Hepatic Venous Congestion in Living Donor Grafts in Liver Transplantation: Is There an Effect on Hepatocellular Carcinoma Recurrence?

A certain degree of graft congestion in living donor liver transplantation (LDLT) using a right liver graft may be inevitable because of the mismatch between the inflow and outflow structures of the liver. The subsequent inflammatory reaction and rapid regeneration of the graft have been suggested as causes of tumor recurrence. Therefore, we investigated the influence of graft congestion on hepatocellular carcinoma (HCC) recurrence after LDLT. Two hundred eighty-nine LDLT patients for HCC within the University of California San Francisco criteria between November 1999 and February 2012 were investigated. Patients were assigned to groups on the basis of the degree of congestion (<= 10% for group A and >10% for group B), which was determined by 3-dimensional reconstruction of posttransplant multidetector helical computed tomography within 2 weeks. Perioperative characteristics, regeneration rates after 6 months, and recurrence rates were compared between the groups, and a multivariate analysis of the influence of congestion on tumor recurrence was subsequently completed. No significant difference in demographics was found. Group B had more elevated peak posttransplant levels of aspartate aminotransferase (296.26 versus 227.53, P=0.05), alanine aminotransferase (382.91 versus 276.98, P=0.04), and highly selective C-reactive protein (5.41 versus 3.55, P<0.001); a higher noncongestive section regeneration rate (25.8% versus 13.6%, P=0.012); and a higher recurrence rate (30.4% versus 9.7%, P=0.01) than group A. Graft congestion >10% [hazard ratio (HR) = 3.10, 95% confidence interval (CI) = 1.15-8.35, P=0.03], microvascular invasion (HR = 5.43, 95% CI = 2.04-14.44, P<0.01), and an alpha-fetoprotein level >200 IU/L (HR = 2.98, 95% CI = 1.10-8.03, P=0.03) were significantly related to tumor recurrence. Liver congestion may promote the recurrence of HCC after LDLT; therefore, it should be minimized. (C) 2014 AASLD.