Construction and Validation of a Predictive Model for Post-TACE Recurrence Risk in Hepatocellular Carcinoma: A...

Transarterial chemoembolization (TACE) is widely used for unresectable hepatocellular carcinoma (HCC) and as adjuvant therapy to prevent postoperative recurrence. However, accurate prognostic models for HCC patients undergoing TACE remain underutilized clinically.

This retrospective study included 265 HCC patients who underwent TACE, randomly assigned to training and validation sets (6:4 ratio). Recurrence-related risk factors were identified using Cox regression and further screened by Least Absolute Shrinkage and Selection Operator (LASSO) regression. A prognostic nomogram was constructed, with decision curve analysis (DCA) assessing its clinical utility.

Univariate Cox regression identified sex, vascular invasion, ascites, modified albumin-bilirubin grade (mALBI), China Liver Cancer (CNLC) staging, white blood cell count, neutrophils, fibrinogen (Fib), and neutrophil-to-lymphocyte ratio as significant risk factors (all HR > 1, P < 0.05). Albumin (ALB), prognostic nutritional index, and alkaline phosphatase-to-albumin ratio (AAPR) were protective factors (all HR < 1, P < 0.05). Multivariate analysis identified elevated Fib and higher CNLC stage as independent predictors (all P < 0.05). LASSO extracted vascular invasion, neutrophils, ALB, activated partial thromboplastin time, Fib, and AAPR as prognostic variables. Areas under the ROC curve at 6 months, 1 year, and 2 years were 0.717, 0.794, and 0.884, respectively. DCA demonstrated greater net clinical benefit than Barcelona Clinic Liver Cancer and CNLC staging systems. Risk stratification by nomogram tertiles showed significantly earlier recurrence in high-risk patients (all P < 0.05).

This TACE-based prognostic nomogram integrating clinical and laboratory parameters accurately predicts postoperative recurrence and enables individualized risk stratification, assisting clinicians in tailoring interventions and delivering personalized therapy.