Failed EUS hepaticogastrostomy, don't give up. EUS rendezvous hepaticogastrostomy can rescue you. The two main reasons for a failed EUSHE are, one, inability to access the bile ducts to perform a cholangiogram or thread a guide wire, and this is mainly due to non-dilated intrahepatic ducts. And two, the inability to dilate the tract or pass the stent, which is often due to cirrhosis or ductal scarring precluding dilation and sometimes from misaligned vector forces precluding stent advancement. The solutions for non-dilated ducts could be for IR to outline the ducts and inflate a balloon to target. For failure to dilate the tract sufficiently to pass a stent, rendezvous with IR allowing percutaneous traction or rendezvous with yourself allowing traction through the mouth could address the problem. We present five cases that failed traditional EUS 8G treated successfully with EUS rendezvous HG. Case 1, a 53-year-old female patient with multiple surgical necrosectomies, Whipple in 2002 for necrotizing pancreatitis including enteroenterostomies, presented with recurrent cholangitis and leakage around two longstanding PTBDs due to an associated afferent limb syndrome. An EUS 8G into segment 2 failed to resolve the leak around the PTBD arising from segment 3, which was not easily identifiable at EUS. Contrast injection alone did not help localizing this segment on EUS. So a 9mm IR balloon pass percutaneously into the leaking segment was inflated with contrast. This was then identified with EUS but in a difficult scope position. A 19-gauge needle was used to puncture the balloon and gain access to this leaking segment. And a new guide wire passed into this segment was grasped by IR and removed percutaneously. Scope position was very challenging preventing dilation via EUS scope despite the wire being secured externally. So a retrograde dilation using a 4-7 dilating catheter was performed. A new 8G was then created using a 10x8 fully covered biliary stent to cover the leaking segment resulting in eventual resolution of the leak. Case 2. A 55-year-old male patient with multiple medical comorbidities and a prior Whipple procedure for chronic pancreatitis, partial right hepatectomy, developed a chronic dehiscence of his hepatic or jejunostomy resulting in a perihepatic abscess requiring a percutaneous drain. The afferent limb was completely inaccessible from severe adhesions and he was a non-surgical candidate. Having failed every other attempt to close this dehiscence, an attempt was made for an EUS 8G to divert the bile away from the leak site. Conventional EUS 8G failed due to non-dilated ducts as did an attempt to inflate an IR balloon which was too deep and not visualized with EUS. So he underwent a complex IR to IR rendezvous. A peripheral branch of the left intrahepatic duct closest to the gastric wall was accessed via the drain at the dehiscence of the hepatic or jejunostomy. This was dilated with a 4mm balloon allowing opening of a loop snare in this cephalad branch. A transhepatic percutaneous needle puncture of this left intrahepatic duct was then performed through the loop of the loop snare. With a subsequent puncture of the gastric lumen allowing placement of a 0.035 inch endoscopic guide wire into the gastric lumen. So the path of this new guide wire was via a new percutaneous access through the left lobe through the left intrahepatic duct into the gastric lumen. A snare was used to grasp this wire. The looping of the guide wire in the stomach had created a knot. So the wire and the knot were pulled out of the mouth and this was undone outside the patient. This guide wire that was passing through the loop snare placed via the existing percutaneous drain via the hepatic or jejunal dehiscence site to the left intrahepatic duct was pulled out percutaneously creating the rendezvous. This allowed a 4mm balloon dilation of the tract under tension and placement of a 10mm by 8cm fully covered biliary SEMS. Contrast injection showed the stent in good position. Case 3. A 36-year-old male patient with prior severe pancreatitis, benign biliary stricture, gastric outlet obstruction, severe portal hypertension and compensated cirrhosis had undergone a surgical gastrojejunostomy and hepaticoduodenostomy. He presented with recurrent succotash cholangitis from food occluding his bile duct at the level of the hepaticoduodenostomy just proximal to the duodenal stricture despite a patent surgical gastrojejunostomy. He had failed double pigtail biliary stents to keep the food out. So he was taken for an EUS 8G to provide a more proximal site of diversion of his biliary tree from his surgical GJ. At EUS, a branch of the left intrahepatic duct is identified but the liver capsule from cirrhosis is hard to puncture but eventually was successful. After this, multiple firm passes are made to try and penetrate the fibrotic bile duct and finally the fourth attempt allows access and a cholangiogram. A guide wire is passed downstream via the hepaticoduodenostomy into the duodenum. Despite using EUS, multiple attempts at dilating with a 4mm balloon only allowed dilation of the liver parenchyma and not the penetration site of the bile duct. A 4-7 French dilating catheter, a Sohindra stent retrieval device, and even a 3-4-5 taper tip catheter failed to traverse the bile duct. So a decision to perform an endo-EUS rendezvous is made by switching to a therapeutic upper endoscope with a cap. During the rendezvous, the guide wire had fallen back into the common hepatic duct and had to be grasped with a rat-toothed forcep under fluoroscopy. Although passage across the fibrotic bile duct is difficult, with tension from the rendezvous wire we are successful, allowing a 4mm dilation. One can see the waste on the balloon. Pulling the scope back to the gastric cardia allows you to visualize the dilation of the gastric wall as well. A 10mm by 8cm fully covered stent is then passed via the hepaticoduodenostomy. The scope is then pulled back to visualize the stent across the gastric wall. And the stent is then deployed. Contrast injection confirms the stent is in good position. A 7 French by 15cm pigtail stent is then similarly deployed. Five patients were treated using the three techniques described, all having failed prior conventional EUS 8G. All were technically successful, but one patient failed to resolve his leak. There were no adverse events over a median follow-up of 28 months. So, if one is unable to access the biliary tree due to non-dilated intraepatic ducts, one can successfully rendezvous with IR by inflating a target balloon. Difficult vector forces or scurrous bile ducts that preclude dilation or stent passage can be similarly rendezvoused with IR or with oneself, allowing traction and successful stent placement. In conclusion, when conventional EUS 8G fails, don't give up. Try a rendezvous with IR or oneself in the same setting or separate procedure to successfully allow internal drainage rather than a long-term PTBD.

EUS rendezvous hepaticogastrostomy

failed EUS hepaticogastrostomy

bile duct access

tract dilation

stent placement