Improved sensitivity for detection of breast cancer by combination of miR-34a and tumor markers CA 15-3 or CEA

Abstract

Background: MicroRNAs biomarkers have shown value for diagnosis and prognosis of various cancers. Combination with established tumor markers has rarely been done. Results: Breast cancer patients had significantly higher serum RNA loads (AUC 0.665), lower miR-34a (AUC 0.772), higher CEA and CA 15-3 levels (AUCs 0.717 and 0.721) than healthy controls. miR-34a correlated with tumor stage and hormone receptor status. There was no significant difference between groups for all other miRNAs. Combination of miR-34a with CEA or CA 15-3 led to improved AUCs of 0.844 and 0.800, respectively. Sensitivity of miR-34a and CA 15-3 reached 56.1% at 95% specificity. When compared with benign breast diseases, combination of miR-34a (AUC 0.719) and CEA (0.623) or CA 15-3 (0.619) resulted in improved performances (0.794 and 0.741). Sensitivity of miR-34a and CA 15-3 reached 53.7% at 95% specificity. Conclusion: While miR-34a provides valuable information for diagnosis and staging, combination with tumor markers CA15-3 or CEA improves the sensitivity for breast cancer detection. Patients and Methods: The diagnostic relevance of the miR-21, miR-34a, miR-92a, miR-155, miR-222 and miR-let-7c was tested in sera of 103 individuals (55 breast cancer, 20 benign breast diseases, 28 healthy controls). MiRNAs were detected by quantitative rt-PCR after extraction and reverse transcription. Cel-miR-39 and miR-16 were used for normalization. Established tumor markers CEA, CA 15-3, CA 19-9 and CA 125 were measured by automatized immunoassays. Diagnostic performance was tested by areas under the curve (AUC) of receiver operating characteristic (ROC) curves and sensitivities at 90% and 95% specificity.