Placement of Self-Expanding Metal Biliary Stents (SEMS) in Patients with Resectable Pancreatic Cancer

Until recently, placement of SEMS in the setting of pancreatic cancer was felt to be relatively contraindicated due to possible interference with surgery or a perceived increased risk of post-operative infection. However, recent small series have suggested a positive outcome in this setting (Lawrence, GIE 2006;63: 804-807). We report a consecutive combined hospital cohort experience. Patients and Methods: All patients (pts) with biliary SEMS in place who underwent surgery in an attempt to perform a pancreaticoduodenectomy were analyzed from 1/03 - 8/08 at two major regionalized centers. Pts were identified through a prospective database, and medical records were analyzed regarding surgical findings, success, and post-operative course. At the time of ERCP, SEMS were carefully placed to ensure the proximal portion lay below the anticipated line of bile duct resection. To that end, non-foreshortening stents were used in 27 cases and foreshortening stents were used in 5. Results: 32 pts (mean age 68 ± 13yrs [range 37-87];18 males) were identified. Pts were either deemed surgical candidates based on preoperative staging (CT ± EUS) or surgically explored if respectability questionable. Of the 32 total, 24 pts underwent Whipple's resection. 8/32 were unresectable at surgical exploration. Review of operative records demonstrated no evidence that factors related to SEMS were responsible for unresectability. Median post-operative hospital stay was 15 days (range 7-32) for pts with Whipple and 3.5 days (range 1-10) for pts undergoing surgical exploration. Of note, in the 8/32 unresectable pts, additional ERCPs were not necessary due to preoperative SEMS placement. Complications included cholangitis prior to surgery n=1, pancreatic fistulas n=1, significant post-operative infections n=1, post-operative wound infections n=1, and postpancreatectomy hemorrhage n=1. In addition, 10/32 pts underwent neoadjuvant chemoradiation, delaying the interval between SEMS placement and surgery by mean of 82 days (range 27-145). 7/10 pts on neoadjuvant therapy had previously received plastic stents with 5/7 having stent occlusion and 2/5 experiencing cholangitis as a result. Conclusions: Biliary SEMS placement does not interfere with subsequent pancreaticoduodenectomy when the stent is placed below the anticipated line of bile duct resection. The frequency of post-operative complications does not appear to have increased as a result of SEMS. Placement of SEMS in pts undergoing neoadjuvant therapy may prevent premature plastic stent occlusion and subsequent delays in both chemoradiation and surgery. The use of biliary SEMS in pts with resectable pancreatic cancer can be advocated.