Premium
T‐cell receptors provide potential prognostic signatures for breast cancer
Author(s) -
Liu Jingjing,
Zhang Jin
Publication year - 2021
Publication title -
cell biology international
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.932
H-Index - 77
eISSN - 1095-8355
pISSN - 1065-6995
DOI - 10.1002/cbin.11562
Subject(s) - nomogram , univariate , proportional hazards model , oncology , lasso (programming language) , medicine , multivariate statistics , breast cancer , multivariate analysis , gene signature , cancer , gene , biology , gene expression , machine learning , computer science , biochemistry , world wide web
Abstract Although T‐cell receptors (TCRs) are related to the progression of breast cancer (BC), their prognostic values remain unclear. We downloaded the messenger RNA (mRNA) profiles and corresponding clinical information of 1413 BC patients from the Cancer Genome Atlas and Gene Expression Omnibus database, respectively. The different expression analysis of 104 TCRs in BC samples was performed, and the consensus clustering based on 104 TCRs was performed by using the K‐mean method of R language. Univariate cox regression analysis was used to screen TCRs significantly associated with the prognosis of BC, and LASSO Cox analysis was applied to optimize key TCRs. The risk score was calculated using the prognostic model constructed based on six optimal TCRs, and multivariate Cox regression analysis was used to determine whether it was an independent prognostic signature. Finally, the nomogram was constructed to predict the overall survival of BC patients. Six optimal TCRs (ZAP70, GRAP2, NFKBIE, IFNG, NFKBIA, and PAK5), which were favorable for the prognosis of BC patients, were screened. Risk score could reliably predict the prognosis of BC patients as an independent prognostic signature. In addition, when bringing into two independent prognostic signatures, age and risk score, the nomogram model could better predict the overall survival of BC patients. Our results suggested that the poor prognosis of BC patients with high risk might be due to an immunosuppressive microenvironment. In summary, a prognostic risk model based on six TCRs was established and could efficiently predict the prognosis of BC patients.