The Emergence of the Potential Therapeutic Targets: Ultrasound-Based Radiomics in the Prediction of Human Epidermal...

Area of Science:

• Radiology
• Oncology
• Medical Imaging

Background:

• HER2 expression is crucial for breast cancer (BC) treatment decisions.
• Accurate prediction of HER2 expression, particularly HER2-low, is vital for personalized therapy.
• Conventional imaging methods have limitations in predicting HER2 status.

Purpose of the Study:

• To evaluate the efficacy of ultrasound (US)-based radiomics features in predicting HER2-low expression in breast cancer patients.
• To develop and validate a radiomics score (RS) for HER2 expression prediction using conventional US images.

Main Methods:

• A multicenter study included 383 patients (training), 233 (validation), and 149 (external validation).
• Radiomics features were extracted from conventional US images.
• Minimum redundancy maximum relevancy and least absolute shrinkage and selector operation algorithms were used to generate an US-based radiomics score (RS).
• Multivariable logistic regression and receiver operating characteristic curve analysis (AUC) were employed to assess diagnostic performance.

Main Results:

• The RS demonstrated strong performance in predicting HER2-low expression across all datasets.
• AUC values for differentiating HER2-zero from HER2-low/positive cases were 0.81 (training), 0.84 (validation), and 0.82 (external validation).
• The RS showed good performance in subgroup analyses, distinguishing HER2-zero from HER2 1+, 2+, and HER2-low tumors (AUC range: 0.79-0.87).

Conclusions:

• The US-based radiomics score (RS) is an effective tool for predicting HER2-low expression in breast cancer.
• This non-invasive imaging-based approach can aid in treatment selection for breast cancer patients.
• Radiomics offers a promising avenue for improving the prediction of HER2 status in breast cancer.