Previous to our RCT, four studies have already reported the optimal extent of pancreaticoduodenectomy. However, each study has been criticized due to the small number of patients, absence of objectively controlled operative techniques, no statistical calculation for required number of enrolled patients, mixed cases with non-pancreatic ductal adenocarcinoma, insufficient clearance of retroperitoneal tissue and lymph node, and no consideration of nerve plexus dissection or extensive dissection of nerve plexus provoking uncontrolled diarrhea [19–23]. To overcome the abovementioned pitfalls of previous RCTs, we designed and executed our own RCT on pancreatic ductal adenocarcinoma based on larger sample size with standardized method of operation and with focus on dissection of nerve plexus as well as lymph node.

In standard resection, lymph node around pancreas head (LN 13, 17) and gallbladder (LN 12c) were only removed without nerve dissection around hepatic artery or superior mesenteric artery (SMA). For extended resection, lymph node around common hepatic artery (LN 8), celiac axis (CA) (LN 9), peripancreatic area (LN 13, 17), hepatoduodenal ligament (LN 12), SMA (LN 14), and para-aortic area (LN16) between CA and inferior mesenteric artery were dissected. All the soft tissues around hepatoduodenal ligament were completely dissected and skeletonized. Nerve plexus or ganglion right side to CA and SMA were dissected semi-circumferentially (Fig. 6.1). Differences in extent of resection between two groups are summarized in Table 6.3 [23].