A novel eight-gene Immune Cell-Associated Predictive Gene model to predict recurrence in triple-negative breast cancer

BACKGROUND: Tumour tissue contains not only tumour cells but also some stromal cells and immune cells. This is one composition of the immune microenvironment of the tumour and causes a significant effect on the prognostic factors and recurrence of malignant tumor. METHODS: In this research, single-cell RNA data from triple-negative breast cancers (TNBCs) were comprehensively analyzed and 1,527 marker genes expressed in immune cells were identified. Subsequently, RNA sequencing and clinical data from 360 patients in the Triple Negative Breast Cancer database at the Fudan University Shanghai Cancer Center (FUSCC) were divided into two groups in a 1:1 ratio, the training group and the validation group. An eight-gene Immune Cell-Associated Predictive Gene (ICAPG) model for predicting breast cancer (BC) recurrence was developed using mRNA data from the training group combined with immune cell marker genes. Based on this model, subjects were divided into two different risk level groups. The predictive power of the model was fully validated using the validation group and The Cancer Genome Atlas (TCGA) database. The localization and expression of these eight genes were then confirmed in a single-cell database. ssGSEA and CIBERSORT algorithms were used to characterize the differences in immune cell infiltration between the two different risk groups. RESULTS: The eight-gene ICAPG model was proven to be effective in the validation group. The low-risk group patients presented higher criterion of infiltration of CD8(+) T cells and higher levels of tumour-infiltrating lymphocytes (TILs). In addition, the relationship between predictive models and homologous recombination deficiency (HRD) was explored and it was revealed that subjects from the high-risk group tended to have higher HRD values. CONCLUSIONS: This research established a new predictive model on the basis of immune cell marker genes that might effectively predict relapse in TNBC patients.