Association of Perioperative Transfusion with Recurrence and Survival After Resection of Distal Cholangiocarcinoma: A 10-Institution Study from the US Extrahepatic Biliary Malignancy Consortium

Background: Perioperative allogeneic blood transfusion is associated with poor oncologic outcomes in multiple malignancies. The effect of blood transfusion on recurrence and survival in distal cholangiocarcinoma (DCC) is not known. Methods: All patients with DCC who underwent curative-intent pancreaticoduodenectomy at 10 institutions from 2000 to 2015 were included. Primary outcomes were recurrence-free (RFS) and overall survival (OS). Results: Among 314 patients with DCC, 191 (61%) underwent curative-intent pancreaticoduodenectomy. Fifty-three patients (28%) received perioperative blood transfusions, with a median of 2 units. There were no differences in baseline demographics or operative data between transfusion and no-transfusion groups. Compared with no-transfusion, patients who received a transfusion were more likely to have (+) margins (28 vs 14%; p = 0.034) and major complications (46 vs 16%; p < 0.001). Transfusion was associated with worse median RFS (19 vs 32 months; p = 0.006) and OS (15 vs 29 months; p = 0.003), which persisted on multivariable (MV) analysis for both RFS [hazard ratio (HR) 1.8; 95% confidence interval (CI) 1.1–3.0; p = 0.031] and OS (HR 1.9; 95% CI 1.1–3.3; p = 0.018), after controlling for portal vein resection, estimated blood loss (EBL), grade, lymphovascular invasion (LVI), and major complications. Similarly, transfusion of ≥ 2 pRBCs was associated with lower RFS (17 vs 32 months; p < 0.001) and OS (14 vs 29 months; p < 0.001), which again persisted on MV analysis for both RFS (HR 2.6; 95% CI 1.4–4.5; p = 0.001) and OS (HR 4.0; 95% CI 2.2–7.5; p < 0.001). The RFS and OS of patients transfused 1 unit was comparable to patients who were not transfused. Conclusion: Perioperative blood transfusion is associated with decreased RFS and OS after resection for distal cholangiocarcinoma, after accounting for known adverse pathologic factors. Volume of transfusion seems to exert an independent effect, as 1 unit was not associated with the same adverse effects as ≥ 2 units.