A preoperative nomogram with MR elastography in identifying cytokeratin 19 status of hepatocellular carcinoma

Yanan Gu; Kaipu Jin; Shanshan Gao; Wei Sun; Minyan Yin; Jing Han; Yunfei Zhang; Xiaolin Wang; Mengsu Zeng; Ruofan Sheng

doi:10.1093/bjr/tqae193

A preoperative nomogram with MR elastography in identifying cytokeratin 19 status of hepatocellular carcinoma

Br J Radiol. 2024 Dec 6:tqae193. doi: 10.1093/bjr/tqae193. Online ahead of print.

Authors

Yanan Gu^{1

2

3}, Kaipu Jin^{4

1}, Shanshan Gao^{1

2}, Wei Sun^{1

2}, Minyan Yin^{2

1}, Jing Han⁵, Yunfei Zhang², Xiaolin Wang^{3

2}, Mengsu Zeng^{1

2

6}, Ruofan Sheng^{1

2}

Affiliations

¹ Department of Radiology, Zhongshan Hospital, Fudan University, Shanghai, 200032, China.
² Shanghai Institute of Medical Imaging, 200032, Shanghai, China.
³ Department of Interventional Radiology, Zhongshan Hospital, Fudan University, Shanghai, 200032, China.
⁴ Shanghai Geriatric Medical Center, Zhongshan Hospital, Fudan University, Shanghai, 201104, China.
⁵ Department of Pathology, Zhongshan Hospital, Fudan University, Shanghai, 200032, China.
⁶ Department of Cancer Center, Zhongshan Hospital, Fudan University, 200032, Shanghai, China.

PMID: 39657213
DOI: 10.1093/bjr/tqae193

Abstract

Objectives: Developing a nomogram integrating MR elastography (MRE)-based tumor stiffness and contrast-enhanced MRI in identifying cytokeratin 19 (CK19) status of hepatocellular carcinoma (HCC) preoperatively.

Methods: 120 CK19-negative HCC and 39 CK19-positive HCC patients undergoing curative resection were prospectively evaluated. All received MRE and contrast-enhanced MRI. Clinical and MRI tumor features were compared. Univariate and multivariate logistic regression analyses identified independent predictors for CK19 status. Receiver operating characteristic curve analysis evaluated diagnostic performance. A nomogram was established with calibration and decision curve analysis.

Results: Multivariate analysis revealed serum alpha fetoprotein (AFP) level (P < 0.001), targetoid appearance (P = 0.007), and tumor stiffness (P = 0.011) as independent significant variables for CK19-positive HCC. The area under the curve for tumor stiffness was 0.729 (95% CI 0.653, 0.796). Combining these features, a nomogram-based model achieved an area under the curve value of 0.844 (95% CI 0.778, 0.897), with sensitivity, specificity, and accuracy of 76.92%, 85.00%, and 83.02%, respectively. Calibration and decision curve analyses demonstrated good agreement and optimal net benefit.

Conclusions: MRE-measured tumor stiffness aids in predicting CK19 status in HCC. The combined nomogram incorporating tumor stiffness, targetoid appearance, and AFP provides a reliable biomarker for CK19-positive HCC.

Advances in knowledge: MRE-measured tumor stiffness can be used to predict CK19 status in HCC. The nomogram, which integrates tumor stiffness, targetoid appearance, and AFP levels, has shown enhanced diagnostic performance. It offers a comprehensive preoperative tool for clinical decision-making, further advancing personalized treatment strategies in HCC management.

Keywords: Cytokeratin 19; magnetic resonance elastography; ：hepatocellular carcinoma.