Impact of Preoperative Biliary Drainage and Bile Colonization On Complications After Pancreatic Head Resection with Bilioenteric Anastomosis in 455 Patients
Frank Makowiec*, Hartwig Riediger, Eva Fischer, Tobias Keck, Ulrich ADAM, Ulrich T. Hopt
Dept. of Surgery, University of Freiburg, Freiburg, Germany
Many patients with malignant or benign diseases requiring pancreatic head resection (PHR) preoperatively undergo (endoscopic or percutaneous) biliary drainage (PBD) to relieve severe jaundice.or symptomatic cholestasis. During the last decade the value and the risks of PBD have been discussed controversially. We, therefore, evaluated our results after PHR with bilioenteric anastomosis and included data on intraoperative bile colonization.
Methods: From 1994 to 3/2006 455 patients underwent PHR with bilioenteric anastomosis (74% PPPD, 16% Whipple, 8% duodenum-preserving PHR, 2% pancreatectomy). Indications for PHR were pancreatic or periampullary cancer (52%), chronic pancreatitis (38%) or others (10%). The results of intraoperative bile cultures (IBC) were available in 76% of the patients. Forty-eight percent had undergone PBD (44% endoscopically (endoPBD), 4% as PTCD). Data were gained by retrospective analysis of our prospective pancreatic database.
Results: Operative mortality was 2.6 %. Surgical complications (SurgComp) occurred in 32.5%, infectious complications (intraabdominal and/or wound infection; InfComp) in 17.8%. Pancreatic leakage was documented after 13.8% of the PHR. EndoPBD (vs. no PBD) was associated with a reduced SurgComp rate (28% vs. 36%, p=0.06) but showed no significant reduction in individual complication types. Patients with PTCD (n=20; vs no PTCD) had a slightly higher SurgComp rate (45% vs 32%, p=0.16) especially because of significantly more InfComp (35% vs 17%, p=0.04). Pancreatic leakage and mortality were not influenced by presence or type of PBD. In the subgroup of 344 cases with available IBC 174 (51%) had undergone PBD. Here, 216 patients (63%) had positive IBC. Positive IBC occurred in 91% after previous PBD but only in 9% in patients without PBD. A positive IBC (vs. negative IBC) did not influence total SurgComp rate (31% vs. 35%) but was associated with significant more abdominal abcesses (12% vs 4%, p<0.05).
Conclusions: Percutaneous PBD increases infectious complication rates after PHR. Endoscopic PBD is not associated with a higher overall complication rate although it promotes bile colonization which may lead to more abdominal abscesses. Our results demonstrate the relative overall safety of endoscopic PBD regarding postoperative complications but the prophylactic use of perioperative antibiotics may be discussed after previous PBD.
2007 Program and Abstracts | 2007 Posters