Contrast-enhanced ultrasound for precise differentiation of hepatitis-related intrahepatic cholangiocarcinoma and hepatocellular carcinoma

BACKGROUND: Chronic hepatitis viral infection is a recognized predisposing factor for both hepatocellular carcinoma (HCC) and intrahepatic cholangiocarcinoma (ICC). However, differentiating hepatitis-associated ICC from HCC remains challenging due to overlapping imaging presentations. Precise differentiation is crucial for optimizing therapeutic strategies and prognostic evaluation. This study aimed to differentiate hepatitis-associated ICC from HCC by integrating B-mode ultrasound (US) and contrast-enhanced ultrasound (CEUS) characteristics. METHODS: Using histopathological confirmation as the gold standard, we conducted a retrospective analysis of patients with chronic hepatitis who had ICC or HCC and underwent CEUS at the Second Hospital of Lanzhou University from July 2020 to February 2025. Based on pathological results, patients were stratified into hepatitis-associated HCC (n = 93) and hepatitis-associated ICC (n = 40) groups. Two radiologists retrospectively analyzed the clinical baseline characteristics and ultrasonographic features. Clinical baseline characteristics and ultrasonographic features were compared between groups. Independent predictors of hepatitis-associated ICC were identified through multivariate logistic regression analysis, and diagnostic performance was evaluated using receiver operating characteristic (ROC) curves. Internal validation was performed using bootstrap resampling (1,000 iterations). RESULTS: Multivariate logistic regression analysis revealed that hepatitis-associated ICC was independently predicted by hypoechoic lesions on B-mode US and the presence of enlarged lymph nodes. On CEUS, significant predictors included peripheral-to-center progressive filling (in the presence of washout), absence of late phase peritumoral rim-like enhancement, intratumoral venous branches in the portal phase, and shorter washout time. Compared to B-mode US alone (AUC = 0.865) or CEUS alone (AUC = 0.919), the combined use of B-mode US and CEUS demonstrated superior discriminatory power for differentiating hepatitis-associated HCC and ICC, achieving an apparent AUC of 0.957 and an optimism-corrected AUC of 0.914. Using the Youden index-derived cutoff, the model yielded 97.5% sensitivity and 63.4% specificity. CONCLUSION: The combination of B-mode US and CEUS features demonstrates superior diagnostic performance compared to either modality alone in differentiating hepatitis-associated ICC from HCC. This approach serves as a valuable adjunctive tool, particularly for clarifying diagnosis in equivocal cases or when cross-sectional imaging is contraindicated, thereby facilitating more precise therapeutic decision-making.