Histologic classification of microscopic portal venous invasion to predict prognosis in hepatocellular carcinoma

SummaryPortal venous invasion is one of the most important prognostic factors after surgical resection of hepatocellular carcinoma. Microscopic portal venous invasion can be evaluated histologically. We examined 280 hepatocellular carcinomas with microscopic portal venous invasion (n = 125) or without it (n = 155) for 3 characteristics: the number of invaded portal vessels, the maximum number of invading carcinoma cells, and the farthest distance from the tumor. Univariate analysis of overall and disease-free survival revealed that the number of invaded portal vessels and the number of invading carcinoma cells were poor prognostic factors. Therefore, we classified patients with microscopic portal venous invasion into 2 groups: a high–microscopic portal venous invasion group, in which there were multiple invaded portal venous vessels (≥2) and more than 50 invading carcinoma cells (n = 57), and a low–microscopic portal venous invasion group, in which microscopic portal venous invasion was observed but with invasion of only a single portal venous vessel or fewer than 50 invading carcinoma cells (n = 68). The high–microscopic portal venous invasion group showed significantly higher α-fetoprotein levels, larger tumor size, and higher frequencies of poorly differentiated histology, capsule infiltration, and intrahepatic metastasis compared with the low–microscopic portal venous invasion group (P = .0496, P < .0001, P = .0431, P = .0180, and P = .0012, respectively). The high–microscopic portal venous invasion group showed poorer overall survival and disease-free survival rates than the low–microscopic portal venous invasion group (P = .0004 and P = .0003), and the high–microscopic portal venous invasion group was an independent prognostic factor for disease-free survival (P = .0259). We proposed a new definition for classifying microscopic portal venous invasion and documented the necessity of definite histologic evaluation of it.