Short-term and long-term outcomes of hepatopancreatoduodenectomy for extrahepatic cholangiocarcinoma and gallbladder carcinoma: A systematic review and meta-analysis with meta-regression.

Objectives: To evaluate short-term and long-term outcomes of hepatopancreatoduodenectomy for extrahepatic cholangiocarcinoma and gallbladder cancer.

Methods: A systematic search of electronic data sources and bibliographic reference lists was conducted. All studies reporting outcomes of hepatopancreatoduodenectomy for extrahepatic cholangiocarcinoma or gallbladder cancer were included, and their risk of bias were assessed. Effect sizes were determined for clinically relevant postoperative pancreatic fistula, clinically relevant posthepatectomy liver failure, bile leak, Clavien-Dindo classification III or greater complications, mortality, and 1- to 5-year survival using random-effects modeling followed by meta-regression analyses.

Results: Twenty-three retrospective studies (789 patients) conducted between 2007 and 2025 were included. Hepatopancreatoduodenectomy for extrahepatic cholangiocarcinoma was associated with clinically relevant postoperative pancreatic fistula: 52.1% (95% confidence interval, 38.8%-65.4%), CR-posthepatectomy liver failure: 31.5% (95% confidence interval, 14.5-48.4%), bile leak: 17.6% (95% confidence interval, 13.3-22.0%), Clavien-Dindo grade III or greater: 59.4% (95% confidence interval, 47.3-71.6%), mortality: 2.8% (95% confidence interval, 0.9-4.6%), and 1-year: 61.8% (95% confidence interval, 49.6-73.9%), 3-year: 30.2% (95% confidence interval, 23.5-36.9%) and 5-year survival: 23.7% (95% confidence interval, 17.3-30.2%). hepatopancreatoduodenectomy for gallbladder cancer was associated with clinically relevant postoperative pancreatic fistula: 48.7% (95% confidence interval, 19.9-77.5%), clinically relevant posthepatectomy liver failure: 15.7% (95% confidence interval, 0.2-31.2%), bile leak: 9.4% (95% confidence interval, 4.0-14.9%), Clavien-Dindo classification III or greater: 45.7% (95% confidence interval, 22.6-68.9%), mortality: 6.7% (95% confidence interval, 1.8-11.6%), and 1-year: 65.0% (95% confidence interval, 44.8-85.1%), 3-year: 19.9% (95% confidence interval, 10.8-29.0%), and 5-year survival: 14.0% (95% confidence interval, 5.2-22.9%). Portal vein resection was associated with clinically relevant postoperative pancreatic fistula (P = .003), clinically relevant posthepatectomy liver failure (P < .001), and Clavien-Dindo grade III or greater (P < .001) in extrahepatic cholangiocarcinoma, and clinically relevant postoperative pancreatic fistula (P < .001) and clinically relevant posthepatectomy liver failure (P < .001) in gallbladder cancer. Arterial resection was associated with clinically relevant posthepatectomy liver failure (P < .001), and Clavien-Dindo classification III or greater (P < .001) in extrahepatic cholangiocarcinoma. Portal vein embolization predicted posthepatectomy liver failure in both extrahepatic cholangiocarcinoma (P < .001) and gallbladder cancer (P < .001).

Conclusion: Hepatopancreatoduodenectomy for both extrahepatic cholangiocarcinoma and gallbladder cancer is associated with significant postoperative morbidity, particularly for extrahepatic cholangiocarcinoma or when combined with vascular resection. However, postoperative mortality is acceptable whilst long-term survival rates are particularly low for gallbladder cancer.