Ischemic complications after preoperative arterial embolization in pancreatectomy with arterial resection: lessons learned from 14 years institutional experience with a narrative review

Abstract Background Preoperative arterial embolization (PAE) of the celiac trunk or hepatic artery may promote recruitment of collateral circulation and thereby prevent hepatic ischemia during pancreatic surgery for locally advanced pancreatic cancer requiring arterial resection (AR) without the need for reconstruction. Despite its theoretical benefits, the role of PAE still remains unclear. Moreover, embolization of variant hepatic arteries before pancreatoduodenectomy (PD), common hepatic artery (CHA) embolization before distal pancreatectomy with celiac axis resection (DP-CAR), and PAE as an adjunct to complex vascular resections represent anatomically and clinically distinct scenarios. Methodology A retrospective single-center analysis was performed to identify patients who underwent PAE prior to pancreatic resection with AR. Patients who did not subsequently undergo pancreatic resection were excluded. The final cohort was stratified into three groups: hepatic artery embolization before PD with arterial resection, CHA embolization before DP-CAR, and PAE as an adjunct to complex or extended vascular resection. Technical complications related to embolization, arterial thrombosis, ischemic complications, major morbidity, reoperation, and mortality were analyzed descriptively. Results Thirty-four patients underwent PAE; eight did not subsequently undergo pancreatic resection, leaving 26 patients for analysis. A total of 28 embolization sessions were performed. Thirteen patients underwent hepatic artery embolization before PD with arterial resection, eight underwent CHA embolization before DP-CAR, and five underwent PAE before complex or extended vascular resection. Six patients (23.1%) experienced at least one technical, thrombotic, and/or ischemic event: 0/13 (0%) in the PD group, 4/8 (50.0%) in the DP-CAR group, and 2/5 (40.0%) in the complex/extended resection group. Major morbidity (Clavien–Dindo ≥IIIb) occurred in 7/26 patients (26.9%), and 5/26 (19.2%) required reoperation. Thirty- and 90-day mortality were both 2/26 (7.7%). All arterial reconstructions were planned preoperatively, and no unplanned arterial reconstruction was required. Conclusion PAE before pancreatic arterial resection encompasses distinct clinical strategies that should not be considered equivalent. No technical, thrombotic, or ischemic events occurred after hepatic artery embolization before PD in this series, whereas such events were observed after CHA embolization before DP-CAR and complex vascular resections. Given the small heterogeneous cohort, absence of a comparator group, and non-standardized assessment of collateral perfusion, these findings cannot establish a protective effect of PAE or determine its contribution to subsequent ischemic events. PAE should therefore be considered an individualized adjunct within the overall vascular surgical strategy.