Prediction of microvascular invasion in HCC by a scoring model combining Gd-EOB-DTPA MRI and biochemical indicators
Kun Zhang1, Shuang-Shuang Xie2, Wen-Cui Li3
1Department of Radiology, First Central Clinical College, Tianjin Medical University, 22 Qixiangtai Road, Heping District, Tianjin, 300070, China.
Objectives:
This study aimed to establish a reliable diagnostic scoring model for the preoperative prediction of microvascular invasion (MVI) in hepatocellular carcinoma (HCC) patients based on gadolinium-ethoxybenzyl-diethylenetriamine pentaacetic acid (Gd-EOB-DTPA)-enhanced magnetic resonance imaging (MRI) and biochemical indicators.
Methods:
This retrospective study included 129 patients with HCC at our hospital from 2014 to 2020. Based on the intratumoral and peritumoral features on Gd-EOB-DTPA MRI and biochemical indicators, a scoring model was developed for preoperative prediction of MVI, and examined for diagnostic efficacy according to postoperative pathological results. The scoring model was further externally validated in an independent cohort of 63 HCC patients.
Results:
Logistic regression analysis was performed to identify five parameters related to MVI, including maximum tumor diameter, peritumoral low intensity in the hepatobiliary phase, incomplete capsule, apparent diffusion coefficient (ADC), and [alkaline phosphatase (ALP) (U/L) + gamma-glutamyl transpeptidase (GGT) (U/L)] / lymphocyte count (× 109/L) ratio (AGLR). Based on these five parameters, a scoring model was developed, and the accuracy, sensitivity, specificity, PPV, and NPV in predicting MVI were 93.6%, 94.7%, 93.2%, 85.7%, and 97.6%, respectively, with a score > 8 set as the threshold.
Conclusion:
The scoring model based on Gd-EOB-DTPA MRI and biochemical indicators provides a reliable tool for preoperative prediction of MVI in HCC patients.
Key Points:
• The scoring model based on Gd-EOB-DTPA MRI and biochemical indicators is practical for preoperative prediction of MVI in HCC patients. • AGLR is an independent risk factor for MVI. • The scoring model could help implement more appropriate interventions, potentially leading to precise and individualized treatments based on the biological characteristics of the tumor.
