Genome Instability-Derived Genes Are Novel Prognostic Biomarkers for Triple-Negative Breast Cancer

Abstract

Background: Triple-negative breast cancer (TNBC) is an aggressive disease. Recent studies have identified genome instability-derived genes for patient outcomes. However, most of the studies mainly focused on only one or a few genome instability-related genes. Prognostic potential and clinical significance of genome instability-associated genes in TNBC have not been well explored.

Methods: In this study, we developed a computational approach to identify TNBC prognostic signature. It consisted of (1) using somatic mutations and copy number variations (CNVs) in TNBC to build a binary matrix and identifying the top and bottom 25% mutated samples, (2) comparing the gene expression between the top and bottom 25% samples to identify genome instability-related genes, and (3) performing univariate Cox proportional hazards regression analysis to identify survival-associated gene signature, and Kaplan-Meier, log-rank test, and multivariate Cox regression analyses to obtain overall survival (OS) information for TNBC outcome prediction.

Results: From the identified 111 genome instability-related genes, we extracted a genome instability-derived gene signature (GIGenSig) of 11 genes. Through survival analysis, we were able to classify TNBC cases into high- and low-risk groups by the signature in the training dataset (log-rank test p = 2.66e-04), validated its prognostic performance in the testing (log-rank test p = 2.45e-02) and Molecular Taxonomy of Breast Cancer International Consortium (METABRIC) (log-rank test p = 2.57e-05) datasets, and further validated the predictive power of the signature in five independent datasets.

Conclusion: The identified novel signature provides a better understanding of genome instability in TNBC and can be applied as prognostic markers for clinical TNBC management.

Keywords: TNBC; copy number variation; genome instability; mutation; prognosis.

FIGURE 1

Identification of genome instability-related genes in triple-negative breast cancer (TNBC). (A) Hierarchical clustering of all the 299 Molecular Taxonomy of Breast Cancer International Consortium (METABRIC) TNBC cases using the expression of 111-genomic instability-related genes. The patients were divided into genomic unstable (GU) and genomic stable (GS) groups. (B) Boxplots of alteration count and FOXM1 expression in GU and GS groups. The alteration counts and FOXM1 expression in GU group were significantly higher than that in GS group. (C) The top 10 Gene Ontology Biological Process (GOBP) terms of functional enrichment results. (D) The top 10 Gene Ontology Cellular Component (GOCC) terms of functional enrichment results.

FIGURE 2

Identification of genome instability-derived gene signature (GIGenSig) for prognostic prediction. (A) Survival curve of overall survival of TNBC patients in the training dataset. Patients were significantly classified into high- and low-risk groups; (B) 5-year receiver operating characteristic (ROC) curve for the GIGenSig in the training dataset; (C) GIGenSig gene expression pattern and alteration distribution and FOXM1 expression level with the increasing overall risk score (ORS) scores for the patients in the training dataset. The blue and red represent the low- and high-risk groups, respectively; (D) distribution of accumulative alteration number and FOXM1 expression in the high- and low-risk groups in the training dataset. The blue and red represent the low- and high-risk groups, respectively.

FIGURE 3

Validation of GIGenSig for prognostic prediction in the testing and METABRIC datasets. (A,B) Survival curves of overall survival of patients in the testing and METABRIC datasets. Patients were significantly classified into high- and low-risk groups. (C,D) The GIGenSig gene expression pattern, alteration distribution, and FOXM1 expression level with the increasing ORS scores for the patients in the testing and METABRIC dataset. The blue and red represent the low- and high-risk groups, respectively. (E,F) The distribution of accumulative alteration number and FOXM1 expression in the high- and low-risk groups in the testing and METABRIC datasets. The blue and red represent the low- and high-risk groups, respectively.

FIGURE 4

Validation of GIGenSig in five additional datasets. (A,B) Boxplots for ORS for TNBC patients with different stage and grade in The Cancer Genome Atlas (TCGA) and Shanghai dataset. (C–E) Boxplots for TFF3 expression among patients with different grade in GSE21653, GSE31448, and GSE25066. (F–H) Boxplots for FOXM1 expression among patients with high and low TFF3 expression in GSE21653, GSE31448, and GSE25066. The comparisons between any two different groups were performed by Wilcox test.

FIGURE 5

Prognostic prediction by GIGenSig is independent of clinical features. (A,B) Survival curve of overall survival (OS) for patients with age <55 and ≥55 in the METABRIC datasets. Patients were significantly classified into high- and low-risk groups. (C,D) Survival curve of OS for patients with premenopausal and postmenopausal status in the METABRIC datasets. Patients were classified into high- and low-risk groups; (E,F) Survival curve of OS for patients with stages I and II and stages III and IV in the METABRIC datasets. Patients were classified into high- and low-risk groups.

FIGURE 6

Better performance of GIGenSig than other prognostic signatures. Five-year ROC comparison of overall survival between GIGenSig signature and other four signatures from AUC (area under the curve), ACC (accuracy), SPE (specificity), and SEN (sensitivity).