CT Radiomics Predicts β-Catenin Expression in HCC - EMJ

This site is intended for healthcare professionals

CT for prediction of nuclear β-catenin expression in HCC

Key Summary:

  • CT-based approaches could complement tissue assessment for pretreatment biological stratification.
  • CT radiomics showed external predictive value for nuclear β-catenin expression in HCC.
  • The stacking model ranked highest but did not clearly improve on radiomics alone.

The study included 1,029 participants with histopathologically confirmed hepatocellular carcinoma (HCC), divided into training, internal validation, and independent external validation cohorts. Nuclear β-catenin expression was present in 47.3% of participants in the training cohort, 47.2% in the internal validation cohort, and 30.3% in the external validation cohort. 

The researchers used portal venous phase CT images and existing whole-tumour segmentations. Tumour regions of interest were manually delineated using ITK-SNAP software by two experienced abdominal radiologists. 

Researchers developed clinical, radiomics, topology-based intratumoral heterogeneity (ITH), core ITH, combined radiomics–ITH and stacking models.  

CT Models Predict β-Catenin Expression in Liver Cancer 

In the external validation cohort, the CT radiomics model achieved an area under the receiver operating characteristic curve (AUC) of 0.862, while the topology-based ITH model achieved an AUC of 0.857. 

The clinical model had lower predictive performance. Combining radiomics and ITH features produced an AUC of 0.863, while the core ITH model achieved an AUC of 0.831. 

The stacking model, which combined predictions from the clinical, radiomics, and ITH models, achieved the highest external validation AUC at 0.886, but its improvement over the radiomics model alone was not statistically significant. 

Directly combining radiomics and ITH features also showed no significant improvement over radiomics alone. 

In the external validation cohort, the models did not fully preserve absolute-risk calibration, which the researchers noted supported the need for local calibration assessment before clinical use. 

Clinical Implications and Need for Further Validation 

Researchers concluded that CT radiomics and topology-based ITH features showed external predictive value for nuclear β-catenin expression in HCC. Although the stacking model achieved the highest external AUC, its improvement over radiomics alone was not statistically significant.  

Further validation and calibration in the intended clinical setting would be needed before these approaches could complement tissue assessment for pretreatment biological stratification. 

Reference: 

Tan Y et al. CT radiomics and topology-based intratumoral heterogeneity for preoperative prediction of nuclear β-catenin expression in hepatocellular carcinoma. Abdom Radiol . 2026. DOI;10.1007/s00261-026-05788-4. 

Featured Image: DC Studio on Adobe Stock 

Featured Image: DC Studio on Adobe Stock

Author:

Each article is made available under the terms of the Creative Commons Attribution-Non Commercial 4.0 License.

Rate this content's potential impact on patient outcomes

Average rating 0 / 5. Vote count: 0

No votes so far! Be the first to rate this content.