Pancreatic stump perfusion assessment using indocyanine green fluorescence and its impact on postoperative pancreatic fistula: A systematic review and meta-analysis.

Abstract

Indocyanine green fluorescence imaging can be used for intraoperative assessment of pancreatic stump perfusion with the aim to guide strategies to prevent postoperative pancreatic fistula in pancreatic surgery. The impact of indocyanine green in this setting is unknown since a systematic review is lacking. This review aimed to assess the relationship between indocyanine green fluorescence imaging of pancreatic stump perfusion and the risk of clinically relevant postoperative pancreatic fistula after pancreatic surgery. A systematic literature search and meta-analysis were conducted, including studies published up to June 2025 that reported postoperative pancreatic fistula rate after pancreatic resection in relation to intraoperative pancreatic stump perfusion assessed by intraoperative indocyanine green fluorescence imaging. Hypoperfusion was defined as a heterogeneous or completely absent signal. Primary outcome was postoperative pancreatic fistula of which only grade B/C were included. Secondary outcome was postpancreatectomy acute pancreatitis. All 3 studies included analyzed patients who underwent pancreatoduodenectomy, comprising a total of 100 patients, with 18 (18%) presenting pancreatic stump hypoperfusion. No studies analyzing left pancreatectomy were identified, whereas only 1 paper analyzed the association between pancreatic hypoperfusion and postpancreatectomy acute pancreatitis. In that study, no patients developed postpancreatectomy acute pancreatitis after revision of the transection line initially found to be hypoperfused. The overall rate of postoperative pancreatic fistula was 13%. After robotic pancreatoduodenectomy (n = 27), stump hypoperfusion was associated with postoperative pancreatic fistula (67% vs 17%; P = .026), compared to the normally perfused group. No significant association of hypoperfusion and postoperative pancreatic fistula was observed after open pancreatoduodenectomy (n = 73). Meta-analysis confirmed the association of stump hypoperfusion with postoperative pancreatic fistula (odds ratio, 8.83; 95% confidence interval, 2.21-35.23; P = .005). A hypoperfused pancreatic stump, assessed intraoperatively using indocyanine green fluorescence imaging, appears to be associated with postoperative pancreatic fistula after pancreatoduodenectomy. Further research is needed to confirm these results in left pancreatectomy and develop a standardized indocyanine green protocol for pancreatic surgery.