Detailed Abstract
[Liver Best Oral Presentation - Liver (Liver Disease/Surgery)]
[LV BO 5] DEVELOPMENT AND EXTERNAL VALIDATION OF AN ONLINE RISK CALCULATOR FOR EARLY MORTALITY AFTER HEPATIC RESECTION IN HEPATOCELLULAR CARCINOMA
Ren-Jie ZHANG 1, Si-Yuan WANG 2, Jia-Hao XU 1, Yu-Ze YANG 3, Yong-Yi ZENG 4, Wei-Min GU 5, Ya-Hao ZHOU 6, Hong WANG 7, Ting-Hao CHEN 8, Xian-Ming WANG 9, Ying-Jian LIANG 10, Jie LI 11, Yong-Kang DIAO 1, Feng SHEN 1, Tian YANG 1
1 Department of Hepatobiliary Surgery, Eastern Hepatobiliary Surgery Hospital, Naval Medical University, China, 2 Hepatopancreatobiliary Center, Beijing Tsinghua Changgung Hospital, China, 3 Department of Hepatobiliary And Pancreatic Surgery, General Surgery Centre, First Hospital of Jilin University, China, 4 Department of Hepatobiliary Surgery, Mengchao Hepatobiliary Hospital, Fujian Medical University, China, 5 The First Department of General Surgery, The Fourth Hospital of Harbin, China, 6 Department of Hepatobiliary Surgery, Pu’er People’s Hospital, China, 7 Department of General Surgery, Liuyang People’s Hospital, China, 8 Department of General Surgery, Ziyang First People’s Hospital, China, 9 Department of General Surgery, First Affiliated Hospital of Shandong First Medical University & Shandong Provincial Qianfoshan Hospital, China, 10 Department of Hepatobiliary Surgery, First Affiliated Hospital of Harbin Medical University, China, 11 Department of Hepatobiliary Surgery, Fuyang People’s Hospital, China
Background : Despite improvements in surgical techniques, early mortality after hepatic resection for hepatocellular carcinoma (HCC) remains a critical concern. Current staging systems and liver function classifications lack individualized risk prediction capability to guide surgical decision-making.
Methods : A multi-institutional retrospective analysis was conducted across 11 hepatobiliary centers in China (2011-2021). The training cohort comprised patients from 9 centers, with the remaining 2 centers serving as external validation. Predictors of postoperative 90-day mortality were identified through multivariable logistic regression analysis. Model performance underwent evaluation using concordance index and calibration plots.
Results : Among 4,966 patients analyzed, the 90-day mortality rate was 4.1%. The final model incorporated six preoperative predictors: performance status, prothrombin time, albumin-bilirubin grade, aspartate aminotransferase to platelet ratio index, tumor burden score, and gross vascular invasion. The model demonstrated excellent discrimination in both training (C-index 0.816) and validation (C-index 0.801) cohorts, significantly outperforming traditional HCC staging systems (AJCC and BCLC) and liver function classifications (P<0.001). Using an optimized cutoff of 54 points, the model achieved 79.0% sensitivity and 71.8% specificity for identifying high-risk patients.
Conclusions : The validated online calculator enables accurate prediction of early mortality risk after hepatic resection for HCC. This practical tool may facilitate surgical candidate selection and support informed decision-making by providing individualized risk estimates. Further validation in international cohorts is warranted to confirm broad applicability.
Methods : A multi-institutional retrospective analysis was conducted across 11 hepatobiliary centers in China (2011-2021). The training cohort comprised patients from 9 centers, with the remaining 2 centers serving as external validation. Predictors of postoperative 90-day mortality were identified through multivariable logistic regression analysis. Model performance underwent evaluation using concordance index and calibration plots.
Results : Among 4,966 patients analyzed, the 90-day mortality rate was 4.1%. The final model incorporated six preoperative predictors: performance status, prothrombin time, albumin-bilirubin grade, aspartate aminotransferase to platelet ratio index, tumor burden score, and gross vascular invasion. The model demonstrated excellent discrimination in both training (C-index 0.816) and validation (C-index 0.801) cohorts, significantly outperforming traditional HCC staging systems (AJCC and BCLC) and liver function classifications (P<0.001). Using an optimized cutoff of 54 points, the model achieved 79.0% sensitivity and 71.8% specificity for identifying high-risk patients.
Conclusions : The validated online calculator enables accurate prediction of early mortality risk after hepatic resection for HCC. This practical tool may facilitate surgical candidate selection and support informed decision-making by providing individualized risk estimates. Further validation in international cohorts is warranted to confirm broad applicability.
SESSION
Liver Best Oral Presentation
Room 1 3/29/2025 10:10 AM - 11:10 AM