Abstract
BACKGROUND: Bone metastasis is a significant contributor to mortality in patients with advanced breast cancer. Its progression is deeply intertwined with tumor metabolic reprogramming and the remodeling of the immune microenvironment. However, the dynamic interplay between metabolic pathways and immune regulation remains incompletely elucidated. METHODS: In this study, leveraging RNA-seq data and clinical information from breast cancer bone metastasis (BCBM) patients sourced from the GEO database, integrated bioinformatic analyses were employed to determine the activity of metabolic pathways significantly associated with prognosis. Metabolism-related genes were identified, and different metabolism-related gene clusters (MRGs) were subsequently identified by unsupervised clustering. Furthermore, a risk model was constructed based on hub prognostic genes, and differences in immune cell infiltration and drug sensitivity were compared between different subgroups. Finally, through single-cell RNA sequencing (scRNA-seq) analysis, we elucidated cellular heterogeneity and cell-cell communication within the tumor microenvironment (TME). RESULTS: This study identified three metabolic pathways (Amino Acid, Cofactor/Vitamin, and Secondary Metabolite) significantly associated with patient prognosis. Two metabolic pathway-related subtypes (C1 and C2) were defined, which exhibited differing prognostic outcomes. Concurrently, MRG1-3 were also identified, and there were significant differences in prognosis and immune infiltration levels between the three clusters, with MRG2 having a significantly better prognosis than MRG1 and MRG3. In addition, metabolism-related risk models based on risk scores were developed. The risk model had strong prognostic predictive ability. Subsequently, scRNA-seq analysis revealed that ALDH1A1 and macrophages may play a key role in BCBM. CONCLUSION: This study reveals the prognostic metabolic pathways and important prognostic target genes in BCBM from the perspective of metabolism-immunity interaction. MRGs can well distinguish the prognosis of different patients, and metabolism-related risk modeling can be used as a good prognostic predictor, which provides valuable insights into the "metabolic-immune" perspective of treatment.