Science Translational Medicine | Peripheral Blood Transcriptomic Analysis Predicts Breast Cancer Subtypes and Response to Neoadjuvant Chemotherapy and Immunotherapy

This study reveals the potential of peripheral blood immune features in predicting response to chemoimmunotherapy in breast cancer patients, offering new insights for the development of dynamic biomarkers based on liquid biopsy in related fields.

 

Literature Overview

The article titled "Peripheral blood transcriptional profiling predicts tumor subtype and neoadjuvant chemoimmunotherapy outcomes in human breast cancer," published in the journal Science Translational Medicine, systematically investigates the dynamic changes in the transcriptome of peripheral blood mononuclear cells (PBMCs) during neoadjuvant chemotherapy or chemotherapy combined with immunotherapy in high-risk stage II/III HER2-negative breast cancer patients. Using RNA sequencing, the authors tracked systemic immune responses at multiple time points and constructed a multiparameter immune biomarker model that successfully predicted outcomes of treatment with pembrolizumab and dostarlimab. This study highlights the value of peripheral blood as a 'liquid biopsy' in precision oncology, providing a new avenue for non-invasive monitoring of anti-tumor immunity.

Background Knowledge

1. The study addresses a critical unmet need in breast cancer: the lack of reliable non-invasive biomarkers to identify patients who will benefit from immunotherapy, especially in HER2-negative cases. Although PD-L1 and tumor-infiltrating lymphocytes (TILs) have been explored, their predictive power is limited and dependent on tumor tissue biopsies, making dynamic monitoring difficult. 2. While immune checkpoint inhibitors (ICIs) have shown efficacy in triple-negative breast cancer (TNBC) and some hormone receptor-positive (HR+) patients, only a subset respond. Existing biomarkers such as tumor mutational burden (TMB) and microsatellite instability (MSI) have low mutation rates in breast cancer and are not widely applicable. 3. The study's approach leverages transcriptomic changes in PBMCs as a 'window' into systemic immune status, using longitudinal sampling to capture early immune activation signals associated with pathologic complete response (pCR). This strategy avoids the invasiveness of repeated tissue biopsies and may more comprehensively reflect the systemic anti-tumor immune state. Key molecules such as TCR diversity, IFN-γ signaling pathway, GZMB, and PRF1 are revealed as potential markers of effector T-cell activity, while dynamic changes in MDSCs and monocytes suggest regulation of immunosuppressive microenvironments.

 

 

Research Methods and Experiments

The authors performed bulk RNA sequencing on 546 longitudinal peripheral blood samples from 160 patients, collected at four time points: baseline (T0), early treatment (T1), mid-treatment (T2), and end of treatment (T3). The patient cohort was derived from the paclitaxel control arm and the pembrolizumab plus paclitaxel arm of the I-SPY2 clinical trial, with pCR as the primary efficacy endpoint. Differential gene expression analysis, GSEA enrichment analysis, and CIBERSORTx deconvolution algorithms were used to systematically analyze immune feature differences between breast cancer subtypes (HR+ vs. TNBC) and treatment response groups. Additionally, TCR clonality analysis was performed to assess changes in T-cell receptor diversity. The study further validated the constructed multiparameter immune scoring model in an independent cohort (n=59) using samples from patients treated with oral paclitaxel plus dostarlimab for external validation.

Key Conclusions and Perspectives

• In triple-negative breast cancer patients, those with higher baseline TCR diversity and rapid clonal expansion after treatment are more likely to achieve pCR, indicating that the breadth of the pre-existing T-cell repertoire is a key determinant of immunotherapy response.
• TNBC patients who respond to chemoimmunotherapy show significant enrichment of GZMB+ cytotoxic CD8 T cells in peripheral blood, and their composite score increases post-treatment, suggesting that effector T-cell activation is directly associated with clinical benefit.
• In HR+ breast cancer, treatment responders exhibit decreased monocyte abundance and reduced leukocyte chemotaxis signaling, suggesting that immunotherapy may reshape the tumor microenvironment by suppressing myeloid-derived suppressor cells (e.g., MDSCs).
• Based on baseline and early-treatment immune features (including T-cell composite score, IFN-γ signaling, TCR clonality, etc.), the predictive model successfully distinguished responders from non-responders to dostarlimab in the independent cohort (AUC=0.65), demonstrating its generalizability across ICI regimens.

Research Significance and Prospects

These findings provide a new pharmacodynamic monitoring tool for drug development, enabling early identification of potential responders and thereby optimizing clinical trial design and patient stratification strategies. Non-invasive monitoring via peripheral blood transcriptomics can help reduce unnecessary exposure to long-term immune-related adverse events (irAEs) and lower treatment costs. Moreover, the model suggests heterogeneous mechanisms of immunotherapy response across breast cancer subtypes: TNBC relies on T-cell activation, whereas HR+ subtypes may depend more on alleviating myeloid-mediated immune suppression.

 

 

Conclusion

This study establishes peripheral blood transcriptomics as a powerful tool for predicting outcomes of neoadjuvant chemoimmunotherapy in breast cancer. The multiparameter immune scoring model not only effectively distinguishes responders within the internal cohort but is also validated in an independent dostarlimab cohort, demonstrating strong generalizability. These results provide robust evidence for a paradigm shift from 'tissue-centric' to 'systemic immune monitoring,' offering significant translational value in clinical settings where repeated tumor sampling is challenging. Future studies could further integrate single-cell transcriptomics and TCR sequencing to resolve the functional states of key immune subsets at higher resolution. Additionally, this model could be incorporated into prospective clinical trials to guide personalized treatment decisions, reduce exposure to ineffective immunotherapies, and improve treatment safety and cost-effectiveness. Ultimately, this blood-based dynamic biomarker strategy may be extended to other solid tumors, advancing the field of precision immuno-oncology.

 

Xiaopeng Sun, Andres A Ocampo, Ann Hanna, Yaomin Xu, and Justin M Balko. Peripheral blood transcriptional profiling predicts tumor subtype and neoadjuvant chemoimmunotherapy outcomes in human breast cancer. Science translational medicine.