Introduction: Portal Vein (PV) resection may be necessary to achieve free resection margins during Pancreaticoduodenectomy (PD). The outcome following PV resection in patients with pancreatic ductal adenocarcinoma has not been thoroughly documented; moreover, a comparative analysis to patients without vascular involvement is lacking.
Methods: We analyzed all consecutive patients with pancreatic ductal adenocarcinoma undergoing PV resection during PD between 1985- mid 2000. Each was matched with a patient in whom there was no evidence of PV involvement. Match (1:1) was based upon age, tumor size, lymph node status, year of the procedure, and surgeon. We included only patients with intentional resection of the PV aiming to achieve free margins; accidental repairs of PV were excluded. Patients were followed by clinic visits and mail correspondence. Unpaired t-test was used to compare parametric variables, and chi-square for contingency tables. The Kaplan-Meier method was used to generate recurrence and survival curves, and the log rank test for comparison (p<0.05 was significant).
Results: We identified 29 patients with PV resection which were matched with 29 controls. There was no difference in any of the factors used for matching, including gender. Grade 3 was prevalent in both groups. Margins of resection were microscopically involved in 7 patients with PV resection and 8 without (p=0.76). Vein was involved histologically by tumor in 14, not involved in 6 and unknown in 9. Venorraphy was performed in 19 patients, interposition graft (PTFE) in 6, primary vein anastomosis in 3, and vein graft in 1. PV was thrombosed in 1 patient postoperatively and stenosed in one; late tumor thrombosis occurred in one patient. For patients with and without PV resection operative time (402 vs 384 min, p=0.52), length of stay (15 vs 13 days, p=0.24), transfusion requirement (2.1 vs 1.2 units RBC, p=0.16), and complication rate (51 vs 38%. P=0.29) were not different. One patient with PV resection died postoperatively. While recurrence-free survival had a trend for significance in the group without PV resection (47 vs 67% at 1-yr, p=0.08) 3-yr actuarial survival was similar to this group (21 vs 28%, p=0.30). Mean follow-up was 802 days.
Conclusions: PV resection can be performed safely with PD and survival was found not to differ significantly from a matched group of controls. In selected patients PV involvement should not be a contraindication to potentially curative resection.