Management of refractory gastrointestinal fistulas using cardiac septal occluded device, a case series. Gastrointestinal fistulas are abnormal epithelialized communication between two or more GI lumens or organs. They are associated with high rate of morbidity and mortality. Most GI fistulas will close spontaneously, however some may remain open, which can be closed by a variety of conventional endoscopic interventions. Techniques for robust fistula closure include organ-plasma coagulation, APC, with endoscopic suturing. APC of tissue surrounding the fistula promotes de-epithelialization for better tissue opposition and fistula closure. Here we see organ-plasma coagulation with over-the-scope suturing for the closure of a defect in the GI tract. And here we see APC with endoscopic suturing using a through-the-scope suturing device. Another option for fistula closure is over-the-scope clip, which is used for full technique closure of smaller fistulas less than two centimeters. While these interventions can be effective, there's often recurrence of the fistula due to epithelialization of fistula lining, fibrotic alteration, and larger chronic fistulas. Cardiac septal occluders, CSOs, offer an alternative management option when these conventional endoscopic interventions have failed. CSOs shown on the left are self-expanding double-dixed device. They are made from nitinol and interwoven polyester that promote tissue engrowth and occlusion. They are FDA approved to treat atrial and ventricular septal defects, but can be used off-label for the management of GI fistulas. CSOs come in two types, a VSD closure device shown on the left and the ASD device shown on the right. They both come in different sizes and can both be used to manage GI fistulas. Choosing the right size and type of CSO is important because the waist diameter for both types determines the maximum length the CSO can include. For ASD CSO, the waist diameter should be equal to or larger than the fistula diameter, while for the VSD CSO it should be larger. Furthermore, in cases where the fistula has a longer length, a VSD CSO may be a better option given its longer waist length. Lastly, CSO can easily be recaptured, redeployed, and stretched for ideal placement. The CSO delivery system as a whole has a maximum length of 80 cm, which makes it incompatible with most endoscopes. To circumvent this, the CSO can be separated from the delivery system and loaded onto an adapter endoscopic biliary catheter to allow for sufficient length for deployment through an endoscope. To prepare the device, we took a 202 cm 10 French biliary pushing catheter and removed its introducer. The catheter was measured alongside the 160 cm pediatric biopsy forceps and cut to reduce its length to approximately 155 cm to match the length of the pediatric biopsy forceps. The pediatric biopsy forceps was then loaded into the biliary catheter and was used to grasp the CSO device, which we can see a close-up view here. We present six cases of successful management of GI fistulas throughout the GI tract using CSOs in patients that have failed other interventions. Mean age and fistula size were 61 and 10 mm, respectively. We present a case of refractory rectal pelvic fistula. A 63-year-old female with a history of rectal adenocarcinoma, status post low anterior colon resection with a coloaneoanastomosis, was referred for a recurrent leak in 5 mm rectal pelvic fistula. Failed closure attempts are listed below. Many of these attempts can be seen here with the use of APC, endoscopic suturing, and endoscopic clips. Despite those interventions, the leak persisted, as shown here, so a decision was made to pursue fistula closure with a CSO device. After preparation of the device, as previously described, the endoscope was advanced through the anus and the 5 mm fistula was identified. A guide wire was passed through the fistula, followed by an extraction balloon catheter, which is used to estimate the size of the fistula and to ensure easy traversability of the fistula prior to deployment of the CSO. The CSO device loaded onto the 10th French catheter was then deployed across the fistula, with the proximal phalange passed into the fistula, and the visualization of the distal phalange can be seen, completely occluding the fistula's tract. The patient was stable post-CSO placement and discharged the next day without any further rectal drainage. Fluoroscopy with contrast injection at 4 months follow-up showed no contrast extravasation, with an intact CSO indicated by the arrow. Colonoscopy at 4 and 23 months follow-up showed significant tissue engrowth without any purulent and rectal drainage. Represent a case of a gastrocutaneous fistula. A 46-year-old male with a history of severe acute necrotizing pancreatitis requiring PEG-2 placement for nutritional support was referred for evaluation for a CSO placement for a persistently leaking 10 mm gastrocutaneous fistula after PEG removal. Failed fistula closure attempts are listed here. Multiple attempts were made to close the fistula include endoscopic suturing and over-the-scope CLIP. These attempts ultimately failed with continued gastric leakage. Given this, a decision was made to close the fistula with a CSO. First, the previously placed sutures around the fistula tract were removed. Then the APC was introduced into the fistula extracutaneously, coagulating the tract as it entered the stomach. Once in the stomach, the APC was used on the gastric opening of the fistula. Then the CSO was introduced into the fistula from the stomach. We can see both the proximal phalange open in the stomach as well as the distal phalange open to the extracutaneous space. Three months post CSO deployment, there was significant tissue engrowth into the CSO and a reported resolution of his gastrocutaneous leakage. Similarly, for case 3, a CSO was placed to treat a persistent 5mm esophageal defect causing a leak after surgical repair of an esophageal perforation in a patient with a history of esophageal stricture requiring multiple dilations. At three months follow-up, he continued to tolerate a solid diet and remained asymptomatic without a leak seen on barium swallow study. For case 4, CSO was deployed for management of a persistent 5mm hepatical jejunostomy in a stomatic fistula associated with abdominal abscess requiring IR drain placement. At three months follow-up, tissue engrowth can be seen and patient was asymptomatic. For case 5, CSO was placed for the management of a persistent 30mm gastro-gastric fistula between the gastric pouch and the remnant stomach in a patient with a history of real and wide gastric bypass who had issues with recurrent weight gain. At three months follow-up, she was asymptomatic and had lost 18 pounds. For case 6, CSO was placed to treat a persistent 5mm jejunocutaneous fistula after patch tube removal. At one month follow-up, patient reported no patch site leakage. These tables give a summary of the six patients including their failed interventions. There were no adverse events. CSOs induce occlusion and promote tissue engrowth for fistula closure. They are safe and effective in the management of GI fistulas and anastomotic leaks. Our case series demonstrate the successful use of CSOs in closure of refractory fistulas that have failed traditional endoscopic and surgical intervention throughout the GI tract. Current conventional endoscopic closure devices requires apposition of fistulas, tract, walls, and can fail to close chronic, large, or fibrotic fistulas. In such cases, fistula tract occlusion with promotion of tissue engrowth with CSO can result in long-term fistula closure. Large studies with long-term follow-up are needed to validate the effectiveness and lasting effect of CSOs.

cardiac septal occluder devices

gastrointestinal fistulas

endoscopic treatments

tissue ingrowth

long-term benefits