Circulating microRNAs have been proposed as disease biomarkers that may aid in risk assessment, diagnosis, prognosis, and monitoring of treatment response. The perceived opportunity has loomed particularly large in neoplastic disease, where alterations in cancer cells are thought to be reflected in the extracellular space as affected cells release upregulated miRNAs or fail to release apparently downregulated species. Despite the promise of miRNA biomarkers, evaluation of the diagnostic specificity and reproducibility of reported markers suggests that realizing this promise remains a work in progress.

This review examines issues of diagnostic specificity and reproducibility that have afflicted circulating miRNA studies. Surveying the breast cancer literature as an example, few miRNAs are reported consistently. Furthermore, it is posited that the assumptions underlying models of direct contributions of diseased tissue to biofluid miRNA profiles may not hold. Suggestions for improving diagnostic specificity and reliability are provided.

To maximize the likelihood of return on investment as miRNAs continue to be evaluated as specific and clinically useful markers, a focus is needed on miRNAs found in specific carriers, such as extracellular vesicles. Alternative sampling techniques should be developed, and nonblood biofluids should be considered. Careful optimization and standardization of preanalytical and analytical methods is needed to ensure that future results, positive or negative, are reliable.