Wenjie Zhu¹, Ying Fan², Mei Liu³, Ningzhi Xu³, Binghe Xu²

¹Department of Medical Oncology, Cancer Hospital and Institute, Chinese Academy of Medical Sciences and Peking union   Medical College, ²Department of Medical Oncology, Chinese Academy of Medical Sciences and Peking union   Medical College, ³Laboratory of Cell and Molecular Biology & State Key Laboratory of Molecular Oncology, Chinese Academy of Medical Sciences and Peking union   Medical College

Objective:Paradigm shifts to neoadjuvant chemotherapy (NACT) for locally advanced and operable breast cancer. Hormone receptor (HR)-positive and human epidermal growth factor receptor 2 (HER2)-negative breast cancer is less responsive to NACT compared to other intrinsic subtypes. Predictors of sensitivity to NACT need to be identified for this subgroup so as to save non-responders from chemotherapy-induced toxicities and potential risk of distant metastasis. Circulating miRNAs have been indicated as predictive biomarker in breast cancer. Method: Patients with primary HR-positive/HER2-negative breast cancer, who were enrolled in a randomized neoadjuvant clinical trial (ClincalTrials.gov NCT02041338), were included and divided into sensitive and insensitive groups based on their responses to NACT as assessed by RESIST v1.1 criteria. We prospectively collected blood samples before random assignment, after two cycles of chemotherapy and before definitive surgery. Baseline and preoperative miRNA profiling of responders and non-responders were compared to reveal candidate miRNAs. Dynamic evaluation of candidate miRNA level in serial plasma was performed by qRT-PCR in the validation cohort. Result: By miRNA array, notable inter-group difference was observed in the level of seven miRNAs at baseline (let-7e, let-7c, miR-20b,miR-26a, miR-222, miR-145 and miR-146a), which further expanded after completion of NACT. For another set of miRNAs (let-7b, miR-9, miR-184, miR-19b,miR-20a and miR-451), the levels before surgery were markedly different between two groups though similar at baseline. Of those, the variation trends of four candidate miRNAs were evaluated in the validation cohort (n=29). MiR-222 was highly expressed in the insensitive group initially and further upregulated with the accumulation of chemotherapy cycles. No significant difference in baseline expression was detected between groups but chemotherapy induced a progressive up- (miR-9 and miR-20a) or down-regulation (miR-451) in the resistant cases. Conclusion: High expression of miR-222 in baseline plasma was indicative of resistance to NACT. Decreased expression of miR-451 or elevated levels of miR-9 and miR-20a in plasma following administration of chemotherapy predicted poor response. We described for the first time that serial measurement of circulating miRNAs might help predict sensitivity to NACT in breast cancer. This inexpensive and non-invasive strategy for monitoring response would promisingly optimize the tailored management of patients with breast cancer.

Key Words: circulating miRNAs breast cancer  neoadjuvant chemotherapy