Integrative Multi-Omics and Humanized Mouse Modelling to Predict Immunotherapy Response in Triple-Negative Breast Cancer.

INTRODUCTION Triple-Negative Breast Cancer (TNBC) is an aggressive variant of breast cancer, which has a high degree of molecular heterogeneity, essentially immunologically cold tumor phenotypes, and heterogeneous responses to immunotherapy. Although immune checkpoint inhibitors and combination immunotherapeutic approaches have broadened treatment opportunities, the activity of these treatments in patients is a key translational concern that is difficult to predict.

Methods

This review summarizes recent advances in multi-omics technologies, mouse models, and computational immuno-oncology to provide a broader picture of predicting immunotherapy outcomes in TNBC. The results of genomics, transcriptomics, proteomics, metabolomics, epigenomics, and single-cell multi-omics research were summarized, and advances made in humanized mouse models (including PBMC-, HSC-, and PDXbased models) were presented. In addition, new computational techniques, such as machine learning, deep learning, and network-based modelling, were tested.

Results

Multi-omics can provide insights into the immunobiology of TNBC, including tumor heterogeneity, interactions between the tumor microenvironment and the immune system, and mechanisms of immune resistance. These approaches illustrate pathways involving regulation of the immune system, tumor evolution, and therapeutic vulnerabilities that can be targeted. Humanized mouse models can, in part, recapitulate the human immune system and therefore offer the opportunity to measure responses to omics and immunotherapies and validate the measured biomarkers as functional.

Discussion

Combining multi-omics data with phenotypic data from the humanized model and complex computational models is an intriguing strategy for predicting oncologic outcomes. Such an integrated omics model AI ecosystem offers a route to better identify patients likely to benefit from immunotherapy and to guide more personalized treatment in TNBC. Realizing this potential, however, depends on clinical validation of the predictions and regulatory acceptance of the underlying models.

Conclusion

An integrated omics model AI approach, supported by clinical proof and regulatory acceptance, has the potential to improve the identification of immunotherapy responders and enable more personalized treatment in TNBC.