Novel insights into triple-negative breast cancer heterogeneity, prognosis, and treatment response based on matrix stiffness: a combined single-Cell and transcriptome analysis.

Background: The heterogeneity of triple-negative breast cancer (TNBC) is closely related to its tumor microenvironment. Tumor matrix stiffness (MS) is a key physical factor regulating tumor progression. This study aimed to explore the impact of MS on the biological behaviors of TNBC and establish its clinical association with patient prognosis. Methods: In this study, by integrating the transcriptome data of TNBC patients from the TCGA and GEO databases and MS-related genes, initially, the cell MS score was calculated based on single-cell RNA-sequencing data to identify malignant cells. Signature genes were selected through differential expression screening, univariate/multivariate Cox regression, and machine-learning methods. An MS-based prognostic model was then constructed and validated in independent cohorts. Furthermore, a nomogram model was constructed to evaluate its prognostic prediction efficacy. Also, the role of the MS-score model in predicting the immune microenvironment and treatment response was assessed. Results: The single-cell atlas revealed that malignant epithelial cells with a high MS score had unique functional states and pathway characteristics. The prognostic model constructed based on MS-related genes could effectively predict the survival risk of patients in multiple cohorts. Multivariate analysis confirmed that the MS score was an independent prognostic factor. The nomogram constructed by incorporating clinical parameters demonstrated good calibration ability and clinical net benefit. At the treatment level, a high MS score was associated with immunotherapy resistance and decreased chemotherapy sensitivity. Conclusion: This study provides a new theoretical basis and translational tools for the prognostic assessment and precision treatment of TNBC.