Portal vein resection in borderline resectable pancreatic cancer: a United Kingdom multicenter study.

BACKGROUND

Until recently, in the United Kingdom, borderline resectable pancreatic cancer with invasion into the portomesenteric veins often resulted in surgical bypass because of the presumed high risk for complications and the uncertainty of a survival benefit associated with a vascular resection. Portomesenteric vein resection has therefore remained controversial. We present the second largest published cohort of patients undergoing portal vein resection for borderline resectable (T3) adenocarcinoma of the head of the pancreas.

STUDY DESIGN

This is a UK multicenter retrospective cohort study comparing pancreaticoduodenectomy with vein resection (PDVR), standard pancreaticoduodenectomy (PD), and surgical bypass (SB). Nine high-volume UK centers contributed. All consecutive patients with T3 (stage IIA to III) adenocarcinoma of the head of the pancreas undergoing surgery between December 1998 and June 2011 were included. The primary outcomes measures are overall survival and in-hospital mortality. Secondary outcomes measure is operative morbidity.

RESULTS

One thousand five hundred and eighty-eight patients underwent surgery for borderline resectable pancreatic cancer; 840 PD, 230 PDVR, and 518 SB. Of 230 PDVR patients, 129 had primary closure (56%), 65 had end to end anastomosis (28%), and 36 had interposition grafts (16%). Both resection groups had greater complication rates than the bypass group, but with no difference between PD and PDVR. In-hospital mortality was similar across all 3 surgical groups. Median survival was 18 months for PD, 18.2 months for PDVR, and 8 months for SB (p = 0.0001).

CONCLUSIONS

This study, the second largest to date on borderline resectable pancreatic cancer, demonstrates no significant difference in perioperative mortality in the 3 groups and a similar overall survival between PD and PDVR; significantly better compared with SB.