Magnetic resonance imaging and diffusion-weighted imaging-based histogram analyses in predicting glypican 3-positive hepatocellular carcinoma. Issue 139 (June 2021)

• Record Type: Journal Article
• Authors: Zhao, Jiangtao; Gao, Shanshan; Sun, Wei; Grimm, Robert; Fu, Caixia; Han, Jing; Sheng, Ruofan; Zeng, Mengsu
• Is Part Of: European journal of radiology. Issue 139(2021)
• Publication Date: 2021-06
• Subjects: HCC Hepatocellular carcinoma -- GPC3 Glypican-3 -- MRI magnetic resonance imaging -- DWI diffusion-weighted imaging -- LI-RADS Liver Imaging Reporting and Data System -- ADC apparent diffusion coefficient -- ICC interclass correlation coefficient -- CI confidence interval -- AUCs areas under the curve -- AFP alpha-fetoprotein -- MVI microvascular invasion; Glypican-3 -- Hepatocellular carcinoma -- Histogram -- Nomogram; Medical radiology -- Periodicals; Radiology -- Periodicals; Radiologie médicale -- Périodiques; Medical radiology; Periodicals; 616.075705
• DOI: 10.1016/j.ejrad.2021.109732
• Languages: English
• ISSNs: 0720-048X

Abstract

Highlights: Elevated AFP and lower 75th percentile ADC were helpful in identifying GPC3(+) HCC. The combined nomogram achieved satisfactory risk prediction of GPC3 expression in HCC. The combined nomogram may provide effective treatment guidance for HCC management. Abstract: Purpose: We aimed to investigate the potential MR imaging findings in predicting glypican-3 (GPC3)-positive hepatocellular carcinomas (HCCs), with special emphasis on diffusion-weighted imaging (DWI)-based histogram analyses. Methods: Forty-three patients with pathologically-confirmed GPC3-negative HCCs and 100 patients with GPC3-positive HCCs were retrospectively evaluated using contrast-enhanced MRI and DWI. Clinical characteristics and MRI features including DWI-based histogram features were assessed and compared between the two groups. Univariate and multivariate analyses were used to identify the significant clinico-radiologic variables associated with GPC3 expressions that were then incorporated into a predictive nomogram. Nomogram performance was evaluated based on calibration, discrimination, and decision curve analyses. Results: Features significantly related to GPC3-positive HCCs at univariate analyses were serum alpha-fetoprotein (AFP) levels >20 ng/mL ( P < 0.0001), absence of enhancing capsule ( P = 0.040), peritumoral enhancement appearance on the arterial phase ( P = 0.049), as well as lower mean ( P = 0.0278), median ( P = 0.0372) and 75th percentile ( P = 0.0085) apparent diffusion coefficient (ADC) values. At multivariate analysis, the AFP levels (odds ratio, 11.236; P < 0.0001) and 75th percentile ADC values (odds ratio, 1.009; P = 0.033) were independent risk factors associated with GPC3-positive HCCs. When both criteria were combined, both sensitivity (79.0 %) and specificity (79.1 %) greater than 75 % were achieved, and satisfactory predictive nomogram performance was obtained with a C-index of 0.804 (95 % confidence interval, 0.729−0.866). Decision curve analysis further confirmed the clinical usefulness of the nomogram. Conclusions: Elevated serum AFP levels and lower 75th percentile ADC values were helpful in differentiating GPC3-positive and GPC3-negative HCCs. The combined nomogram achieved satisfactory preoperative risk prediction of GPC3 expression in HCC patients.