Abstract
PURPOSE: Hepatocellular carcinoma (HCC) represents a significant global health problem, requiring precise prognostic tools for optimal treatment stratification. This study aimed to develop a new risk prediction score, called AD score, based on the serum markers alpha-fetoprotein (AFP) and des-γ-carboxy prothrombin (DCP), to offer an objective and accurate preoperative assessment of HCC in patients undergoing transarterial chemoembolization (TACE). PATIENTS AND METHODS: This was a retrospective study that included 295 HCC patients who were subjected to TACE (training set, n=147; testing set, n=148). Serum AFP and DCP levels were log-transformed to construct the AD score. Multivariate Cox regression analysis on cirrhosis subgroups validated the objectivity of the model. Performance comparison of established models (Child Pugh, BCLC, ALBI, Up-to-seven, Six-and-twelve, Four and seven, HAP score, mHAP-II, FAIL-T score), was assessed through time-dependent receiver operating characteristic (ROC) curves and risk stratification. RESULTS: The AD score, incorporating lgAFP and lgDCP, demonstrated superior predictive accuracy than the existing models. Time-dependent ROC curve revealed the consistent superiority of the AD score over a 5-year period. The risk stratification into low, intermediate, and high group based on the AD score showed a significant survival difference in both training and testing set. CONCLUSION: For HCC patients undergoing TACE, the AD score serves as an objective and straightforward prognostic tool, enhancing predictive accuracy and showcasing its clinical utility. It demonstrates potential significance as a crucial addition to preoperative risk assessment for TACE.