Purpose: To develop and validate a preoperative nomogram model that incorporates clinical factors, serological markers and liver imaging reporting and data system (LI-RADS v2018) MRI features for predicting early recurrence (ER) of hepatocellular carcinoma (HCC) treated with transarterial chemoembolization (TACE).
Methods: One hundred and fourteen patients with HCC who underwent MRI scanning before TACE were enrolled retrospectively and divided into a training cohort (n=80) and a test cohort (n=34). The clinical factors, serological markers and LI-RADS v2018 MRI features associated with ER were determined by univariable and multivariable analyses. A nomogram model predicting ER after TACE was developed, and its discriminatory ability, goodness-of-fit and clinical application were evaluated by receiver operating characteristic (ROC) curve, calibration curve and decision curve analysis (DCA), respectively.
Results: In total, 74 (64.9%) patients were diagnosed with ER according to the follow-up results. Increased alpha fetoprotein (AFP) levels, larger tumor size, nonsmooth margin, mosaic architecture satellite nodules and corona enhancement were independent predictors associated with ER (p < 0.05). For the established nomogram model that incorporated these six significant predictors, the AUC values were 0.94 (95% CI: 0.89-0.99) and 0.95 (95% CI: 0.88-1.00) for predicting ER after TACE in the training and test cohorts, respectively. The calibration curve and DCA results demonstrate the good goodness-of-fit and clinical benefits of this nomogram.
Conclusion: A preoperative nomogram model based on serological markers and LI-RADS v2018 MRI features could adequately predict ER in HCC patients after TACE, which may provide personalized guidance for predicting prognosis.
Keywords: Hepatocellular carcinoma (HCC); early recurrence (ER); liver imaging reporting and data system (LI-RADS); nomogram; transarterial chemoembolization (TACE).
Copyright © 2022 The Association of University Radiologists. Published by Elsevier Inc. All rights reserved.