Short- and long-term outcomes after venous resection during left-sided and total pancreatic resection: a nationwide cohort study

Abstract

Background: Venous resection (VR) during left-sided pancreatic resection (LPR) is typically more challenging than during pancreatoduodenectomy, because the pancreatic head remains in situ. Theoretically, converting from LPR to total pancreatectomy (TP) could resolve this issue. This study compared clinical outcomes after LPR and TP with and without VR. Methods: This nationwide retrospective study included all patients who underwent LPR for all indications or single-stage TP in the Netherlands from 2014 to 2019. Results: One-thousand-seventy-five patients were analyzed: 996 underwent LPR and 79 TP. Twenty-eight (3%) LPRs were with VR (VR+), of which 17 wedge and 11 segmental resections. Of 21 (27%) TPs with VR, 11 were wedge and 10 segmental resections. In the LPR-group, VR+ was associated with a higher incidence of portomesenteric venous thrombosis (PVT) than VR- (3/28 vs. 5/968, p < 0.001), but not with major complications or mortality. Similarly, in the TP-group, higher rates of PVT were reported after VR+ (2/21 vs. 0/58, p = 0.004). The incidence of major complications and PVT were comparable between LPR-VR+ and TP-VR+, but postoperative mortality was higher in TP-VR+ (6/21 vs. 1/28 p = 0.032). Conclusion: Our findings suggest that, despite technical impediments of VR during LPR, switching to TP may not benefit patients.