Ecotypes of triple-negative breast cancer in response to chemotherapy

Key Points:

• The ARTEMIS Trial (NCT02276443) enrolled patients with triple-negative breast cancer (TNBC), defined by specific receptor negativity or low positivity and HER2 negativity, with ethical approval and informed consent obtained; tumor biopsies were collected pre-treatment for multi-omics analyses including scRNA-seq and spatial transcriptomics.
• Single-cell RNA sequencing (scRNA-seq) and spatial transcriptomics (Visium, VisiumHD, Xenium platforms) were performed on fresh frozen tumor biopsies using standardized protocols with modifications optimized for breast tissue, enabling detailed cellular and spatial profiling.
• Comprehensive bioinformatics pipelines were developed for data preprocessing, cell type identification, aneuploidy detection, gene expression normalization, clustering, and integration, including canonical correlation analysis (CCA) to correct batch effects and sophisticated methods such as CopyKAT for copy number alteration inference.
• Cancer cell archetypes and transcriptional metaprograms were identified via non-negative matrix factorization (NMF) on pseudo-bulk and single-cell data, with marker genes defined and linked to chemotherapy response; tumor microenvironment (TME) cell states were characterized through subclustering and label transfer from scRNA-seq to spatial data.
• Predictive models for neoadjuvant chemotherapy (NAC) response were constructed using both cell-state frequencies and gene expression signatures, validated in external TNBC cohorts, and survival analyses were performed using Cox proportional-hazards models to assess prognostic significance.