Is Pancreatectomy with Vascular Resection and Reconstruction for Locally Advanced Pancreatic Neuroendocrine Tumors Safe?

Abstract

Introduction: While pancreatectomy with vascular resection and reconstruction (VR) is commonly performed for locally advanced pancreatic adenocarcinoma, little is known regarding outcomes for pancreatic neuroendocrine tumors (pNETs). Methods: Non-parenchyma-sparing pancreatectomies for pNETs at Mayo Clinic 2000-2020 were retrospectively reviewed. Propensity score matching was performed to adjust for selection bias. Differences between groups were compared with chi-square, Fisher‘s exact, and Mann-Whitney tests and survival assessed using Kaplan-Meier analysis. Results: Of 813 eligible patients, 36 (4.4%) required VR: 22 underwent pancreaticoduodenectomy or total pancreatectomy and 14 distal pancreatectomy. 34 patients had venous resection, one arterial, and one both. Patients requiring VR had larger tumors (median 55 vs. 25 mm, p<0.001), higher grade (84.6% vs. 58.5% grade 2-3, p=0.008), and higher stage (44.4% vs. 19.2% stage 4, p<0.001). After matching, the VR group had longer operative times (median 410 vs. 331 minutes, p<0.001), greater blood loss (47.2% vs. 25.0% >1000 ml, p=0.035), and higher transfusion rates (63.9% vs. 29.2%, p=0.001). Major complications (Clavien-Dindo >/=3) occurred in 41.7% after VR and 25.0% after standard resection (p=0.121). 90-day mortality was 2.8% after VR and 4.2% after standard resection (p>0.999). Over the entire period, overall (p=0.172) and progression-free (p=0.105) survival were not significantly different. On landmark analysis, VR patients had worse overall survival starting at 60 months (p=0.043) and worse progression-free survival starting at 18 months (p=0.044). Conclusion: For patients with locally advanced pNETs, pancreatectomy with VR can be performed in selected patients at high-volume centers with acceptable perioperative morbidity and short- and long-term survival.