Related Experiment Video
Updated: Sep 16, 2025

A Three-Dimensional Spheroid Model to Investigate the Tumor-Stromal Interaction in Hepatocellular Carcinoma
Published on: September 30, 2021
A Novel Risk Prediction Model for Hepatocellular Carcinoma in MASLD: A Multinational, Multicenter Cohort Study
Ho Soo Chun1, Minjong Lee1, Hye Ah Lee2
1Department of Internal Medicine, Ewha Womans University College of Medicine, Seoul, Republic of Korea; Department of Internal Medicine, Ewha Womans University Medical Center, Seoul, Republic of Korea.
Background & Aims:
It is unclear that which cardiometabolic risk factors (CMRFs) are significantly associated with hepatocellular carcinoma (HCC) development in metabolic dysfunction-associated steatotic liver disease (MASLD). We aimed to develop and validate a novel CMRF-based HCC risk prediction model in MASLD.
Methods:
This multicenter cohort study recruited 77,677 MASLD patients from 20 medical centers in Korea and other Asian and Western countries (2004-2023). A novel CMRF-based HCC risk prediction model (MASLD-HCC score) was developed based on time-varying Cox multivariable analysis in a training cohort (n = 36,800, Korea), which was validated internally (n = 36,799, Korea) and externally (n = 4078, 11 other Asia and Western countries).
Results:
In the training cohort, 71 (0.2%) patients developed HCC (median follow-up 5.1 years). Overweight/obesity or central obesity and prediabetes/diabetes were independently associated with HCC development, along with age, sex, and platelets. The MASLD-HCC score with these 5 risk factors showed a Harrell's C-index of 0.84 for HCC development, which was maintained in the internal (C-index 0.83) and external validation cohorts (C-index 0.93), and the model was well calibrated. Decision curve analyses showed that patients had positive net benefits from the model. When stratified by the MASLD-HCC score, the risk of HCC development in the high-risk group was significantly higher than the low-risk group (training: subdistribution hazard ratio [sHR], 11.44; 95% confidence interval [CI], 7.10-18.41; internal validation: sHR, 12.36; 95% CI, 7.72-19.79; external validation: sHR, 56.84; 95% CI, 12.88-250.73; all P < .001).
Conclusions:
Overweight/obesity or central obesity and prediabetes/diabetes with fibrotic burden were significantly associated with the increased HCC risk in MASLD. The MASLD-HCC score may enable physicians to stratify the HCC risk.
More Related Videos
04:09Predicting Treatment Response to Image-Guided Therapies Using Machine Learning: An Example for Trans-Arterial Treatment of Hepatocellular Carcinoma
Published on: October 10, 2018
07:13Comparison of Predictive Performance of Three Lymph Node Staging Systems in Colorectal Signet Ring Cell Carcinoma Based on Machine Learning Model
Published on: April 18, 2025