Circulating microRNAs in the early prediction of disease recurrence in primary breast cancer

Chara Papadaki, Michalis Stratigos, Georgios Markakis, Maria Spiliotaki, Georgios Mastrostamatis, Christoforos Nikolaou, Dimitrios Mavroudis, Sofia Agelaki

In primary breast cancer metastases frequently arise from a state of dormancy that may persist for extended periods of time. We investigated the efficacy of plasma micro-RNA (miR)-21, miR-23b, miR-190, miR-200b and miR-200c, related to dormancy and metastasis, to predict the outcome of patients with early breast cancer. miRNAs were evaluated by RT-qPCR in plasma obtained before adjuvant chemotherapy. miRNA expression, classified as high or low according to median values, correlated with relapse and survival. Receiver operating characteristic (ROC) curves were constructed to determine miRNA sensitivity and specificity. miR-21 (p < 0.001), miR-23b (p = 0.028) and miR-200c (p < 0.001) expression were higher and miR-190 was lower (p = 0.013) in relapsed (n = 49), compared to non-relapsed patients (n = 84). Interestingly, miR-190 was lower (p = 0.0032) in patients with early relapse (at < 3 years; n = 23) compared to those without early relapse (n = 110). On the other hand, miR-21 and miR-200c were higher (p = 0.015 and p < 0.001, respectively) in patients with late relapse (relapse at ≥ 5 years; n = 20) as compared to non-relapsed patients. High miR-200c was associated with shorter disease-free survival (DFS) (p = 0.005) and high miR-21 with both shorter DFS and overall survival (OS) (p < 0.001 and p = 0.033, respectively) compared to low expression. ROC curve analysis revealed that miR-21, miR-23b, miR-190 and miR-200c discriminated relapsed from non-relapsed patients. A combination of of miR-21, miR-23b and miR-190 showed higher sensitivity and specificity in ROC analyses compared to each miRNA alone; accuracy was further improved by adding lymph node infiltration and tumor grade to the panel of three miRs (AUC 0.873). Furthermore, the combination of miR-200c, lymph node infiltration, tumor grade and estrogen receptor predicted late relapse (AUC 0.890). Circulating miRNAs are differentially expressed among relapsed and non-relapsed patients with early breast cancer and predict recurrence many years before its clinical detection. Our results suggest that miRNAs represent potential circulating biomarkers in early breast cancer.