Reoperation for pancreatic fistula: a systematic review of completion pancreatectomy vs. pancreas-preserving-procedures and outcomes.

Background: Consensus on the nomenclature and indications for reoperation for post-operative pancreatic fistula (POPF) after pancreatoduodenectomy (PD) is lacking. This study explores the available literature to classify the different types of reoperations and report outcomes.

Methods: A systematic literature search was performed, including articles from 2010 to 2024 reporting reoperations for POPF after PD. The primary outcome was 30- or 90-day-mortality. Secondary outcomes included reoperation date, additional relaparotomy, ICU-admission, hospital stay, rate of pancreatic-exocrine-insufficiency, diabetes and long-term survivors.

Results: Twenty-five studies were reviewed with 766 patients reoperated for POPF after PD, 283 (37 %) undergoing completion pancreatectomy (CP) and 483 (63 %) pancreas-preserving-procedures (PPPs). Among PPPs, drainage (30 %), wirsungostomy (14 %), pancreatic anastomosis repair (6 %), "sinking" of pancreatic stump (6 %) and re-do pancreatic anastomosis (4 %) were identified. The main indications for reoperation were post-pancreatectomy hemorrhage, necrotizing acute pancreatitis, sepsis and peritonitis. PPPs were preferred with severe hemodynamic instability. Mortality rates after CP and PPPs ranged from 20 to 56 % and 0-67 %, respectively. Early reoperation was associated with reduced ICU-recovery after "sinking" (p = 0.049).

Conclusion: Reoperation for POPF after PD is rarely needed. When it is, early timing seems critical for better outcomes, and PPPs seems to be the best bail out option in patients with severe hemodynamic instability.