Lasso-Based Nomogram for Predicting Early Recurrence Following Radical Resection in Hepatocellular Carcinoma.

Abstract

Background: Hepatocellular carcinoma (HCC) is a common malignancy with a high recurrence rate following curative resection. This study aimed to identify factors contributing to early recurrence (within 2 years) and develop a Lasso-based nomogram for individualized risk assessment.

Methods: We conducted a retrospective analysis of 206 hCC patients who underwent curative resection at Taizhou Hospital, Zhejiang Province, from January 2019 to August 2022. Patients were randomly divided into training (n=144) and validation (n=62) cohorts. Lasso regression was used to identify potential recurrence risk factors among 17 candidate predictors. A Cox proportional hazards model was constructed based on variables selected by Lasso. Model performance was assessed using receiver operating characteristic (ROC) curves, calibration plots, and decision curve analysis (DCA).

Results: Five independent predictors of early HCC recurrence were identified: age, serum alanine aminotransferase (ALT) levels, cirrhosis, tumor diameter, and microvascular invasion (MVI). The nomogram demonstrated area under the curve (AUC) values for recurrence-free survival (RFS) of 0.828 (95% confidence interval [CI]: 0.753-0.904) at 1 year, 0.799 (95% CI: 0.718-0.880) at 2 years, and 0.742 (95% CI: 0.642-0.842) at 5 years in the training cohort. The corresponding AUCs in the validation cohort were 0.823 (95% CI: 0.686-0.960), 0.804 (95% CI: 0.686-0.922), and 0.857 (95% CI: 0.722-0.992) at 1, 2 and 5 years, respectively. Calibration curves and DCA confirmed the nomogram's high accuracy and clinical utility.

Conclusion: The Lasso-Cox regression nomogram effectively predicts HCC recurrence within two years post-hepatectomy, providing a valuable tool for personalized postoperative management to improve patient outcomes.