Dynamic Contrast-Enhanced Computed Tomography Findings that may Predict Poorly-Differentiated Hepatocellular Carcinoma Prior to Treatment

Abstract

Objective: To identify CT findings that predict poorly-differentiated hepatocellular carcinoma (p-HCC). Methods: This retrospective study included pathologically proven HCC patients during January 2010 to December 2017 who underwent dynamic contrast-enhanced computed tomography (CT) imaging within 12 weeks before the pathological diagnosis. CT findings were reviewed and graded by consensus opinion of two abdominal radiologists. The relationship between imaging findings and histological differentiation of HCC was analyzed using chi-square test. Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy for diagnosis of p-HCC were calculated. Results: Of 200 HCCs during the study period, 18 were well-differentiated, 170 were moderately-differentiated, and 12 were poorly differentiated. Irregular rim enhancement in arterial phase (p<0.001) and presence of lymphadenopathy (p=0.003) were both statistically significantly different among the three types of histological differentiation of HCC. Sensitivity, specificity, PPV, NPV, and accuracy for prediction of p-HCC by the presence of irregular rim enhancement in arterial phase and lymphadenopathy were 58.3%, 97.3%, 58.3%, 97.3%, and 95%, and 50%, 88.8%, 22.2%, 96.5%, and 86.5% - all respectively. Conclusion: The presence of irregular rim enhancement in arterial phase and lymphadenopathy are potentially useful CT findings for prediction of p-HCC prior to treatment.

Keywords

• Computed tomography
• histological differentiation
• poor differentiation
• hepatocellular carcinoma