Development of Risk Prediction Model for Hepatocellular Carcinoma Progression of Indeterminate Nodules in Hepatitis B Virus-Related Cirrhotic Liver

Abstract

OBJECTIVES: This study was performed to evaluate long-term outcome of indeterminate nodules detected on cirrhotic liver and to develop risk prediction model for hepatocellular carcinoma (HCC) progression of indeterminate nodules on hepatitis B virus (HBV)-related cirrhotic liver. METHODS: Indeterminate nodules up to 2 cm with uncertain malignant potential detected on computed tomography of cirrhotic liver during HCC surveillance were analyzed retrospectively. HCC risk prediction model of indeterminate nodules in HBV-related cirrhotic liver was deduced based on result of Cox regression analysis. RESULTS: A total of 494 indeterminate nodules were included. Independent risk factors of HCC progression were old age, arterial enhancement, large nodule size, low serum albumin level, high serum alpha-fetoprotein (AFP) level, and prior HCC history in all included subjects. In subjects with chronic hepatitis B, old age (year: hazard ratio (HR)=1.06: P<0.001), arterial enhancement (HR=2.62: P=0.005), large nodule size (>1 cm: HR=7.34: P<0.001), low serum albumin level (</=3.5 g/dl: HR=3.57: P=0.001), high serum AFP level (>/=100 ng/ml: HR=6.04: P=0.006), prior HCC history (HR=4.24: P=0.001), and baseline hepatitis B e antigen positivity (HR=2.31: P=0.007) were associated with HCC progression. We developed a simple risk prediction model using these risk factors and identified patients at low, intermediate, and high risk for HCC: 5-year cumulative incidences were 1%, 14.5%, and 63.1%, respectively. The developed risk score model showed good performance with area under the curve at 0.886 at 3 years, and 0.920 at 5 years in leave-one-out cross-validation. CONCLUSIONS: We developed a useful and accurate risk score model for predicting HCC progression of indeterminate nodules detected on HBV-related cirrhotic liver.