Validation of combined AFP, AFP-L3, and PIVKA II for diagnosis and monitoring of hepatocellular carcinoma in Chinese patients

BACKGROUND: In this study, we aimed to investigate the performance of GALAD, GALAD-C, and GAAP models in Chinese population in comparison to our newly build statistical model. METHODS: In this study, we built the AALP model based on age, α-fetoprotein (AFP), AFP-L3, and prothrombin induced by vitamin K absence-II (PIVKA II) to differentiate between patients with HCC and patients with CLD. We then compared the serum levels of AFP-L3 and PIVKA II in patients with HCC who were defined as remission or progression and showed the prognostic value of combined biomarkers. RESULTS: The AUC value of the AALP model for HCC detection was 0.939 and AALP model exhibited a sensitivity of 81 % and a high specificity of 95 %. AALP model also exhibited good performance in the subgroups of patients with CLD. Furthermore, we demonstrated the consistency between imaging results and serum levels of AFP-L3 and PIVKA II. CONCLUSIONS: The AALP model achieved a good diagnostic performance and a high sensitivity for predicting HCC patients. Our research also showed that AFP-L3 and PIVKA II are complementary to each other but irreplaceable in the clinical detection and monitoring of HCC.