Assessment of the clinical value with the application of serum abnormal prothrombin for auxiliary diagnosis of hepatocellular carcinoma:a multicenter Chinese population-based case-control study
Objective To comprehensively evaluate the clinical value of Elecsys serum abnormal prothrombin(PIVKA-Ⅱ)test reagent for auxiliary diagnosis of hepatocellular carcinoma(HCC)in the Chinese population.Methods A multicenter case-control design was used in the study.Samples from patients with first-time confirmed,diagnosed,and untreated HCC,benign liver disease and interfering controls were collected continuously.Elecsys PIVKA-Ⅱ and alpha-fetoprotein(AFP)were tested for analysis.Various clinical details of the subjects were collected and analyzed.The efficacy of PIVKA-Ⅱ(21.29 ng/ml)and AFP(400 ng/ml)for HCC diagnosis was calculated at specific positive cut-off values.Statistical analysis was performed using the Kruskal-Wallis test or receiver operating characteristic curve.Results A total of 448 subjects were eventually enrolled from five centers,including 185 HCC cases.PIVKA-Ⅱ had a higher diagnostic sensitivity and accuracy than AFP(84.86％vs.30.81％and 89.01％vs.63.66％)when the benign liver disease group was used as the control group,while the specificity was slightly lower.A sensitive analysis showed that PIVKA-Ⅱ had a sensitivity of＞80％at this specific positive cut-off value in the subgroup of AFP-negative subjects,patients with different etiologies,and HCC patients with multiple Barcelona Clinic liver cancer stages(including early-stage HCC).At the same time,the PIVKA-Ⅱ subject had a slightly higher area under the receiver operating characteristic curve than the AFP(0.920 0 vs.0.880 9).Conclusion The clinical efficacy of Elecsys PIVKA-Ⅱ is good and stable in the Chinese population.Additionally,it has the clinical potential to improve the current missed diagnosis status of AFP-negative HCC and HCC monitoring at an early stage,as well as the effectiveness of accuracy promotion for HCC auxiliary diagnosis in China.
NETL
NSTL
万方数据
