Back to Annual Meeting Program

Background: Preoperative endobiliary drainage (PBD) is commonly performed in patients with malignant obstruction for whom operative resection is eventually planned. The aim of this study was to compare the safety of pancreatoduodenectomy (PD) following PBD with self-expandable metallic stents (SEMS) versus standard plastic stents (PS).

Methods: We conducted a single center, retrospective cohort study of patients undergoing PBD followed by PD for periampullary cancer between January 1998 and December 2009. Patients were excluded if PBD was solely performed at a different facility because details regarding PBD complications and stent type were unavailable. To determine the potential impact of SEMS on PBD and postoperative outcomes, we compare patient characteristics as well as PBD and postoperative complications between groups. If patients had more than one PBD procedure, we present complications during 1) any PBD and 2) the last PBD procedure. SEMS were routine placed at least 1cm from the hepatic bifurcation. Variables are presented as simple proportions or medians, with two-way statistical comparisons (Fisher’s exact or Mann-Whitney-U test).

Results: We identified 147 patients who underwent PBD and then PD at our center, 17 (12%) of who underwent placement of a fully covered (10) or uncovered (7) SEMS. Among PS patients, 74% had a 10Fr PS placed during their final PBD. In all patients, 29% had previously undergone 1 (25%) or 2 (4%) PBDs with PS. Neoadjuvant therapy was used in 8.0% of PBD and 17.6% of SEMS patients (p=0.20). Including previous PBD attempts, patients had at least one failed PBD (17%) or other complication (18%) including early stent occlusion (n=22), pancreatitis (n=1) and perforation (n=1). Complications specific to the last PBD were similar (5.9% v. 7.1%, p=0.85) in SEMS and PS patients, with no cases of early stent occlusion in the SEMS group. The median number of PBD procedures was significantly greater in SEMS v. PS, 2 v. 1, p<0.001; no SEMS cases required repeat PBD.
The median time from last PBD to surgery was longer in SEMS v. PS patients, 31 v. 18 days, p=0.004. Postoperative complications Clavien grade ≥3 occurred in 22% of SEMS versus 11% of PS patients, p=0.37. Infection-specific complications after PD were similar in the SEMS (11.1%) and PS groups, 26.8%, p=0.31. Perioperative cholangitis occurred in one SEMS patient while one postoperative bile leak was recorded in the PS group. Median postoperative length of stay was similar in SEMS and PS patients, 8.5 v. 8.0 days; p=0.87.

Conclusion: Complications, particularly stent occlusion, occur frequently after PBD in patients scheduled to undergo PD for malignancy. Compared to PS, SEMS do not impact on the incidence of major postoperative complications and may be cost-effective as first-line endobiliary drainage given the high incidence of early stent occlusion with PS.