A Prognostic Model of Triple-Negative Breast Cancer Based on miR-27b-3p and Node Status

OBJECTIVE Triple-negative breast cancer (TNBC) is an aggressive but heterogeneous subtype of breast cancer. This study aimed to identify and validate a prognostic signature for TNBC patients to improve prognostic capability and to guide individualized treatment. METHODS We retrospectively analyzed the prognostic performance of clinicopathological characteristics and miRNAs in a training set of 58 patients with invasive ductal TNBC diagnosed between 2002 and 2012. A prediction model was developed based on independent clinicopathological and miRNA covariates. The prognostic value of the model was further validated in a separate set of 41 TNBC patients diagnosed between 2007 and 2008. RESULTS Only lymph node status was marginally significantly associated with poor prognosis of TNBC (P = 0.054), whereas other clinicopathological factors, including age, tumor size, histological grade, lymphovascular invasion, P53 status, Ki-67 index, and type of surgery, were not. The expression levels of miR-27b-3p, miR-107, and miR-103a-3p were significantly elevated in the metastatic group compared with the disease-free group (P value: 0.008, 0.005, and 0.050, respectively). The Cox proportional hazards regression analysis revealed that lymph node status and miR-27b-3p were independent predictors of poor prognosis (P value: 0.012 and 0.027, respectively). A logistic regression model was developed based on these two independent covariates, and the prognostic value of the model was subsequently confirmed in a separate validation set. The two different risk groups, which were stratified according to the model, showed significant differences in the rates of distant metastasis and breast cancer-related death not only in the training set (P value: 0.001 and 0.040, respectively) but also in the validation set (P value: 0.013 and 0.012, respectively). CONCLUSION This model based on miRNA and node status covariates may be used to stratify TNBC patients into different prognostic subgroups for potentially individualized therapy.