Prognostic implications of tumor invasion or adhesion to peripancreatic vessels in resected pancreatic cancer

Abstract

The purpose of this study was to evaluate the operative risk and the prognostic implications of pancreatectomy plus resection and reconstruction of peripancreatic vessels (PPV) in patients with pancreatic adenocarcinoma. One hundred ten patients who underwent pancreatectomy with PPV resection and reconstruction (Study Group; SG) were retrospectively compared with 62 patients without distant metastasis who were palliated, (Control Group 1; CG-1), as well as 197 patients who underwent "conventional"pancreatectomy (Control Group 2; CG-2). Postoperative morbidity and mortality were similar in SG (33% and 3%), in CG-1 (26% and 3%), and in CG-2 (40% and 6%) patients. Median survival time (MST) of SG patients (15 months) was longer than that of CG-1 patients (6 months; P < .0001) and similar to that of CG-2 patients (18 months). Patients undergoing isolated venous resection (n = 84) had the best outcome (MST: 15 months) ( P < .0001 vs CG-1 patients), while patients undergoing resection of multiple PPV (n = 14) had the worst outcome (MST: 8 months). PPV infiltration, histologically proven in 64 patients (65%), was associated with decreased MST only if the tunica intima was infiltrated (26%) (11 months; P < .001). Multivariate analysis showed that no adjuvant therapy, intimal invasion, and poorly differentiated histology were associated with a higher hazard of death by 2.2, 2.2, and 2.5-fold, respectively. In properly selected patients, pancreatectomy plus resection and reconstruction of PPV was performed as safely as palliation or "conventional" pancreatectomy and was associated with better survival when compared to palliation.