Endoscopic Clipping of a Post-Whipple Pancreatic Bleed
Tonya Kaltenbach MD, Sherry Wren MD, Lana Schumacher MD and Roy Soetikno MD
Bleeding following whipple surgery most often occurs from the gastro or duodeno jejunal anastomosis. Bleeding at the pancreatic surface would be the most difficult to treat, as it typically requires surgical resection of the pancreatic remnant – a surgery that carries a significant morbidity and mortality.
We present a case of a 69 yo man who had massive upper gastrointestinal bleeding post operative day two following a whipple procedure for cholangiocarcinoma.
He was found at surgery to have a 2 cm cancer that was adherent to the left hepatic artery and portal vein, and underwent a pyloric sparing whipple procedure with partial resections and reconstructions of the vessels.
Two days post operatively, he developed shock with bloody nasogastric tube output.
He was stabilized with massive colloid transfusions, and endoscopy was then performed. Though there is no absolute contraindication to post operative endoscopy, given the fresh anastomosis in this case, we carefully planned the endoscopy in order to minimize the procedure time, endoscope and bowel manipulation and insufflation,
We endoscopically approached this massive upper gastrointestinal bleed by first intubating the patient. We used the therapeutic megachannel endoscope with water jet capabilities. We reviewed the anatomy to consider all of the potential bleeding sources, and we prepared for mechanical hemostasis. We used carbon dioxide in lieu of air for insufflation.
Clots were removed using the 6mm channel therapeutic scope to improve visualization. However, upon further inspection, there was no bleeding at the duodenojejunostomy.
The hepaticojejunstomy was identified. There is no active bleeding at the site of anastomosis or evidence of hemobilia.
The endoscope was changed to a pediatric colonoscope with water jet function in order to reach the site of the pancreaticjejunostomy, where there was fresh blood.
Active bleeding could be seen originating from the cut pancreatic surface. We used reopenable clips in order to have the ability to close, and assess hemostasis, prior to final deployment. Complete hemostasis was achieved following placement of 2 clips.
There was concern for recurrent bleeding 2 days later when the patient required another 4 units of blood transfusion. We elected to re-attempt endoscopy in a last effort to spare the patient from a re-exploration with takedown of the pancreaticojejunostomy.
Repeat endoscopy showed a clot with no active bleeding. One of the clips had been dislodged. We removed the clot using a cold snare.
And identified an exposed nonbleeding visible vessel, as seen here in the endoscopic and schematic image just below the remaining clip.
We chose to clip using a minature clip. This clip is the most difficult to deploy but hosts the strongest properties.
Following repeat clipping, active bleeding occurred. This allowed precision in localization and clipping of the bleeding site for complete hemostasis.
We repeated endoscopy on post operative day 6 for surveillance of the site. All clips were in place and there was no evidence of bleeding. He did well without recurrent bleeding and was discharged home. At six months follow up he had no further bleeding episodes, development of fistulas or complications from his surgery.
Endoscopic treatment of a bleeding site on the pancreatic surface of pancreaticojejunostomy can be successful in the immediate post-operative period after a whipple procedure.
Such attempt a treatment may prevent surgery of completion panecreatectomy.