Abstract
Purpose: No consensus exists regarding the most effective form of pancreaticojejunostomy (PJ) following pancreaticoduodenectomy (PD). Methods: Data were gathered through the American College of Surgeons-National Surgical Quality Improvement Program, Pancreatectomy Demonstration Project. A total of 1781 patients underwent a PD at 43 institutions. After appropriate exclusions, 890 patients were analyzed. Patients were divided into duct-to-mucosa (n = 734, 82%) and invagination (n = 156, 18%) groups and were compared by unadjusted analysis. Type of PJ was included in eight separate morbidity and mortality multivariable analyses. Results: Invagination patients had higher serum albumin (p < 0.01) and lower body mass index (p < 0.01), were less likely to have a preoperative biliary stent (p < 0.01), and were more likely to have a soft gland (p < 0.01). PJ anastomosis type was not associated with morbidity but was associated with mortality (duct-to-mucosa vs. invagination, odds ratio = 0.22, p < 0.01). Among patients who developed a clinically relevant pancreatic fistula, none of the 119 duct-to-mucosa, compared with 5 of 21 invagination, patients died (p < 0.01). Conclusion: Patients who undergo a PJ by duct-to-mucosa or invagination differ with respect to preoperative and intraoperative variables. When an invagination PJ leaks, there may be a greater influence on mortality than when a duct-to-mucosa PJ leaks.