Novel MicroRNA-Based Risk Score Identified by Integrated Analyses to Predict Metastasis and Poor Prognosis in Breast Cancer
- 327 Downloads
The use of biomarkers that allow early therapeutic intervention or intensive follow-up evaluation is expected to be a powerful means for reducing breast cancer mortality. MicroRNAs (miRNAs) are known to play major roles in cancer biology including metastasis. This study aimed to develop a novel miRNA risk score to predict patient survival and metastasis in breast cancer.
An integrated unbiased approach was applied to derive a composite risk score for prognosis based on miRNA expression in primary breast tumors in 1051 breast cancer patients from The Cancer Genome Atlas (TCGA). Further analysis of the risk score with metastasis/recurrence was performed using the TCGA data set and validated in a separate patient population using small RNA sequencing.
The three-miRNAs risk score (miR-19a, miR-93, and miR-106a) was developed using the TCGA cohort, which predicted poor prognosis (p = 0.0005) independently of known clinical risk factors. The prognostic value was validated in another three following independent cohorts: GSE19536 (p = 0.0009), GSE22220 (p = 0.0003), and the Molecular Taxonomy of Breast Cancer International Consortium (METABRIC) (p = 0.0023). The three-miRNAs risk score predicted bone recurrence in TCGA (p = 0.0052), and the findings were validated in another independent population of patients who experienced bone recurrence and age/stage-matched patients without any recurrence. The three-miRNAs risk score enriched multiple metastasis-related gene sets such as angiogenesis and epithelial mesenchymal transition in a gene-set-enrichment analysis.
The authors developed the novel miRNA-based risk score, which is a promising biomarker for prediction of worse survival and bone recurrence potential in breast cancer.
Kazuaki Takabe is supported by NIH/NCI Grant R01CA160688, Susan G. Komen Investigator Initiated Research Grant IIR12222224, and Institutional Grant 71-4085-01. The authors acknowledge Bioinformatics and Biostatistics Shared Resources supported by National Cancer Institute (NCI) Grant P30CA016056, as well as Pathology Resource Network and Clinical Data Delivery and Honest Broker services provided by the Clinical Data Network, which is funded by the National Cancer Institute and Roswell Park Cancer Institute Cancer Center Support Grant shared resource.
Conflict of interest
There are no conflicts of interest.
- 4.SEER Stat Fact Sheets: Female Breast Cancer, 2016. http://seer.cancer.gov/statfacts/html/breast.html. Accessed 20 Dec 2016.
- 20.Sobin LH, Gospodarowicz MK, Wittekind C. TNM classification of malignant tumours. 7th ed. New York: John Wiley & Sons; 2009.Google Scholar