Multi-institutional study of outcomes from portal vein reconstruction during pancreatic cancer surgery.

[OBJECTIVE] Portal vein reconstruction (PVR) is undertaken to enhance resectability during pancreatic cancer surgery. We hypothesize that the type of reconstructions for patients undergoing PVR may not significantly influence overall outcomes.

[METHODS] Patients undergoing PVR for pancreatic cancer surgery (2013-2023) were identified at three academic institutions. χ and Kaplan-Meier analyses were used to evaluate outcomes.

[RESULTS] Eighty patients were captured, with 53% men and 45% women. Preoperative chemoradiation was undertaken in 70% of patients. Pancreaticoduodenectomy was undertaken in 79 patients (92%), with distal pancreatectomy in one (1.2%). Primary repair was undertaken in 46 patients (58%), bovine patch in 22 (28%), other vein patch in two (3%), with interposition conduits such as the internal jugular vein, great saphenous vein, superficial femoral vein, and PTFE in 10 (13%). Thirty-day complications included four deaths (5%), one stroke (1.3%), and one major adverse cardiac event (1.3%). Overall complications included death in 45 patients (56%), mesenteric ischemia in one (1.3%), and reintervention in one patient (1.3%). Postoperative medications included aspirin in 43 patients (54%), clopidogrel in one (1.3%), dual antiplatelet therapy in three (4%), and anticoagulation in 17 (21%). Over 21 months mean follow-up, 18 reconstructions (23%) occluded, with three (4%) reintervened upon at a mean of 6.3 months from index procedure with no association with type of repair. At 36 months, estimated survival was 66%, of which primary patency in survivors was 78%.

[CONCLUSIONS] PVR during pancreatic cancer surgery is safe and durable. One-quarter of all reconstructions fail at medium-term follow-up, and very few require reintervention for clinical symptoms. Although primary repair has been suggested as the most optimal reconstruction, our findings imply that long-term patency of PVR may not be clinically important. Furthermore, outside of infectious considerations, it is likely that conduit or reconstruction type does not affect outcome.