Downregulated circulating microRNAs after surgery: potential noninvasive biomarkers for diagnosis and prognosis of early breast cancer

Abstract

Success in curing breast cancer largely depends on the stage at diagnosis. Circulating microRNAs are becoming a promising noninvasive biomarker. We postulate that a postoperative decline in circulating microRNAs might have diagnostic and prognostic value. Applying high-throughput microarrays, we screened the dysregulated microRNAs in paired serum samples before and after surgery. The relative concentrations of putative markers between the early breast cancer and cancer-free groups were evaluated in the training set and verified in the validation set. Sensitivity, specificity, and receiver operating characteristic (ROC) curves were used to assess diagnostic value. Survival analysis was performed using Kaplan-Meier estimates and a Cox proportional hazards model. Five microRNAs significantly reduced after surgery were selected for the training set. We found that miR-130b-5p, miR-151a-5p, miR-206, and miR-222-3p were significantly higher in the breast cancer group. Each of the four microRNAs had potential diagnostic value. The combined four microRNAs (training set: area under the curve (AUC) 0.8457; validation set: AUC 0.9309) had better diagnostic value than each single microRNA. MiR-222-3p was an independent prognostic factor for disease-free survival (HR = 13.19; 95% CI, 1.06-163.59; P = 0.045). Patients with no fewer than three highly expressed miRNAs had shorter DFS than patients with 0-2 highly expressed miRNAs (HR = 2.293; 95% CI, 1.128-0.662; P = 0.022). Our findings indicate that postoperatively downregulated circulating miR-130b-5p, miR-151a-5p, miR-206, and miR-222-3p may be potential biomarkers for breast cancer diagnosis and prognosis.

Fig. 1. Heat map and hierarchical clustering of downregulated miRNA expression signatures compared postoperative with preoperative signatures in nine matched-pair breast cancer patients’ samples.

Each row represents a miRNA, and each column represents a sample. The miRNA clustering tree is shown on the left. The color scale shown on the top illustrates the relative expression level of a miRNA in a certain slide: red color represents a high relative expression level and green color represents a low relative expression level. Those miRNAs whose fold change is more than 1.5 and P value less than 0.05 are listed here

Fig. 2. Receiver-operating characteristic (ROC) curve analysis of the four miRNAs for detecting breast cancer.

(A) ROC curve analysis of combined miR-130b-5p,miR-151a-5p,miR-206 and miR-222-3p in the training set (AUC 0.8457,sensitivity 85.00%, specificity 65.22%). (B) ROC curve analysis of combined miR-130b-5p,miR-151a-5p,miR-206 and miR-222-3p in the validation set (AUC 0.9309,sensitivity 84.31%, specificity 83.33%).

Fig. 3. The Kaplan-Meier survival curve for disease-free survival in the breast cancer patients.

(A) Analysis by the expression of miR-222-3p. (B) Analysis by the number of highly expressed miRNAs.