Background: Benefit from pancreaticoduodenectomy (PD) combined with superior mesenteric-portal vein (SMV-PV) resection for pancreatic adenocarcinoma with local venous invasion is controversial.
Methods: Using formal decision analysis, we compared outcomes for PD plus SMV-PV resection (group I) versus those for palliative chemoradiotherapy (no resection, group II) in patients with pancreatic adenocarcinoma with local venous invasion. Studies were identified using MEDLINE. 2,144 group I and 709 group II patients were evaluated. Patients with arterial invasion or distant metastases were excluded.
Results: Group I patients had a 40% higher 1-yr survival rate (p<0.0001) than group II patients. 1-way sensitivity analyses indicated the robustness of this finding is contingent on perioperative mortality rate and the percentage of cases in which true venous invasion by cancer is documented histologically. In the studies evaluated, median perioperative mortality was 3.6% (range, 0 - 29%), and the median percentage of cases with true venous invasion was 63.6% (range, 2.9 - 100%)We conducted a 2-way sensitivity analysis to evaluate the impact of varying these two parameters (Figure 1). The shaded portion of the figure (lower perioperative mortality rate and lower percentage of cases with true venous invasion) represents the area for which resection confers a higher 1-year survival than palliative treatment.
Conclusions: Our analysis suggests that pancreaticoduodenectomy with SMV-PV resection may confer a survival advantage over non-resectional palliation in select patients believed to have pancreatic adenocarcinoma with local venous invasion. Given the difficulty of identifying true vascular invasion preoperatively, these procedures should be done only if low perioperative mortality rates can be achieved.