Abstract 148P
Background
The abnormal expression of microRNAs is a key hallmark of breast cancer (BC). Based on the miRNA expression information of BC in TCGA database, we aimed to screen and identify new prognostic markers for BC.
Methods
The differential expression of miRNAs was identified between tumor and normal samples in TCGA database by Morpheus software. Using Cox proportional hazard regression model, a miRNA-based prognosis model was established. Kaplan-Meier survival analysis, Log-rank test, receiver operating characteristic (ROC) curves and multivariate Cox regression analysis were used to test the prognosis model of BC.
Results
Through the differential expression analysis of miRNAs in 103 paired BC and adjacent normal tissues in TCGA database, we screened 100 most significantly altered miRNAs, including 50 up-regulated miRNAs and 50 down-regulated miRNAs. Then we randomly divided 839 breast cancer patients with complete clinicopathological characteristics into training set and testing set groups. By using Cox proportional risk regression model, a three-miRNA signature (miR-1247-5p, miR-29a-3p, and miR-148b-3p) was established in the training set, and validated in the testing set and the entity set. The results showed that the prognosis of low-risk group was better than that of high-risk group, and the three-miRNA signature had high diagnositic sensitivity and specificity. Multivariate Cox regression analysis showed that the three-miRNA signature was an independent prognostic factor for BC patients.
Conclusions
This study revealed a three-miRNA signature as a promising prognostic marker for breast cancer.
Clinical trial identification
Editorial acknowledgement
Legal entity responsible for the study
Beilei Zeng.
Funding
Has not received any funding.
Disclosure
All authors have declared no conflicts of interest.