Imaging features based on CT and MRI for predicting prognosis of patients with intrahepatic cholangiocarcinoma: a single-center study and meta-analysis

SiameseNet based on multiple instance learning for accurate identification of the histological grade of ICC tumors

Background

After hepatocellular carcinoma (HCC), intrahepatic cholangiocarcinoma (ICC) is the second most common primary liver cancer. Timely and accurate identification of ICC histological grade is critical for guiding clinical diagnosis and treatment planning.

Method

We proposed a dual-branch deep neural network (SiameseNet) based on multiple-instance learning and cross-attention mechanisms to address tumor heterogeneity in ICC histological grade prediction. The study included 424 ICC patients (381 in training, 43 in testing). The model integrated imaging data from two modalities through cross-attention, optimizing feature representation for grade classification.

Results

In the testing cohort, the model achieved an accuracy of 86.0%, AUC of 86.2%, sensitivity of 84.6%, and specificity of 86.7%, demonstrating robust predictive performance.

Conclusion

The proposed framework effectively mitigates performance degradation caused by tumor heterogeneity. Its high accuracy and generalizability suggest potential clinical utility in assisting histopathological assessment and personalized treatment planning for ICC patients.

Associations between Dynamic Contrast-Enhanced Magnetic Resonance Imaging with Histopathological Features in Cholangiocarcinoma

INTRODUCTION

The relationships between histopathology and imaging remain elusive, and investigating the underlying reasons for tumor microstructure leading to an imaging phenotype is of clinical importance. In the present study, a cross-sectional guided biopsy specimen was used to correlate prebioptic magnetic resonance imaging (MRI) with immunohistochemical staining of the histopathologic specimen using precise spatial biopsy localization.

METHODS

Twenty-seven patients with mass-forming cholangiocarcinoma (CCA) were included in the present analysis. All patients were imaged with a 1.5 T clinical scanner at least 1 month prior to biopsy. The contrast-enhanced dynamic sequences were analyzed with quantified signal intensities. The bioptic specimens were obtained by cross-sectional guided biopsy and further analyzed for cell density, proliferation index (Ki67), tumor-infiltrating lymphocytes, tumor-stroma ratio (TSR), and collagen.

RESULTS

There were no statistically significant correlations between MRI signal intensities and cell count, TSR, Ki67 index, and CD45 count. Only a moderate correlation was observed between relative signal intensities of the venous phase and the collagen-stained area (r = 0.40, p = 0.04).

CONCLUSION

DCE-MRI is not associated with histopathological features in CCA. The complex interactions of tumor and tumor microenvironment are not reflected in the MRI phenotype.

Diagnosis of intrahepatic cholangiocarcinoma with CEUS

Intrahepatic cholangiocarcinoma (ICC) is a rare, heterogeneous, highly lethal tumor of the biliary tract. Due to the lack of effective treatments, an early identification of ICC is essential to achieve the best outcome in terms of therapy and prognosis aiming for a curative intent. ICC may arise on a normal liver or with an underlying liver disease, making the diagnosis more difficult and complex. Contrast-enhancement ultrasound (CEUS) is an accurate procedure able to detect ICC-specific contrast vascular pattern, and thus facilitating the correlation between radiological and histopathological findings with high specificity and sensitivity. CEUS has been shown to have a high diagnostic potential in the diagnosis of ICC thanks to the possibility of studying in real time the intralesional microcirculation and evaluating the precocity of the enhancement of the lesion during the arterial phase. All these features allow to differentiate the ICC from hepatocarcinoma (HCC) with high sensitivity and specificity. Furthermore, CEUS is a rapid, non-invasive, non-nephrotoxic or non-allergenic tool. The only limitations CEUS may have are related to the disease site and patient characteristics (obesity) and compliance, including the operator's experience. A clinical evaluation of the patient, together with tumor markers and biochemical tests assessment, to differentiate ICC from HCC are highly suggested.