Preoperative prediction tertiary lymphoid structures of hepatocellular carcinoma on gadoxetate disodium-enhanced MRI
10.3760/cma.j.cn112149-20250415-00197
- VernacularTitle:钆塞酸二钠增强MRI术前预测肝细胞癌瘤内三级淋巴结构的研究
- Author:
Lin CHEN
1
;
Yiman LI
;
Jie CHENG
;
Fengxi CHEN
;
Ping CAI
;
Wei CHEN
;
Qingrui LI
;
Huarong ZHANG
;
Xiaoming LI
Author Information
1. 陆军军医大学第一附属医院7 T磁共振转化医学研究中心/放射科，重庆 400038
- Publication Type:Journal Article
- Keywords:
Carcinoma, hepatocellular;
Magnetic resonance imaging;
Tertiary lymphoid structures;
Gadoxetate disodium
- From:
Chinese Journal of Radiology
2025;59(6):674-680
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To evaluate the efficacy of gadolinium ethoxybenzyl- diethy-lenetriamine pentaacetic acid (Gd-EOB-DTPA) enhanced MRI features in the preoperative prediction of tertiary lymphoid structures (TLS) within hepatocellular carcinoma (HCC) lesions.Methods:This retrospective cross-sectional study included clinical and pathological data from 297 HCC patients treated at the Southwest Hospital, Army Medical University between June 2021 and November 2022. Based on postoperative pathology, patients were categorized into TLS-negative ( n=93) and TLS-positive ( n=204) groups. MRI features of HCC lesions using Gd-EOB-DTPA enhancement and relevant clinical data were analyzed. Intergroup comparisons of imaging features and laboratory findings were performed using independent sample t-test, Mann-Whitney U test, χ2 test, or Fisher exact test, as appropriate. The logistic regression analysis was conducted to identify independent predictors of TLS positivity. A predictive model was constructed and visualized using a nomogram. The model′s predictive performance and clinical utility were assessed using the receiver operating characteristic (ROC) curves, calibration curves, and decision curve analysis (DCA). The area under the ROC curve (AUC) was compared using the DeLong test. Results:Significant differences were observed between the TLS-negative and TLS-positive groups in alpha-fetoprotein (AFP) levels, intratumoral hemorrhage, and peritumoral satellite nodules in the hepatobiliary phase ( P<0.05). Multivariate logistic regression identified intratumoral hemorrhage ( OR=0.123, 95% CI 0.070-0.216, P<0.001) and peritumoral satellite nodules in the hepatobiliary phase ( OR=0.236, 95% CI 0.093-0.596, P=0.002) as independent predictive factors for TLS-positivity. The imaging model based on these two features yielded an AUC of 0.764 (95% CI 0.709-0.809) for predicting TLS-positivity. When combined with AFP levels, the resulting clinical-imaging model achieved a superior AUC of 0.784 (95% CI 0.732-0.829), which was significantly higher than that of the imaging model alone ( Z=2.20, P=0.028). A nomogram was constructed based on the clinical-imaging model. The calibration curve demonstrated good predictive performance of the nomogram, and the DCA showed that the curve remained above the default line across a range of reasonable threshold probabilities, indicating that patients could derive clinical benefit. Conclusion:A nomogram model based on Gd-EOB-DTPA enhanced MRI features combined with AFP levels can effectively predict the presence of TLS in HCC.
