BACKGROUND: The accurate staging of hepatocellular carcinoma (HCC) is important in establishing treatment strategies and prognosis. Among tumor factors, microvascular invasion, one of TNM staging components and prognostic factors, is underestimated preoperatively, due to inaccuracy of imaging modalities. We investigated preoperative predictors of microvascular invasion. METHODS: We reviewed 190 consecutive HCC patients given curative resection from 1999 to 2006. All were treatment-naive and monitored every 3 months after resection. Tumor recurrence, survivals, and clinicopathological factors associated with microvascular invasion were analyzed. RESULTS: The 5-year disease-free survival (DFS) rate was 39.4%(median follow-up duration: 35 months). On resection pathology, 38.9% (74/190 patients) had microvascular invasion undetected preoperatively, using liver spiral computed tomography (CT) or angiography. Independent predictors of microvascular invasion were tumor size (P = 0.043), number (P = 0.011), and Edmondson grade (P = 0.001). Patients with Edmondson grade 1 and size <5 cm had no microvascular invasion, while those with grade > or =2 had higher incidences (7/18 patients, 38.8%) even in small tumor (<2 cm). When tumors recurred, presence of microvascular invasion independently increased incidences of multiple tumors, portal vein invasion, and diffuse-infiltrative patterns significantly. CONCLUSIONS: Preoperative predictors of microvascular invasion are tumor size, number, and Edmondson grade, which may be useful for making clinical decisions in both non-surgical and surgical candidates.
BACKGROUND: The accurate staging of hepatocellular carcinoma (HCC) is important in establishing treatment strategies and prognosis. Among tumor factors, microvascular invasion, one of TNM staging components and prognostic factors, is underestimated preoperatively, due to inaccuracy of imaging modalities. We investigated preoperative predictors of microvascular invasion. METHODS: We reviewed 190 consecutive HCC patients given curative resection from 1999 to 2006. All were treatment-naive and monitored every 3 months after resection. Tumor recurrence, survivals, and clinicopathological factors associated with microvascular invasion were analyzed. RESULTS: The 5-year disease-free survival (DFS) rate was 39.4%(median follow-up duration: 35 months). On resection pathology, 38.9% (74/190 patients) had microvascular invasion undetected preoperatively, using liver spiral computed tomography (CT) or angiography. Independent predictors of microvascular invasion were tumor size (P = 0.043), number (P = 0.011), and Edmondson grade (P = 0.001). Patients with Edmondson grade 1 and size <5 cm had no microvascular invasion, while those with grade > or =2 had higher incidences (7/18 patients, 38.8%) even in small tumor (<2 cm). When tumors recurred, presence of microvascular invasion independently increased incidences of multiple tumors, portal vein invasion, and diffuse-infiltrative patterns significantly. CONCLUSIONS: Preoperative predictors of microvascular invasion are tumor size, number, and Edmondson grade, which may be useful for making clinical decisions in both non-surgical and surgical candidates.