Abstract 2216: Construction of a prognostic 6-gene signature for breast cancer based on multi-omics data

Abstract

Abstract Breast cancer (BC) is a heterogeneous disease. Distinguishing BC patients with different prognosis can help to make clinical strategies. In our study, we tried to identify novel prognostic genetic biomarkers for BC by using multi-omics data and establish a risk score (RS) signature with genetic biomarkers which have high predictive value. A combined significance (pcombined) of each gene was calculated by Fisher’s method based on the p value of the difference examination of the respective index in RNA-seq, CNV, and DNA methylation data of corresponding genes from TCGA-BRCA dataset. Genes with a pcombined < 0.010 were subjected to univariate cox and Lasso regression to screen out hub genes, whereby a RS signature consisted of six genes, C15orf52, C1orf228, CEL, FUZ, PAK6, and SIRPG, was established. With the Kaplan-Meier (KM) method, the receiver operating characteristic curves (ROC), multivariate cox regression, Nomogram, and decision curve analysis (DCA), the predicted performance of the RS signature was assessed in TCGA-BRCA, GSE7390, and GSE20685. The predicted performance in TNBC patients was emphatically analyzed in TCGA-BRCA and GSE103091, while immune characteristics were also explored. Basing on the data from Genomics of Drug Sensitivity in Cancer (GDSC) database and related R package, we further examined the drug sensitivity of TNBC patients with different RS. With survival rates having significant differences in all KM analyses and most area under ROCs reaching 0.700, the RS signature could distinguish the prognosis of patients, even stratified by disease stages or subtypes. The RS also showed a stronger predictive ability than traditional clinical indicators, including age, tumor topology (T), regional lymph node (N), metastasis (M), and TNM stage. The down-regulated expressions of many immune checkpoints were observed in TNBC patients with higher RS, while the sensitivity of many antitumor drugs was reduced in those with higher RS, except for BI2536 (CAS No.: 755038-02-9), which demonstrated a smaller half maximal inhibitory concentration (which mean a higher drug sensitivity) in TNBC patients with higher RS. In conclusion, the six genes RS signature established based on multi-omics data exhibited well performance in predicting the prognosis of BC patients, regardless of disease stages or subtypes. Contributing to a more personalized treatment, the RS signature might benefit the outcome of BC patients. Citation Format: Zeyu Xing, Yuting Hong. Construction of a prognostic 6-gene signature for breast cancer based on multi-omics data [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 2216.