Objective To develop and evaluate nomograms incorporating clinical features and quantitative CT parameters for the preoperative prediction of microvascular invasion (MVI) and pathological grading in patients with hepatocellular carcinoma (HCC). Materials and methods This retrospective multicenter study involved 684 consecutive patients with pathologically confirmed HCC. In this context, 553 patients from Center 1 were randomly split into training (70%) and internal validation (30%) cohorts, and 131 patients from Center 2 served as an external validation cohort. Predictive factors for MVI-positive and high-grade HCC were identified through univariate and multivariate logistic regression. Two nomograms combining clinical factors and CT quantitative parameters were developed. Results For MVI prediction, multivariate analysis identified HBV infection, an AFP concentration > 20 ng/mL, a larger tumor diameter, a higher arterial absolute enhancement value (AAEV), and a lower portal relative enhancement ratio (PRER) as independent predictors. The nomogram achieved area under the curve (AUC) values of 0.769 (95% CI: 0.720–0.818), 0.771 (0.692–0.850), and 0.760 (0.648–0.872) in the training, internal validation, and external validation cohorts, respectively. For pathological grade prediction, younger age, an AFP concentration > 20 ng/mL, a larger tumor diameter, and a lower PRER were independent predictors. The corresponding nomograms had AUCs of 0.696 (0.639–0.752), 0.644 (0.547–0.741), and 0.768 (0.681–0.856) across cohorts. Both nomograms demonstrated excellent calibration and significant clinical utility for decision curve analysis. Conclusion Nomograms integrating clinical features and quantitative CT parameters facilitate accurate preoperative prediction of MVI status and pathological grading in HCC patients, demonstrating strong potential for clinical implementation. Key Points <Emphasis Type="BoldItalic">Question</Emphasis> Predicting microvascular invasion (MVI) and pathological grading preoperatively in hepatocellular carcinoma (HCC) patients using a clinically practical and cost-effective method remains challenging. <Emphasis Type="BoldItalic">Findings</Emphasis> Quantitative parameters derived from contrast-enhanced CT demonstrated the ability to predict MVI and pathological grade, with validated models demonstrating robust diagnostic performance and clinical applicability. <Emphasis Type="BoldItalic">Clinical relevance</Emphasis> Nomograms, which integrate clinical features and quantitative CT parameters, represent a clinically applicable tool for the preoperative prediction of MVI status and pathological grading in HCC patients, facilitating personalized treatment planning and optimized patient management. Graphical Abstract

Quantitative analysis of enhanced CT in predicting microvascular invasion and pathological grading of hepatocellular carcinoma

期刊论文
发表日期：

2025年10月31日

查看全文

Dan Shi,
Qian Kou,
Hai-Yan Chen,
Jie-Yu Chen,
Lu-Fei Jin,
Chun-Yan Jiang,
Lei Shi,
Ri-Sheng Yu,
Chao Wang

摘要

Objective

To develop and evaluate nomograms incorporating clinical features and quantitative CT parameters for the preoperative prediction of microvascular invasion (MVI) and pathological grading in patients with hepatocellular carcinoma (HCC).

Materials and methods

This retrospective multicenter study involved 684 consecutive patients with pathologically confirmed HCC. In this context, 553 patients from Center 1 were randomly split into training (70%) and internal validation (30%) cohorts, and 131 patients from Center 2 served as an external validation cohort. Predictive factors for MVI-positive and high-grade HCC were identified through univariate and multivariate logistic regression. Two nomograms combining clinical factors and CT quantitative parameters were developed.

Results

For MVI prediction, multivariate analysis identified HBV infection, an AFP concentration > 20 ng/mL, a larger tumor diameter, a higher arterial absolute enhancement value (AAEV), and a lower portal relative enhancement ratio (PRER) as independent predictors. The nomogram achieved area under the curve (AUC) values of 0.769 (95% CI: 0.720–0.818), 0.771 (0.692–0.850), and 0.760 (0.648–0.872) in the training, internal validation, and external validation cohorts, respectively. For pathological grade prediction, younger age, an AFP concentration > 20 ng/mL, a larger tumor diameter, and a lower PRER were independent predictors. The corresponding nomograms had AUCs of 0.696 (0.639–0.752), 0.644 (0.547–0.741), and 0.768 (0.681–0.856) across cohorts. Both nomograms demonstrated excellent calibration and significant clinical utility for decision curve analysis.

Conclusion

Nomograms integrating clinical features and quantitative CT parameters facilitate accurate preoperative prediction of MVI status and pathological grading in HCC patients, demonstrating strong potential for clinical implementation.

Key Points

Question Predicting microvascular invasion (MVI) and pathological grading preoperatively in hepatocellular carcinoma (HCC) patients using a clinically practical and cost-effective method remains challenging.

Findings Quantitative parameters derived from contrast-enhanced CT demonstrated the ability to predict MVI and pathological grade, with validated models demonstrating robust diagnostic performance and clinical applicability.

Clinical relevance Nomograms, which integrate clinical features and quantitative CT parameters, represent a clinically applicable tool for the preoperative prediction of MVI status and pathological grading in HCC patients, facilitating personalized treatment planning and optimized patient management.

Graphical Abstract