, followed by an over-the-scope clip. Lastly, this case highlights the superior patient care that was delivered by a multidisciplinary collaboration between advanced surgical endoscopy, vascular surgery, and interventional radiology. For more information visit www.FEMA.gov

Aortic Prosthetic Graft

Eroded Aortic Graft

Mycotic Aortic Aneurysm

Endoscopic Management

Aortoesophageal Fistula