ABSTRACT
Background:
Lysosomes are known to have a significant impact on the development and
recurrence of
breast cancer. However, the
association between
lysosome-related
genes (LRGs) and
breast cancer remains unclear. This study aims to explore the potential
role of LRGs in predicting the
prognosis and
treatment response of
breast cancer.
Methods:
Breast cancer gene expression profile data and clinical information were downloaded from TCGA and GEO databases, and
prognosis-related LRGs were screened for
consensus clustering analysis. Lasso Cox
regression analysis was used to construct
risk features derived from LRGs, and immune
cell infiltration, immune
therapy response,
drug sensitivity, and clinical pathological feature differences were evaluated for different molecular subtypes and
risk groups. A
nomogram based on
risk features derived from LRGs was constructed and evaluated.
Results:
Our study identified 176 differentially expressed LRGs that are associated with
breast cancer prognosis. Based on these
genes, we divided
breast cancer into two molecular subtypes with significant prognostic differences. We also found significant differences in immune
cell infiltration between these subtypes. Furthermore, we constructed a prognostic
risk model consisting of 7 LRGs, which effectively divides
breast cancer patients into high-
risk and low-
risk groups.
Patients in the low-
risk group have better prognostic characteristics, respond better to
immunotherapy, and have lower
sensitivity to
chemotherapy drugs, indicating that the low-
risk group is more likely to benefit from
immunotherapy and
chemotherapy. Additionally, the
risk score based on LRGs is significantly correlated with immune
cell infiltration, including CD8
T cells and
macrophages. This
risk score model, along with age,
chemotherapy, clinical stage, and N stage, is an independent
prognostic factor for
breast cancer. Finally, the
nomogram composed of these factors has excellent performance in predicting overall
survival of
breast cancer.
Conclusions:
In conclusion, this study has constructed a novel LRG-derived
breast cancer risk feature, which performs well in prognostic prediction when combined with clinical pathological features.