Yubo Jing Xinzhu Huang Yiyang Wang Junyi Wang Yongxiang Li Dlraba Yelihamu Chenming Guo ^*

• First Affiliated Hospital of Xinjiang Medical University, Urumqi, China

Background: Breast cancer (BC) is the prevailing malignant tumor, with its prevalence and death rate steadily rising over time. BC often does not show obvious symptoms in its early stages and is difficult to distinguish from benign breast disease. We aimed to find a distinct group of miRNAs utilizing serum as a non-invasive biomarker for early BC diagnosis.Methods: Herein, we mainly include the screening stage, testing stage, and verification stage. In the screening stage, 8 miRNAs associated with BC were selected and analyzed via literature reading, and the expression of the above miRNAs in BC was further verified by bioinformatics and included in the research analysis. In the testing phase, quantitative reverse transcription polymerase chain reaction (qRT-PCR) was deployed to select the five miRNAs with the most significant expression differences in 15 BC patients and 15 benign breast controls to proceed to the next stage. In a subsequent validation phase, the five miRNAs obtained from serum samples from an additional 75 BC patients and 50 benign control patients were evaluated using RT-qPCR.The diagnostic capacity, specificity, and sensitivity of candidate miRNAs were estimated with the receiver operating characteristic (ROC) curve and area under the curve (AUC). Finally, the optimal diagnostic combination model with high sensitivity and strong specificity was constructed by using the above 5 miRNAs.The BC patients reported a significant decline in mir-10b-5p, mir-133a-3p, mir-195-5p, and mir-155-3p levels in serum levels contrasted with those in benign controls. Additionally, BC patients experienced elevated mir-195-3p levels than in benign controls. We implemented ROC analysis to evaluate its diagnostic capacity for BC. We demonstrated that all five miRNAs had robust diagnostic capability, with an AUC above 0.8. We developed a conclusive diagnostic combination model consisting of these 5 miRNAs in order to enhance the diagnosis accuracy. This model demonstrated a high diagnostic value, as shown by an AUC of 0.948.The serum biomarker panels composed of five miRNAs identified in this study (mir-10b-5p, mir-133a-3p, mir-195-5p, mir-195-3p, and mir-155-3p) provide hope for early, non-invasive, and accurate diagnosis of BC.