Differentiating tumor vs bland portal vein thrombosis (PVT) is essential in determining liver transplant (LT) candidacy for patients with hepatocellular carcinoma (HCC). The primary aim of this study was to evaluate cross-sectional imaging characteristics combined with clinical parameters to develop non-invasive criteria that could reliably differentiate tumor vs bland PVT. Contrast-enhanced abdominal CT or MRI in patients with PVT were independently reviewed by an experienced Radiologist blinded to the original diagnosis and patient outcome, to specifically evaluate the following diagnostic criteria for tumor PVT: enhancement of thrombus, venous expansion, neo-vascularity and continuity with HCC or treatment site. Of the 470 consecutive patients with HCC listed for LT between January 2004 and February 2011, 62 (13%) were found to have PVT, and 13 of 62 (21%) PVT were deemed to be from tumor invasion. Concordance with the independent radiologic review was 100% in the diagnosis of tumor vs bland PVT. When comparing tumor vs. bland PVT, thrombus enhancement was seen in 100% vs 9%, venous expansion in 92% vs 11%, neo-vascularity in 58% vs 2%, and continuity with HCC or treatment site in 100% vs 21%, respectively (all p<0.001). When combining these 4 imaging characteristics with alpha-fetoprotein (AFP) >1000 ng/dL at time of PVT, the presence of ≥3 criteria best characterized tumor PVT (p<0.001) with 100% sensitivity, 94% specificity, 80% positive predictive value (PPV), and 100% negative predictive value (NPV). Of these 5 criteria, neo-vascularity had the highest PPV of 88% while each of the 5 criteria had NPV >90%. No patient with diagnosis of tumor PVT underwent LT and no patient with bland PVT by these criteria had macro-vascular invasion on explant. Bland PVT was not associated with micro-vascular invasion (p=0.69), and did not adversely impact post-LT outcome. The post-LT survival at 1 and 5 years for patients with bland PVT vs those without PVT was 97% vs 93% and 75% vs 78%, respectively (p=0.43). Recurrence-free probabilities at 1 and 5 years were 91% vs 96% and 85% vs 85%, respectively (p=0.72). In conclusion, in LT candidates with HCC, non-invasive criteria based on ≥3 of the following: thrombus enhancement, venous expansion, neo-vascularity, continuity with either HCC or treatment site, and AFP > 1000 ng/dL, can accurately differentiate tumor vs bland PVT.

« Back to 2015 American Transplant Congress