Differential expression profile of microRNAs associated with human breast cancer progression

Materials and methods MicroRNA expression profiles were generated by microarray analysis (Agilent) of 70 samples from primary breast tumors with different clinical staging. Data were analyzed using bioinformatics software (R, Cluster and Tree View), which allowed sample discrimination (metastatic vs. non metastatic tumors) by hierarchical clustering of breast cancer expression signatures. We performed a second analysis using the ROC curve method that allowed us to identify miRNAs with greater predictive potential of malignancy. Furthermore, by Venn diagram analysis, we were able to identify 8 miRNAs that were differentially expressed in all metastatic tumors of different clinical staging, compared to the primary non-metastatic tumors. SPSS 19 software was used to generate a Kaplan-Meier curve (p<0.05) to estimate the disease-specific survival (considering the specific event such as death by cancer), based on miRNA expression patterns.