Prediction of dual-phenotype hepatocellular carcinoma based on ultrasound and contrast enhanced ultrasound - a preliminary study.

Abstract

Aims: To investigate the diagnostic value of ultrasound features with dual-phenotype hepatocellular carcinoma (DPHCC). Material and methods: A retrospective analysis was conducted on patients diagnosed with hepatocellular carcinoma (HCC) confirmed by pathology results from October 2016 to December 2023. Patients were categorized into DPHCC and non-DPHCC (NDPHCC) groups based on immunohistochemical findings. Clinical baseline characteristics and ultrasound features were compared between these groups. Multivariable logistic regression was employed to identify predictive factors. A logisticregression model and nomogram were developed to predict DPHCC.

Results: This study included 219 HCC patients (average age: 55.84±9.00, 173 males), with 61 cases of DPHCC and 158 cases of NDPHCC. The results of multivariate logistic regression indicated that halo sign (OR=1.93, p=0.049), PVP hypo-enhancement (OR=2.73, p=0.002) and nodule-in-nodule sign (OR=4.23, p=0.003) were independent predictors of DPHCC. The model's AUC value was 0.73 (95%CI: 0.65-0.80). The calibration curve displayed good predictive accuracy.

Conclusions: Ultrasound features, including CEUS, may provide additional information for the non-invasive diagnosis of DPHCC.