Analysis of contrast-enhanced ultrasound features of hepatocellular adenoma according to different pathological molecular classifications.

OBJECTIVE To explore the specific contrast-enhanced ultrasound (CEUS) features of hepatocellular adenomas (HCA) according to their pathological molecular classifications. METHODS & MATERIALS In this retrospective study, fifty-three histopathologically proved HCA lesions (mean size, 39.7±24.9 mm) were included. Final histopathological diagnosis of HCA lesions were identified by surgical resection (n = 51) or biopsy (n = 2) specimens. CEUS imaging features were compared among four subgroups according to World Health Organization (WHO) 2019 pathological molecular classifications standards. Analysis of variance (ANOVA) were used for statistical analysis of continuous variables. Fisher's exact test were used for categorical variables. The sensitivity (SE), specificity (SP), and accuracy of CEUS feature in diagnosis of each HCA subtype were calculated and compared. RESULTS Final histopathological diagnosis included HNF-1α inactivated HCAs (H-HCA, n = 12), β-catenin activated HCAs (B-HCA, n = 8), inflammatory HCAs (I-HCA, n = 31), and unclassified HCAs (U-HCA, n = 2). During arterial phase of CEUS, all HCAs were hyper-enhanced, 66.6% (8/12) of H-HCAs and 50% (4/8) of B-HCAs displayed complete hyperenhancement, whereas 58.0% (18/31) of I-HCAs showed centripetal filling hyperenhancement pattern (P = 0.016). Hyper-enhanced subcapsular arteries could be detected in 64.5% (20/31) I-HCAs during early arterial phase. During portal venous and late phase, sustained hyper- or iso-enhancement were observed in 91.7% (11/12) of H-HCAs, while most of I-HCAs (61.3%, 19/31) and B-HCAs (7/8, 87.5%) were hypo-enhanced (P = 0.000). Central unenhanced areas were most commonly observed in I-HCAs (29.0%, 9/31) (P = 0.034). CONCLUSION Depending on its unique imaging features including enhancement filling pattern, hyper-enhanced subcapsular artery and presence of washout, CEUS might provide helpful diagnostic information for preoperative prediction of various HCA molecular subtypes.