Background: The accuracy of des-γ -carboxyprothrombin (DCP) in the detection of hepatocellular carcinoma (HCC) in those infected hepatitis B virus (HBV) from cross-sectional or case-control studies is contradictory.
Objective: To resolve this contradiction using a prospective study.
Methods: Three hundred male individuals persistently infected with HBV were recruited from the Chinese cohort and followed up once per year from 2012 to 2015. Each subject was screened for HCC by measurements of serum alpha-fetoprotein (AFP), lectin-bound α -fetoprotein (AFP-L3), DCP concentrations and ultrasonographic examinations.
Results: Nineteen HCC cases were identified. The area under receiver operating characteristic (AUROC) at first, second and third visit for AFP, AFP-L3 and DCP ranges from 0.710-0.897, 0.566-0.637 and 0.520-0.595, respectively. The rate of elevated DCP is not significantly different between the HCC cases and controls (52.6% vs. 47.4%) (P > 0.05). The incidence of HCC in subjects with elevated DCP is not significantly higher than that of those with normal DCP (9.5% vs. 4.6%) (P > 0.05). The AUROC of combinations of these biomarkers was higher than that of AFP alone at the first visit. However, it was reduced at the second visit. At the third visit, the AUROCs of AFP + DCP and AFP + AFP-L3 + DCP, but not that of AFP + AFP-L3, were higher than that of AFP alone.
Conclusions: AFP but DCP or AFP-L3 remains a valuable biomarker for HCC in those chronically infected with HBV. The combination with AFP-L3 and DCP may not increase the accuracy of AFP in differentiating HCC cases from controls, among those infected with HBV.
Keywords: Hepatocellular carcinoma; hepatitis B virus; serum biomarker; the area under receiver operating characteristic.