Vascular surgery collaboration during pancreaticoduodenectomy with vascular reconstruction

Background: Once thought to have unresectable disease, pancreatic cancer patients with portal venous involvement are now reported to have comparable survival after pancreaticoduodenectomy (PD) with vascular reconstruction (VR) as compared with patients without vascular involvement. We hypothesize that a multidisciplinary approach involving a vascular surgeon will minimize morbidity and improve patency of VRs. Methods: We identified 204 patients who underwent PD for pancreatic adenocarcinoma from 1997 to 2008. Patients who underwent PD with VR (N = 42) were compared with those who underwent standard PD (N = 162). VRs were performed by a vascular surgeon and involved primary repair (N = 8), vein patch (N = 25), or interposition grafting (N = 9) with femoral or other venous conduit. Results: Patients undergoing PD with VR had larger tumors (3.0 cm vs. 2.5 cm, P < 0.01) but did not have different rates of tumor-free margins (73% vs. 72%, P = 0.84) or lymph nodes metastases (50% vs. 38%, P = 0.14). The VR group had higher median blood loss (875 mL vs. 550 mL, P = 0<0.01), but no differences in mortality, complication rates, length of stay, or readmission rates were found in a median follow-up of 29 months. Overall survival rates were similar. Predictors of mortality on multivariate analysis included increasing histological grade (P = 0.01), positive lymph nodes (P = 0.01), and increasing tumor size (P = 0.01), but not VR (P = 0.28). When evaluated by computed tomography scans within 6 months postoperatively, 97% of reconstructions remained patent. Conclusions: The need for VR is not a contraindication to potentially curative resection in patients with pancreatic adenocarcinoma. Assistance of a vascular surgeon during VR may allow moderate-volume centers to achieve outcomes comparable with high-volume centers.