Application of a new collagen-based sealant for the treatment of pancreatic injury

Pancreatic injury is the most frequent serious morbidity that can occur following laparoscopic splenectomy. The presumed mechanism of injury is direct trauma to the pancreatic tail during hilar dissection or transection with endovascular staplers. It was our aim to study the direct application of topical sealants to a pancreatic injury to prevent a pancreatic leak. A porcine model of pancreatic injury in a normal pancreas was developed. Fourteen animals underwent midline laparotomy and subtotal (8 cm) distal pancreatectomy. The pancreas was sharply divided and minimal cauterization used. A (1/4)-inch round Jackson Pratt drain was secured to the pancreatic stump in all animals. The control group (n = 7) underwent no further treatment. Two groups underwent treatment with a sealant. Group 1 (n = 3) had a hydrogel applied to the pancreatic remnant. Group 2 (n = 4) had a biodegradable sealant of PEG (polyethylene glycol)-collagen formulation applied to the pancreatic stump by aerosol. During the postoperative period, animals were fed 2.2 lb/d of a high fat (10% lard) diet to stimulate pancreatic exocrine function. Total daily drain outputs were recorded, and drain amylase content and serum amylase concentration were measured on postoperative days 3, 7, and 10. A significant pancreatic leak was defined as a drain to serum amylase content of greater than 3:1. Animals were killed on day 10 and explored for undrained fluid collections. The pancreas was removed for histologic evaluation. Control and treatment groups were compared using repeated measures ANOVA of log-transformed data. All animals survived until elective necropsy. In group 1, average total drain outputs were not significantly less than the controls (125 mL versus 269; P = .66). The mean drain/serum amylase concentration on days 3, 7, and 10 were not significantly different between group 1 and the control group. Group 2 had significantly less overall average drainage output than controls (40 mL versus 269 mL; P = .0006). Furthermore, group 2 had a significant reduction in pancreatic leaks on days 3, 7, and 10 as measured by mean drain/serum amylase concentration. No undrained fluid collections were identified at autopsy. Based on our experimental porcine model, a novel collagen-based, biodegradable hydrogel can prevent a ductal leak following pancreatic injury. These encouraging data have fostered application by aerosol treatment to the hilar bed at laparoscopic splenectomy to diminish capsular disruption from mechanical injury.