Objectives: Liver cirrhosis is a major risk factor for hepatocellular carcinoma (HCC), and all liver study societies recommend HCC surveillance in patients with cirrhosis. However, no ideal modality for HCC surveillance has been determined. The aim of this study is to assess the effectiveness of α-fetoprotein (AFP) measurement in HCC surveillance.
Methods: In this retrospective analysis, all patients with cirrhosis, who received HCC surveillance through ultrasound (US) and AFP measurement between January 2002 and July 2010, were followed up until June 2013. The performance effectiveness of surveillance using AFP, US, or both in HCC detection was compared.
Results: Overall, 1,597 patients were followed for a median duration of 4.75 (range 1.42-12) years. Over the 8563.25-person-year follow-up period, 363 patients (22.7%) developed HCCs. For HCC detection, the area under the receiver operator characteristic curve of surveillance AFP was 0.844 (95% confidence interval: 0.820-0.868, P<0.001). When the traditional cutoff value of 20 ng/ml was used, the sensitivity and specificity of AFP were 52.9% and 93.3%, respectively. US exhibited a sensitivity and specificity of 92.0% and 74.2%, respectively. A combination of US and AFP exhibited a sensitivity and specificity of 99.2% and 68.3%, respectively. By using cut-off at 20 ng/ml and AFP level increase ≥2 × from its nadir during the previous 1 year, the combination of US and AFP yielded a sensitivity of 99.2% and an improved specificity of 71.5%.
Conclusions: The complementary use of AFP and US improved the effectiveness of HCC surveillance in patients with cirrhosis.