A Potential Prognostic Long Noncoding RNA Signature to Predict Recurrence among ER-positive Breast Cancer Patients Treated with Tamoxifen

Abstract

Limited predictable long noncoding RNA (lncRNA) signature was reported in tamoxifen resistance among estrogen receptor (ER)-positive breast cancer (BC) patients. The aim of this study was to identify and assess prognostic lncRNA signature to predict recurrence among ER-positive BC patients treated with tamoxifen. Cohorts from Gene Expression Omnibus (GEO) (n = 298) and The Cancer Genome Atlas (TCGA) (n = 160) were defined as training and validation cohort, respectively. BC relapse associated lnRNAs was identify within training cohort, and the predictable value of recurrence was assessed in both cohorts. A total of 11lncRNAs were recognized to be associated with relapse free survival (RFS) of ER-positive BC patients receiving tamoxifen, who were divided into low-risk and high-risk group on basis of relapse risk scores (RRS). Multivariate cox regression analyses revealed that the RRS is an independent prognostic biomarker in the prediction of ER-positive BC patients' survival. GSEA indicated that high-risk group was associated with several signaling pathways in processing of BC recurrence and metastasis such as PI3K-Akt and Wnt signaling. Our 11-lncRNA based classifier is a reliable prognostic and predictive tool for disease relapse in BC patients receiving tamoxifen.

Figure 1

11-LncRNA signature biomarker characteristics in the GSE17705 cohort (N = 298). (A) 11-lncRNA expression and risk score distribution by z-score. The risk scores for all patients in GSE17705 cohort are plotted in ascending order and marked as low risk (black) or high risk (red), as divided by the threshold (vertical black line). (B) Patients’ recurrence status and time; (C) heatmap of the lncRNA expression profiles. Rows represent lncRNAs, and columns represent patients. Red indicated higher expression and black indicated lower expression.

Figure 2

ROC cures and Kaplan-Meier estimates of 3-year and 5-year RFS. ROC curves showed sensitivity and specificity of 3-year and 5-year RFS prediction by the 11-lncRNA risk score within (A) GSE17705 cohort (n = 298) and (B) TCGA cohort (n = 160). Kaplan-Meier plots were used to visualize the survival probabilities for the low-risk versus high-risk group of patients determined on the basis of the third quartile risk score from the training-set patients within (C) GSE17705 cohort (n = 298) and (D) TCGA cohort (n = 160) (all P < 0.0001).

Figure 3

11-LncRNA signature biomarker characteristics in the TCGA cohort (N = 160). (A) 11-lncRNA expression and risk score distribution by z-score. The risk scores for all patients in TCGA cohort are plotted in ascending order and marked as low risk (black) or high risk (red), as divided by the threshold (vertical black line). (B) Patients’ recurrence status and time; (C) heatmap of the lncRNA expression profiles. Rows represent lncRNAs, and columns represent patients. Red indicated higher expression and black indicated lower expression.

Figure 4

Nomogram, ROC curves and the internal calibration for predicting 3-year and 5-year RFS within TCGA cohort (n = 160). The nomogram was built using 11-lncRNA risk score, age, positive lymph nodes ratio, AJCC stage, and chemotherapy. (A) Nomogram for predicting proportion of patients with 3-year and 5-year RFS. (B) ROC curve showed sensitivity and specificity of the nomogram. (C) Plots depict the calibration of model in terms of agreements between predicted and observed 3-year and 5-year RFS. Model performance is shown by the plot, relative to the 45-degree line, which represents perfect prediction.

Figure 5

(A) Gene Set Enrichment Analysis Delineates Biological Pathways and Processes associated with risk score within GSE17705 cohort. (A) Volcano Plot with t-statistics. Circles show potentially differential genes between low-risk and high-risk group based on 11-lncRNA risk scores. The green and red circle were up and down genes, respectively. (B) Cytoscape and Enrichment Map were used for visualization of the GSEA results. Nodes represent enriched gene sets, which are grouped and annotated by their similarity according to related gene sets. Enrichment results were mapped as a network of gene sets (nodes). Node size is proportional to the total number of genes within each gene set. Proportion of shared genes between gene sets is represented as the thickness of the green line between nodes.