Development and validation of a circulating microRNA panel for the early detection of breast cancer

Abstract

Background: Mammography is widely used for breast cancer screening but suffers from a high false-positive rate. Here, we perform the largest comprehensive, multi-center study to date involving diverse ethnic groups, for the identification of circulating miRNAs for breast cancer screening.

Methods: This study had a discovery phase (n = 289) and two validation phases (n = 374 and n = 379). Quantitative PCR profiling of 324 miRNAs was performed on serum samples from breast cancer (all stages) and healthy subjects to identify miRNA biomarkers. Two-fold cross-validation was used for building and optimising breast cancer-associated miRNA panels. An optimal panel was validated in cohorts with Caucasian and Asian samples. Diagnostic ability was evaluated using area under the curve (AUC) analysis.

Results: The study identified and validated 30 miRNAs dysregulated in breast cancer. An optimised eight-miRNA panel showed consistent performance in all cohorts and was successfully validated with AUC, accuracy, sensitivity, and specificity of 0.915, 82.3%, 72.2% and 91.5%, respectively. The prediction model detected breast cancer in both Caucasian and Asian populations with AUCs ranging from 0.880 to 0.973, including pre-malignant lesions (stage 0; AUC of 0.831) and early-stage (stages I-II) cancers (AUC of 0.916).

Conclusions: Our panel can potentially be used for breast cancer screening, in conjunction with mammography.

Fig. 1. Biomarker discovery.

a Volcano plot for 324 miRNAs profiled in 183 breast cancer patients as compared to 106 healthy individuals in the discovery cohort. Eighty-six miRNAs with p-values of less than 0.01 and magnitudes of log2 (fold change) of more than 0.5 are highlighted in red. b Heat-map of 663 cancer and non-cancer samples clustered using the expression of 33 selected miRNA biomarkers identified in the Discovery cohort. The expression levels (copy/ml) of miRNAs are presented in log2 scale and standardised to zero mean. The colour scale represents the concentrations of miRNA. Hierarchical clustering was carried out for both dimensions (miRNAs: Y-axis, samples: X-axis) based on the Euclidean distance. c Correlation of log2 (fold change) for 33 selected miRNA biomarker candidates in the Discovery cohort and Validation 1 cohort. The miRNA biomarker candidates with p-values larger than 0.05 in the validation 1 cohort are highlighted in red.

Fig. 2. Optimisation of multi-miRNA biomarker panels.

a Boxplots of AUC of multi-miRNA biomarker panels (with 2–12 miRNAs), in both the training and test sets, calculated from 200 iterations of the two-fold cross-validation process. The boxplot presents the 25th, 50th, and 75th percentiles in panel AUC. b Median AUC for the training and test sets from 200 iterations of the two-fold cross-validation process for multi-miRNA panels with 2–12 miRNAs. ***p < 0.001 (Student’s t-test). c ROC curves for breast cancer prediction performance of the optimal eight-miRNA biomarker panel in the Discovery and Validation 1 cohorts. The point with the maximum classification accuracy is shown as the red box. The sensitivity and specificity values at the maximum accuracy point are also shown. The 95% CI for these values is shown in the brackets.

Fig. 3. Validation of optimal eight-miRNA biomarker panel.

a ROC curve for breast cancer prediction performance of the optimal eight-miRNA biomarker panel in the Validation 2 cohort. The point with the maximum classification accuracy is shown as the red box. The sensitivity and specificity values at the maximum accuracy point are also shown. The 95% CI for these values is shown in the brackets. b AUC of eight-miRNA biomarker panel in detecting breast cancer in the Validation 1 and Validation 2 cohorts separated by sample source. c ROC curves for performance of the optimal eight-miRNA biomarker panel in predicting early (stages 0, I and II) and late (stages III and IV) breast cancer in the Validation 2 cohort.

Fig. 4. Prediction algorithm based on the expression of eight-miRNA biomarker panel.

a Prediction algorithm scores of cancer and non-cancer samples calculated from the expression of eight-miRNA panel expression. b Prediction algorithm scores of non-cancer samples and cancer samples by tumour stage (0, I, II, III, IV, and unknown).