Multidimensional MRI radiomics-based model predicts recurrence risk in triple-negative breast cancer

Area of Science:

• Oncology
• Radiology
• Medical Informatics

Background:

• Triple-negative breast cancer (TNBC) presents a significant clinical challenge with poor prognoses.
• Current staging systems lack accuracy in predicting TNBC recurrence due to substantial tumor heterogeneity.
• There is a critical need for reliable prognostic assessment methods in TNBC management.

Purpose of the Study:

• To develop and validate a combined multimodal prognostic model for triple-negative breast cancer.
• To enhance the accuracy of recurrence risk assessment beyond traditional staging systems.
• To provide a more comprehensive evaluation of TNBC biological aggressiveness for personalized medicine.

Main Methods:

• Systematic integration of clinicopathological indicators and multi-scale radiomics features from a large TNBC cohort (n=700).
• Innovative incorporation of intratumoral habitat heterogeneity analysis and peritumoral microenvironment characteristics.
• Validation using consistency index (C-index) and time-dependent ROC analysis for disease-free survival (DFS).

Main Results:

• The combined multimodal model demonstrated superior prognostic capabilities compared to existing methods.
• Time-dependent ROC analysis confirmed robust Area Under the Curve (AUC) for 1-, 3-, and 5-year DFS.
• The model effectively captured tumor biological behavior and aggressiveness.

Conclusions:

• The developed predictive model offers a more comprehensive assessment of TNBC biological aggressiveness.
• This approach overcomes limitations of traditional staging systems in individualized prognosis.
• The model can assist clinicians in formulating personalized treatment and follow-up strategies for TNBC patients.