Dynamic evaluation of the aMAP score for predicting HBV-related HCC risk in a cohort of hospitalized Chinese patients using time-dependent ROC analysis
Institute of Genomics and Precision Medicine, Gannan Medical University, Ganzhou, Jiangxi Province, China
Ganzhou Key Laboratory of Interventional Medicine, Department of Interventional Radiology and Vascular Surgery, Ganzhou Hospital-Nanfang Hospital, Southern Medical University (Ganzhou People’s Hospital), Ganzhou, Jiangxi, China
Department of Hepatology, Institute of Hepatology, The Affiliated Fifth People’s Hospital of Ganzhou, Gannan Medical University, Ganzhou, Jiangxi Province, China
Henan Province Engineering Research Center for of Intractable Digestive Tract Tumor Precision Therapy, Henan Province Engineering Technology Research Center for Digestive Tract Tumor Precision Therapy, Zhengzhou Key Laboratory of Precision Therapy of Gastrointestinal Cancer, Department of Oncology, The Affiliated Cancer Hospital of Zhengzhou University and Henan Cancer Hospital, Zhengzhou, Henan Province, China
The Affiliated Ganzhou Hospital, Jiangxi Medical College, Nanchang University, Ganzhou, Jiangxi, China
Clin Exp HEPATOL 2026; 12, 4
Aim of the study
To validate the dynamic predictive performance of the aMAP risk score for hepatocellular carcinoma (HCC) development in hospitalized chronic hepatitis B (CHB) patients.
Material and methods
A retrospective cohort study was conducted to explore the predictive performance of the aMAP risk score for HCC development among 2914 CHB patients with at least 1 year of follow-up. The aMAP score (incorporating age, sex, albumin, bilirubin, and platelets) was calculated at baseline. Patients were stratified into low-, medium-, and high-risk groups using the X-tile method. Cox regression, Kaplan-Meier analysis, and time-dependent ROC analysis were used to assess the discriminative ability and predictive accuracy of the aMAP risk score for HCC.
Results
Multivariable analysis identified drinking status, cirrhosis, and a higher aMAP risk score as independent predictors of HCC. Kaplan-Meier curves showed significantly increased cumulative HCC incidence across risk groups (log-rank p < 0.0001). The aMAP score achieved an AUC of 0.837 (95% CI: 0.808-0.867) for baseline prediction, with minimal deviation in 2- or 5-year calibration. Time-dependent ROC analysis demonstrated higher discrimination at later time points (AUC: 0.782 at 24 months vs. 0.876 at 60 months).
Conclusions
A higher aMAP risk score was significantly associated with a higher risk of HBV-related HCC development, and showed satisfactory dynamic predictive performance for incident HCC.
Keywords
aMAP risk score, hepatocellular carcinoma, chronic hepatitis B virus, cohort study, time-dependent ROC
Coverage in
Integrated with
