Pancreatic head resection with segmental duodenectomy: safety and long-term results

To evaluate the usefulness and long-term results with pancreatic head resection with segmental duodenectomy (PHRSD; Nakao's technique) in patients with branch-duct type intraductal papillary mucinous neoplasms (IPMNs). A prospective study from Nagoya (Japan) and Barcelona (Spain).

Surgery should be the first choice of treatment of IPMNs. An aggressive surgery (eg, pancreatoduodenectomy) should be questioned in patients with an indolent disease or with noninvasive tumors. Recently, organ-preserving pancreatic resections for benign and noninvasive IPMN located in the head of the pancreas have been described. We have PHRSD in which the pancreatic head can be completely resected and the major portion of the duodenum can be preserved by this procedure. There have been only 4 reports concerning PHRSD with <8 patients (each one) in the English literature.

Thirty-five patients underwent PHRSD (20 men, 15 women), mean age 65.1 +/- 9.0 (range, 55-75). Mean maximal diameter of the cystic lesion was 26.4 +/- 5.3 mm (range, 20-33 mm) and mean diameter of the main pancreatic duct was 3.3 +/- 0.5 mm (range, 3.0-4.0 mm). Alimentary tract reconstruction was performed in 20 patients by pancreatogastrostomy, duodenoduodenostomy, and choledochoduodenostomy (type A) and 15 patients by pancreaticojejunostomy, duodenoduodenostomy and choledochojejunostomy (Roux-en-Y; type B). Surgical parameters, postoperative complications, endocrine function, exocrine function, and long-term outcomes were evaluated. To compare the perioperative factors, a matched-pairs analysis between PHRSD patients and patients with pylorus preserving pancreaticoduodenectomy (PPPD) was performed. In the latter group were included 32 patients with branch-duct type of IPMN operated during the same time period that patients with PHRSD. The mean follow-up period was 48.8 months.

Mean operative time after PHRSD was 365 +/- 50 and mean surgical blood loss was 615 +/- 251 mL. There was no mortality. Pancreatic fistula occurred in 10% and 13% with types (alimentary tract reconstruction) A and B, respectively. Noninvasive IPMN was found in 31 patients and invasive IPMN in 4 patients (11.4%). In the matched-pairs analysis between PHRSD and PPPD, the 2 procedures were comparable in regard to operation time and intraoperative blood loss. The overall incidence of pancreatic fistula was higher after PPPD than after PHRSD; the difference was not statistically significant. When fistulas occurred after PHRSD they were grade A (biochemical). In contrast, pancreatic fistulas after PPPD were grade A in 78% of cases and grade B in 22% (clinically relevant fistula). The incidence of delayed gastric emptying was significantly higher in the PPPD group compared with the PHRSD group (P < 0.01). Endocrine pancreatic function, measured by fasting blood glucose levels and HbA1, levels was unchanged in 94.28% of patients, in the PHRSD group, and in 87.87% in the PPPD group. Body weight was unchanged in 80% after PHRSD and in 59% after PPPD. Postoperative enzyme substitution was needed in 20% of patients after PHRSD and in 40% patients after PPPD. The 5-year survival rate was 100% in patients with benign IPMN and 42% in patients with invasive IPMN.

PHRSD is a safe and reasonable technique appropriate for selected patients with branch-duct IPMN. The major advantages of PHRSD are promising long-term results in terms of pancreatic function (exocrine and endocrine) with important consequences in elderly patients. Long-term outcome was satisfactory without tumor recurrence in noninvasive carcinoma. PHRSD should therefore be considered as an adequate operation as an organ-preserving pancreatic resection for branch-duct type of IPMN located at the head of the pancreas.