[Rational imaging of hepatocellular carcinoma. The challenge of multimodal diagnostic criteria]

Computed Tomography Radiomics to Differentiate Intrahepatic Cholangiocarcinoma and Hepatocellular Carcinoma

AIMS

Intrahepatic cholangiocarcinoma (iCCA) and hepatocellular carcinoma (HCC) differ in prognosis and treatment. We aimed to non-invasively differentiate iCCA and HCC by means of radiomics extracted from contrast-enhanced standard-of-care computed tomography (CT).

MATERIALS AND METHODS

In total, 94 patients (male, n = 68, mean age 63.3 ± 12.4 years) with histologically confirmed iCCA (n = 47) or HCC (n = 47) who underwent contrast-enhanced abdominal CT between August 2014 and November 2021 were retrospectively included. The enhancing tumour border was manually segmented in a clinically feasible way by defining three three-dimensional volumes of interest per tumour. Radiomics features were extracted. Intraclass correlation analysis and Pearson metrics were used to stratify robust and non-redundant features with further feature reduction by LASSO (least absolute shrinkage and selection operator). Independent training and testing datasets were used to build four different machine learning models. Performance metrics and feature importance values were computed to increase the models' interpretability.

RESULTS

The patient population was split into 65 patients for training (iCCA, n = 32) and 29 patients for testing (iCCA, n = 15). A final combined feature set of three radiomics features and the clinical features age and sex revealed a top test model performance of receiver operating characteristic (ROC) area under the curve (AUC) = 0.82 (95% confidence interval =0.66-0.98; train ROC AUC = 0.82) using a logistic regression classifier. The model was well calibrated, and the Youden J Index suggested an optimal cut-off of 0.501 to discriminate between iCCA and HCC with a sensitivity of 0.733 and a specificity of 0.857.

CONCLUSIONS

Radiomics-based imaging biomarkers can potentially help to non-invasively discriminate between iCCA and HCC.

Dual-energy CT based material decomposition to differentiate intrahepatic cholangiocarcinoma from hepatocellular carcinoma

PURPOSE

To assess the potential of material decomposition in dual-energy CT (DECT) to differentiate intrahepatic cholangiocarcinoma (iCCA) from hepatocellular carcinoma (HCC).

METHOD

In this retrospective study, we included 94 patients (26 female (27.7 %), median age 64.5 (interquartile range 55.5-74.5) years) with either iCCA or HCC who underwent abdominal contrast-enhanced DECT in arterial phase. To test for differences between iCCA (n = 47) and HCC (n = 47), we evaluated mean attenuation and DECT material density values including iodine density (ID), normalized iodine uptake (NIU), fat fraction, and lesion-to-liver parenchyma ratio. Histopathology served as reference standard for all lesions. We used univariate logistic regression models for the outcome iCCA versus HCC. ROC curve analysis was applied to assess discriminative ability of the model. Model accuracy was evaluated by calculating the Brier score. Youden index was applied to establish thresholds to differentiate between iCCA and HCC.

RESULTS

Comparison of quantitative image parameters revealed significant differences between iCCA and HCC for ID (1.6 ± 0.5 mg/ml vs 2.8 ± 0.8 mg/ml, p < 0.001), NIU (14.5 ± 4.8 vs 24.8 ± 10.3, p < 0.001), attenuation (41.9 ± 10.1 HU vs 47.9 ± 8.9 HU, p = 0.003), and fat fraction (12.0 ± 7.8 % vs 9.0 ± 6.4 %, p = 0.045). ROC curve analysis revealed highest ability to differentiate iCCA from HCC for ID (AUC = 0.93, 95 % CI 0.89-0.98). For ID, an optimal threshold of 2.33 mg/dl was determined to discriminate between iCCA and HCC (sensitivity 89.4 %, specificity 76.6 %).

CONCLUSIONS

DECT-based iodine quantification can serve as a tool for the differentiation of iCCA and HCC in contrast-enhanced CT. ID yielded the highest diagnostic performance and may assist in clinical routine CT diagnostics.