To investigate the perioperative risk and effect evaluation of pancreatico- duodenectomy (PD) with vascular resection and reconstruction.

A retrospective analysis was carried out on three hundred and sixty-three patients who had undergone PD from January 2010 to February 2016 in the First Affiliated Hospital of Sun Yat-sen University. The combination group included 38 cases underwent PD and treated with vascular resection and reconstruction, and the control group included 325 cases just undergoing PD. The differences in operative time, blood loss, postoperative hospital stay, rate of post-PD pancreatic fistula, rate of intra-abdominal or alimentary tract hemorrhage, rate of biliary fistula and perioperative mortality rate were compared between the two groups.

All cases were successfully completed. The operative time in the combination group was significantly longer than control group [(7.5±2.2) h vs (5.9±1.7) h, t=-5.153, P<0.001]. The median intraoperative blood loss in combination group was 500 (100-4 500) ml, which was more than 200 (20-12 000) ml of control group (Z=-4.028, P<0.001). The combination group had no advantage in the postoperative hospital stay [(20.3±14.2) d vs (18.5±13.1) d, t=-0.811, P>0.05]. The incidence of complication and mortality in the combination group were 26.3% (10/38) and 2.6% (1/38), while 35.4% (115/325) and 1.8% (6/326) in the control group, with no significant differences between two groups (P=0.266, 0.539). Postoperative vascular pathology in the combination group confirmed that 15 patients were invaded by cancer cells, 35 of graft patency, and 3 of portal vein thrombosis complications.

Compared with normal PD, PD combined with vascular resection and reconstruction does not increase postoperative hospital stay, perioperative complication rate and mortality, and may help patients with vascular invasion.