Thrombosis and anticoagulation after portal vein reconstruction during pancreatic surgery: a systematic review.

Background: Portal vein (PV) resection and reconstruction, which includes the resection and reconstruction of the PV and superior mesenteric vein, enable surgical removal of borderline resectable and locally advanced pancreatic cancer. Thrombosis of the reconstructed PV represents a major cause of early postoperative and long-term morbidity and mortality. No universally accepted standard for anticoagulation exists. This study aimed to assess early and late thrombosis rates after PV reconstruction with special regard to the type of PV reconstruction and anticoagulation regimen and to comprehensively assess thrombotic events and their clinical effect in patients receiving pancreatic surgery with venous resection and reconstruction.

Methods: The Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines were followed. Studies reporting on PV resection and reconstruction providing data on thrombosis rates were included. The following parameters were assessed: study type, year of publication, number of patients, type/number of PV reconstruction, follow-up period, postoperative mortality, thrombosis rate of the reconstructed PV axis, intraoperative blood loss, and anticoagulation.

Results: A total of 23 studies with 2751 patients were included in the final analysis. Of note, 670 patients received tangential resection of the PV with venorrhaphy or patch repair, 1505 patients had segmental resection with end-to-end reconstruction, and 576 patients received reconstruction with an interposition graft/conduit. The pooled overall thrombosis rate was 15%. Reconstruction of tangential defects with either venorrhaphy or patch repair and end-to-end repair of segmental defects resulted in a thrombosis rate of 12%. Subgroup analysis according to the type of graft reconstruction revealed the highest occlusion rates of 55% in patients with allogeneic grafts, followed by up to 27% in patients with synthetic PV conduits. Autologous conduits had a thrombosis rate of 10%. Early thrombotic events were detected in 5% of patients after venorrhaphy/patch reconstruction and end-to-end reconstruction. Early events were most common in the allogeneic graft subgroup (22%), followed by synthetic conduits (15%). There were fewer early events in the autologous graft group (7%). Early PV thrombosis was associated with relevant mortality of up to 26%. Anticoagulation regimens varied between studies.

Conclusion: The overall thrombosis rate after PV resection is low. However, among the different reconstruction techniques, allogeneic interposition grafts/conduits had the highest thrombosis rates among the different types of reconstruction after PV resection. No specific anticoagulation strategy can be considered beneficial based on the existing literature.