Seoul St. Mary’s Hospital Team Develops AI Model to Predict Risk and Guide Treatment for Liver Cancer Patients Before Therapy
Professor Hajin Won's Team at Seoul St. Mary's Hospital
Early Assessment of Treatment Risks
Demonstrated Potential for Improved Outcomes
A research team at Seoul St. Mary’s Hospital has developed an artificial intelligence model that predicts the risk of sudden deterioration of liver function in hepatocellular carcinoma patients before the initiation of systemic therapy.
The research group led by Professor Hajin Won of the Department of Gastroenterology at Seoul St. Mary’s Hospital, Catholic University of Korea, analyzed data from 2,026 hepatocellular carcinoma patients treated at eight hospitals under the Catholic Medical Center from 2010 to 2024, and built the Machine Learning-based Hepatic Safety Score (MHSS) model. This model comprehensively incorporates blood test results, liver function indicators, platelet counts, tumor size and number, presence of vascular invasion, and tumor markers, and can be used to select safe and effective personalized treatment options for patients.
Professor Jaewon Han, Department of Gastroenterology, Seoul St. Mary's Hospital. Seoul St. Mary's Hospital
View original imageTraditionally, liver function has been evaluated using tools such as the Child-Pugh score (an indicator of cirrhosis severity based on bilirubin, albumin, and coagulation levels), ALBI (a simple indicator calculated using only albumin and bilirubin), MELD (a scoring system used to prioritize liver transplantation in patients with end-stage liver disease), and FIB-4 (an index estimating liver fibrosis based on age and blood test results). However, these tools focus primarily on blood test results and assess only liver function, without accounting for tumor characteristics such as size or the presence of vascular invasion, which is a notable limitation.
In contrast, the newly developed AI-based tool incorporates both liver function indicators and tumor-related information, and was found to predict variceal bleeding and post-treatment liver function deterioration more accurately than existing tools. The model also demonstrated stable results in an independent validation cohort composed of patients from other institutions.
Patients classified as high-risk by this model had a 3.25-fold higher risk of liver function deterioration during treatment, a 4.90-fold higher risk of variceal bleeding, and a 2.21-fold higher risk of mortality compared to the low-risk group. This demonstrates, through AI analysis, that deterioration in liver function during treatment is not determined solely by baseline liver function scores, but is a result of complex factors such as tumor size and vascular invasion by cancer cells.
The research team’s analysis of treatment selection by medication also yielded meaningful insights. Simulation results showed that, among low-risk patients, a specific immunotherapy combination (Atezolizumab-Bevacizumab) provided a superior survival benefit compared to other therapies. However, among high-risk patients, the increased risk of variceal bleeding meant that this survival benefit was not as apparent.
Based on these findings, the research team implemented a personalized treatment simulation that prioritized this immunotherapy combination for low-risk patients, and recommended therapies with relatively lower bleeding risk for high-risk patients. As a result, the simulation predicted a 24% reduction in the risk of liver function deterioration, a 40% reduction in the risk of variceal bleeding, and a 26% reduction in all-cause mortality compared to the non-application group.
This study is considered significant in that it provides clinical evidence not only for determining "which treatment is most effective" for liver cancer patients, but also for assessing "which treatment is safest for which patient." The findings are expected to enable more precise application of pre-treatment endoscopic evaluation and bleeding prevention strategies, as well as tailored planning of treatment intensity and follow-up for individual patients.
Professor Hajin Won, who led the study, stated, "This research is meaningful in that it establishes an objective foundation for proposing a safe and rational treatment pathway for each patient by comprehensively evaluating tumor characteristics, liver function, and the risk of portal hypertension within a single AI tool," and added, "We plan to further develop this into a practical, personalized precision medicine tool that can be used in clinical practice to help patients continue treatment safely, through prospective studies and validation with diverse datasets."
Meanwhile, this study was supported by research funding from the Global Physician-Scientist Training Program of the Ministry of Health and Welfare and the Korea Health Industry Development Institute, and was published in the prestigious international journal 'npj Digital Medicine' (IF 18.0) in the field of digital healthcare and medical artificial intelligence.
Hot Picks Today
"Two Spend 500,000 Won in a Flash, Yet Wait Two Hours?" ... 'Dreamland' Shunned by Kids and Foreign Tourists
- The Era of Getting 200 Million Won After 30 Years Is Over... A Japanese Bank's "Perpetual Department Head" Who Just Clocked In Cries Out
- Spain Hit by Its Worst Tragedy... Lee Kangin, Moreno and Real Madrid Join Soccer World in Mourning
- They Said, “Make It Your Home for Life” - Then Suddenly Kicked 200 Passengers Off... What Happened to the Around-the-World Cruise?
- "130 Rice Bowls Sold Out in a Day"... 'Unfiltered Face' Adds 8 Million Won to Night Market Sales: China's Bungeunwoo
The research team has also released the developed predictive model as a web-based calculator, making it freely available to improve clinical accessibility for both patients and healthcare professionals.
© The Asia Business Daily. All rights reserved. Unauthorized AI training and use prohibited.