Background & aims: Liver stiffness measurement (LSM) via vibration-controlled transient elastography (VCTE) accurately assesses fibrosis. We aimed to develop a universal risk score for predicting hepatocellular carcinoma (HCC) development in patients with chronic hepatitis.
Methods: We systematically selected predictors and developed the risk prediction model (HCC-LSM) in the HBV training cohort (n = 2,251, median follow-up of 3.2 years). The HCC-LSM model was validated in an independent HBV validation cohort (n = 1,191, median follow-up of 5.7 years) and a non-viral chronic liver disease (CLD) extrapolation cohort (n = 1,189, median follow-up of 3.3 years). A HCC risk score was then constructed based on a nomogram. An online risk evaluation tool (LEBER) was developed using ChatGPT4.0.
Results: Eight routinely available predictors were identified, with LSM levels showing a significant dose-response relationship with HCC incidence (P < .001 by log-rank test). The HCC-LSM model exhibited excellent predictive performance in the HBV training cohort (C-index = 0.866) and the HBV validation cohort (C-index = 0.852), with good performance in the extrapolation CLD cohort (C-index = 0.769). The model demonstrated significantly superior discrimination compared to six previous models across the three cohorts. Cut-off values of 87.2 and 121.1 for the HCC-LSM score categorized participants into low-, medium-, and high-risk groups. An online public risk evaluation tool (LEBER; http://ccra.njmu.edu.cn/LEBER669.html) was developed to facilitate the use of HCC-LSM.
Conclusion: The accessible, reliable risk score based on LSM accurately predicted HCC development in patients with chronic hepatitis, providing an effective risk assessment tool for HCC surveillance strategies.
© The Author(s) 2024. Published by Oxford University Press. All rights reserved. For permissions, please email: journals.permissions@oup.com.