TEMs but not DKK1 could serve as complementary biomarkers for AFP in diagnosing AFP-negative hepatocellular carcinoma

Background & aims Hepatitis B virus (HBV)-related hepatocellular carcinoma (HCC) is prevalent worldwide. Despite its limitations, serum alpha-fetoprotein (AFP) remains the most widely-used bio-marker for the diagnosis of HCC. This study aimed to assess whether measurement of peripheral plasma Dickkopf-1 (DKK1) and Tie2-expressing monocytes (TEMs) could overcome the limitations of AFP and improve the diagnostic accuracy of HCC. Methods Plasma DKK1 level and the percentage of TEMs in peripheral CD14+CD16+monocytes from HCC patients (n = 82), HBV-related liver cirrhosis (LC) patients (n = 29), chronic hepatitis B (CHB) infected patients (n = 28) and healthy volunteers (n = 31) were analyzed by ELISA and flow cytometry. Receiver operating characteristic (ROC) curves were used to analyze a single biomarker, or a combination of two or three biomarkers. Univariate and multivariate analyses were performed to assess the significance of each marker in prediction of HCC and AFP-negative HCC from LC patients. Results The percentage of TEMs in peripheral CD14+CD16+monocytes and plasma level of DKK1 in HCC group were significantly higher than those in LC, CHB and healthy control groups (all P-values < 0.05). The percentage of TEMs alone was also significantly higher in AFP-negative HCC group than that in LC, CHB and healthy control groups (all P-values < 0.05). Plasma DKK1 level alone could not distinguish between AFP-negative HCC and LC patients. ROC curves showed that the optimal diagnostic cutoff value was 550.93 ng/L for DKK1 and 4.95% for TEMs. There was no significant difference in AUC of DKK1, TEMs and AFP in HCC diagnosis between the four groups (all P>0.05). A combination of DKK1, TEMs and AFP measurements increased the AUC for HCC diagnosis as compared with either marker alone (0.833; 95% CI 0.768-0.886). The AUC for TEMs was 0.692 (95% CI 0.564-0.819) in differentiating AFP-negative HCC from LC, with a sensitivity of 80.0% and a specificity of 65.52%. Only TEMs prevailed as a significant predictor for AFP-negative HCC differentiating from LC patients in univariate and multivariate analyses (P = 0.016, P = 0.023). Conclusions TEMs and DKK1 may prove to be potential complementary biomarkers for AFP in the diagnosis of HCC. TEMs rather than DKK1 could serve as a complementary biomarker for AFP in the differential diagnosis of AFP-negative HCC versus LC patients.