Laparoscopic cholecystectomy remains the standard of care for acute cholecystitis. Bioleaks are a known complication of cholecystectomy that occur at a rate of 0.4 to 2%. The majority of bioleaks are treated with a biliary stent in the common bile duct. However, up to 9% of bioleaks can be refractory to plastic stent placement. We present two cases where refractory bioleaks were treated with endovascular coils. Case 1, an 83-year-old female with a history of coronary disease, status post-CABG, Parkinson's disease, and type 2 diabetes presents with a worsening of the abdominal pain. A CT of the pelvis shows findings consistent with acute cholecystitis. Due to significant inflammation, the patient eventually underwent subtotal cholecystectomy with drain placement. The cystic duct was left open, and unfortunately, the drain output remained bilious. A 10 French plastic stent is placed to treat the leak. Unfortunately, the patient's surgical drain remains bilious four weeks after stenting. At this time, the patient had a bile leak which was refracted to a 10 French plastic stent. The endoscopist considered placement of a covered metal stent. Traditionally, bile leaks are treated with a stent across the major papilla, which creates a pressure gradient and encourages preferential drainage of bile into the duodenum. Although a covered metal stent may increase this pressure gradient, we do not feel that this would be sufficient given the patient's aberrant anatomy of a high cystic duct takeoff and a high-grade leak. We therefore chose to increase resistance in the cystic duct in an effort to control the leak. The endoscopist considered placement of a vascular plug, sclerosing agent of the cystic duct, and coiling of the cystic duct. Vascular coils have several advantages, including being able to be deployed through a catheter, a slim profile which allows for maneuverability through a tortuous cystic duct, and being pliable so that once they are deployed, they take the shape of the target they are in and do not compress adjacent structures, such as the common hepatic or common bile duct. We did identify prior studies from interventional radiology where vascular coils had been used. In these studies, patients had had a cystic duct stump leak, which was complicated by a biloma. The biloma was treated with a percutaneous drain, and once the biloma had resolved, the interventional radiologist used a percutaneous route to cannulate the cystic duct and deploy vascular coils. Therefore, these studies showed that percutaneous transcathetamolization of the cystic duct is feasible, that coiling successfully treated cystic duct leaks that proved refractory to traditional methods, and that this treatment option was robust, and that patients did not have a recurrence of bile leak. Building off this, we attempted to identify a way of deploying vascular coils endoscopically. We will now present the next few demonstrations of vascular coil deployment. We initially chose a catheter that was slim enough to be able to cannulate the cystic duct, but large enough to accommodate a vascular coil. We initially chose a 3 French to 5 French catheter, which could accommodate a 0.018 inch wire. This was loaded with a 0.018 inch vascular coil. Unfortunately, this combination did not work because the coil prematurely formed in the catheter, and a guide wire could not be used to advance the coil. In our experience, a 5.5 French catheter loaded with a 0.035 inch vascular coil is the optimal combination. We use a 0.035 inch guide wire to push the vascular coil with excellent one-to-one response. With this combination, the coil does not prematurely form in the catheter, and the guide wire is long enough to fully deploy the coil within the target. We now return to our case. The inclusion cholangiogram shows filling of the left-front intra-hepatic ducts, takeoff of the cystic duct, and a bile leak. The cystic duct is identified on cholangioscopy and cannulated with a 0.25 inch guide wire. Long exchange is performed, leaving the guide wire in place. With the wire in place, a 5.5 French catheter, which can accommodate a 0.035 inch vascular coil, is advanced into the cystic duct. Once an appropriate combination of catheter, coil, and wire as a pusher is obtained, three coils are deployed in the cystic duct. At this time, the endoscopist chooses to re-evaluate the common bile duct with cholangioscopy to confirm that the coils are in the cystic duct and not in the common bile duct. During cholangioscopy, the bifurcation is near the distance, and the cystic duct takeoff is appreciated in the bottom right of the image. Although there is inflammation, the vascular coils are not protruding from the cystic duct orifice. A plastic stent is also placed in the common bile duct for preferential drainage into the duodenum. Prior to coiling, the patient's surgical drain had to be emptied three times a day. Two weeks after coiling, the surgical drain had minimal output and was pulled. Three months after initial coiling, the patient underwent repeat ear CP for plastic stent removal. A lunoclosion cholangiogram shows appropriate filling of the intra-hepatics and common bile duct without any active extravasation. A final fluoroscopic image shows contrast draining from the common bile duct and that the coils are in appropriate position. The patient continues to do well six months after initial coiling, all external genes have been removed, and the patient has been discharged home. Case 2. A 58-year-old man with a history of compensated alcohol-induced cirrhosis undergoes laparoscopic colostasectomy at an outside institution. Unfortunately, his case is complicated by a cystic duct stump leak. Initially, the fluid seen on post-operative imaging was thought to be a site of secondary cirrhosis rather than a bile leak. He did not undergo any intervention for this and was transferred to our facility one month post-operatively. Initial CAT scans in our institution show a large fluid collection in the abdomen, thought to be a biloma with a mature wall. Our first intervention was near CP, where a bile leak was noted off the cystic duct stump. To treat this leak, a 10-millimeter by 8-centimeter covered metal stent was placed in the common bile duct. Next, the biloma was assessed on US exam, and a 15-millimeter by 15-millimeter luminoposy metal stent was placed into the collection to promote drainage. Although the biloma partially resolved, the patient had a persistent leak from the cystic duct. The persistent leak was due to the significant delay between the initial surgery and the endoscopic intervention on the leak. The patient's cystic duct leak appeared to be refractory to covered metal stent and plastic stent placement. Therefore, we elected to attempt control of the leak with cystic duct coiling. Initial scalp film shows a purgatory strain in appropriate position. The common bile duct is cannulated, and the intrapadics are appreciated. The cystic duct stump leak is once again appreciated on balloon occlusion cholangiogram. A 5.5 French catheter is advanced into the cystic duct, and then three 0.035-inch 20-millimeter coils are advanced and deployed into the target. Repeat cholangiogram does show a persistent leak after coil deployment. This is highlighted by the arrow. Therefore, many forceps are advanced into the duct and used to push the coils together in an effort to help control the cystic duct stump leak. A repeat cholangiogram shows control of the leak. A plastic stent is placed for preferential drainage of bile into the duodenum. The patient did well after procedure. All external drains were removed, and a follow-up CT scan showed resolution of the biloma. These cases highlight the feasibility of intraductal coiling of the cystic duct in the setting of a refractory bile leak. This method of treatment may have a role in chronic leaks or for patients with aberrant anatomy.

laparoscopic cholecystectomy

bile leaks

endovascular coils

refractory treatment

biliary stent