摘要Background:Diagnostic panels based on multiple biomarkers and clinical characteristics are considered more favorable than individual biomarker to diagnose hepatocellular carcinoma(HCC).Based on age,sex,alpha-fetoprotein(AFP),and protein induced by vitamin K absence Ⅱ(PIVKA-Ⅱ)with/without AFP-L3,ASAP and GALAD models are potential diagnostic panels.The diagnostic performances of these two panels were compared relative to HCC detection among patients with various etiologies of chronic liver diseases(CLDs).Methods:A multicenter case-control study recruited CLDs patients with and without HCC from 14 Chi-nese hospitals.The etiologies of CLDs included hepatitis B virus(HBV),hepatitis C virus(HCV),alcoholic liver disease(ALD),and nonalcoholic fatty liver disease(NAFLD).Using area under the receiver operating characteristic curve(AUC)values,the diagnostic performances of ASAP and GALAD models were com-pared to detect HCC among patients with various etiologies of CLDs.Results:Among 248 HCC patients and 722 CLD controls,the ASAP model demonstrated the highest AUC(0.886)to detect HCC at any stage,outperforming the GALAD model(0.853,P=0.001),as well as any individual biomarker(0.687-0.799,all P＜0.001).In the subgroup analysis of various CLDs etiologies,the ASAP model outperformed the GALAD model to HCC independent of CLDs etiology.In addition,the ASAP model performed better in detecting early-stage(BCLC stage O/A)HCC versus the GALAD model.Conclusions:Despite using one less laboratory variable(AFP-L3),the ASAP model demonstrated better diagnostic performance than the GALAD model to detect all-stage HCC among patients with various eti-ologies of CLDs-related HCC.