Spatial-temporal radiogenomics in predicting neoadjuvant chemotherapy efficacy for breast cancer: a comprehensive review

Abstract

Radiomics is undergoing a paradigm shift from single-omics to multi-omics, from single-temporal to multi-temporal analysis, and from global to subregional analysis. These transformations have shown great potential in addressing key challenges related to imaging changes before and after neoadjuvant chemotherapy (NAC) in breast cancer. Furthermore, radiomics has achieved remarkable progress in tasks such as exploring tumor heterogeneity and uncovering underlying biological mechanisms. Integrating imaging data with gene data offers novel perspectives for understanding imaging changes driven by specific genetic alterations. However, current radiomics studies on neoadjuvant chemotherapy for breast cancer have not yet achieved a close integration of imaging changes with underlying biological mechanisms. They are largely limited to simple associations between models and genomic data, without in-depth interpretation of the biological significance inherent in imaging features, which is essential to directly link these features with the dynamic progression of the disease. This review seeks to explore the spatial-temporal heterogeneity of imaging alterations observed during NAC for breast cancer, while assessing their biological implications using established analytical approaches. It highlights the distinct advantages of spatial-temporal radiomics in predictive model development and examines potential correlations between imaging dynamics and gene expression profiles before and after NAC. Additionally, we critically examines previous radiogenomics studies, providing theoretical insights into their limitations. Finally, the review proposes future directions and innovative approaches for applying spatial-temporal radiogenomics in NAC for breast cancer, serving as a valuable reference and roadmap for researchers and clinical practitioners in this field.

Keywords: Breast cancer; Magnetic resonance imaging; Neoadjuvant chemotherapy; Radiogenomics; Spatial-temporal heterogeneity.

Fig. 1

Workflow of mutiomics model. (A) Radiomics Workflow Diagram: A diagram illustrating the steps involved in radiomics analysis. (B) Pathomics Workflow Diagram: A diagram showing the process of pathomics analysis. (C) Genomics Workflow Diagram: A diagram outlining the steps in genomics data analysis. (D) Metadata Diagram: A diagram displaying the structure of associated metadata. (E) Multi-Omics Data Diagram: A diagram representing the integration of multi-omics data

Fig. 2

Timeline of Spatial-Temporal heterogeneity on radiomics and radiogenomics

Fig. 3

The structure of Spatial-Temporal heterogeneity in radiomics. A. Spatial Heterogeneity: Tumor Subregions; B. Temporal Heterogeneity: Tumor Shrinkage Before and After Neoadjuvant Chemotherapy; C. Spatial-Temporal Heterogeneity: Four Response States of Tumors After Neoadjuvant Chemotherapy and Subregional Changes; D. Spatial-Temporal Radiomics-Genomics Integrated Model. Diagram D uses partial responders after neoadjuvant therapy from Diagram C as an example, integrating tumor subregional changes before and after neoadjuvant chemotherapy with genetic information to provide a biological interpretation of these changes. (Complete Response, CR; Partial Response, PR; Stable Disease, SD; Progressive Disease, PD)