Image_4_A Novel Ferroptosis-Related Gene Signature for Overall Survival Prediction in Patients With Breast Cancer.TIF (1.05 MB)

Image_4_A Novel Ferroptosis-Related Gene Signature for Overall Survival Prediction in Patients With Breast Cancer.TIF

Download (1.05 MB)
posted on 2021-06-17, 05:11 authored by Lizhe Zhu, Qi Tian, Siyuan Jiang, Huan Gao, Shibo Yu, Yudong Zhou, Yu Yan, Yu Ren, Jianjun He, Bin Wang

Breast cancer is the most common malignant tumor in women worldwide. However, advanced multidisciplinary therapy cannot rescue the mortality of high-risk breast cancer metastasis. Ferroptosis is a newly discovered form of regulating cell death that related to cancer treatment, especially in eradicating aggressive malignancies that are resistant to traditional therapies. However, the prognostic value of ferroptosis-related gene in breast cancer remains unknown.

Materials and Methods

In this study, a total of 1,057 breast cancer RNA expression data with clinical and follow-up information were downloaded from the TCGA cohort, multivariate Cox regression was used to construct the 11-gene prognostic ferroptosis-related gene signature. The breast cancer patients from the GEO cohort were used for validation. The expression levels of core prognostic genes were also verified in erastin-treated breast cancer cell lines by real-time polymerase chain action (PCR).

Results and Discussion

Our results showed that 78% ferroptosis-related genes were differentially expressed between breast cancer tumor tissue and adjacent non-tumorous tissues, including 29 of them which were significantly correlated with OS in the univariate Cox regression analysis. Patients were divided into high-risk group and low-risk group by the 11-gene signature. Patients with high-risk scores had a higher probability of death earlier than the low-risk group both in the TCGA construction cohort and in the GEO validation cohort (all P < 0.001). Meanwhile, the risk score was proved to be an independent predictor for OS in both univariate and multivariate Cox regression analyses (HR > 1, P < 0.01). The predictive efficacy of the prognostic signature for OS was further verified by the time-dependent ROC curves. Moreover, we also enriched many immune-related biological processes in later functional analysis; the immune status showed a statistical difference between the two risk groups. In addition, the differences in expression levels of 11 core prognostic genes were examined in ferroptosis inducer-treated breast cancer cell lines.


In conclusion, a novel ferroptosis-related gene model can be used for prognostic prediction in breast cancer. New ferroptosis-related genes may be used for breast cancer targeting therapy in the future.