Development and validation of a herb-related gene signature for prognosis prediction and therapeutic response assessment in breast cancer

Background: Breast cancer is a highly heterogeneous disease, and there is a continuing need for robust prognostic tools that can also inform therapeutic strategies. This study aimed to develop a novel prognostic signature for breast cancer by leveraging the multi-target philosophy of traditional Chinese medicine (TCM). Methods: We systematically analyzed 221 TCM prescriptions for breast cancer to identify potential herb-related targets. A 20-gene herb-related risk score (HRS) was constructed using least absolute shrinkage and selection operator (LASSO) Cox regression in The Cancer Genome Atlas (TCGA) breast cancer cohort (n=1,097). The model's prognostic performance was validated in three independent cohorts: Molecular Taxonomy of Breast Cancer International Consortium (METABRIC) (n=1,466), GSE20685 (n=327), and GSE10886 (n=226). We further investigated the association of the HRS with the tumor immune microenvironment and potential drug sensitivity using bioinformatics algorithms. Results: The HRS served as an independent prognostic factor for overall survival (OS) [TCGA: hazard ratio (HR) =2.03, 95% confidence interval (CI): 1.25-3.30, P=0.004; METABRIC: HR =1.82, 95% CI: 1.59-2.10, P