Outcomes after pancreatoduodenectomy and total pancreatectomy in patients with a high-risk pancreatic anastomosis: An entropy balance analysis

Background: The benefit of performing a total pancreatectomy in high-risk patients is largely debated. Our aim is to evaluate what would have been the short-term outcomes of patients who underwent a high-risk pancreatoduodenectomy if a total pancreatectomy was performed instead. Methods: Perioperative data from patients who underwent pancreatoduodenectomy or total pancreatectomy at 5 tertiary hepato-pancreato-biliary centers (2016e2022) were collected prospectively. The alternative fistula risk score was calculated, and patients with a risk of developing a clinically relevant postoperative pancreatic fistula >20% were analyzed. Hainmueller's "entropy balance" method was applied. Results: A total of 1,172 pancreatoduodenectomy and 448 total pancreatectomy procedures were evaluated; 277 patients were at high risk of a clinically relevant postoperative pancreatic fistula. It was observed that total pancreatectomy resulted in an obvious nullification of a clinically relevant postoperative pancreatic fistula and a decreased minor complications rate, with an odds ratio of 0.427 (95% confidence interval: 0.198, 0.919; P = .030). Notably, in patients with a performance status scored as 1 or an American Society of Anesthesiologists score of II-III, total pancreatectomy considerably reduced major complications (odds ratio: 0.317 [95% confidence interval: 0.151, 0.666; P = .002] and 0.607 [95% confidence interval: 0.404, 0.912; P = .016]) and mortality (odds ratio: 0.063 [95% confidence interval: 0.014, 0.376; P = .001] and 0.14 [95% confidence interval: 0.021, 0.953; P = .046]). Failure to rescue (20.5% vs 15.8%) was also reduced in total pancreatectomy pseudopopulation, and no major differences in postoperative glycemic control were observed. Conclusion: In our simulation patients with a high-risk pancreatic anastomosis, total pancreatectomy improved short-term surgical outcomes. A strong impact on major complications is expected in specific subpopulations, which should be the real target for the assessment and application of this extreme mitigation strategy. Further randomized studies are required to assess quality of life and long-term complications in this setting. (c) 2025 Elsevier Inc. All rights are reserved, including those for text and data mining, AI training, and similar technologies.