Serum miRNA-based prediction of axillary lymph node metastasis in breast cancer

Purpose: Sentinel lymph node biopsy (SLNB) is the gold-standard procedure for evaluating axillary lymph node (ALN) status in patients with breast cancer. However, the morbidity of SLNB is not negligible, and the procedure is invasive for patients without ALN metastasis. Here, we developed a diagnostic model for evaluating ALN status using a combination of serum microRNAs (miRNAs) and clinicopathological factors as a novel less-invasive biomarker. Experimental design: Preoperative serum samples were collected from patients who underwent surgery for primary breast cancer or breast benign diseases between 2008 and 2014. A total of 958 serum samples (921 cases of primary breast cancer, including 630 cases in the no ALN metastasis group and 291 cases in the ALN metastasis group, and 37 patients with benign breast diseases) were analyzed by miRNA microarray. Samples were randomly divided into training and test sets. Logistic LASSO regression analysis was used to construct diagnostic models in the training set, which were validated in the test set. Results: An optimal diagnostic model was identified using a combination of four miRNAs (miR-629-5p, miR-629-3p, miR-4710, and miR-4492) and three clinicopathological factors (T stage, lymphovascular invasion, and ultrasound findings), which showed a sensitivity of 0.876, a specificity of 0.712, an accuracy of 0.824, and an area under the receiver operating characteristic curve of 0.86 in the test set. Conclusions: Serum miRNA profiles may be useful for the diagnosis of ALN metastasis before surgery in a less-invasive manner than SLNB.

https://ift.tt/2r9pChR