Reliable miRNA biomarker quantification in clinical practice - are we there yet?

Blood-borne miRNAs serve as disease diagnostic biomarkers and await clinical validation. Here, we evaluated Cel-miR-39-3p and miRNA16-5p as calibrator for the quantification of 15 miRNAs linked to hepatic impairment. We added defined copy numbers of Cel-miR-39-3p to plasma of healthy controls (N = 5) and patient samples undergoing liver resection (N = 51). The miRNAs were isolated according to SOPs and quantified by RT-qPCR using the 2^-(ΔΔ-CT)-method. Although miRNA16-5p and the spike-in control behaved similar in qPCR assays (R² = 0.8591) the spike-in control suffered from high inter-patient variability (median 7.6-fold) and low recoveries (median 5.6%, 95% CI 1.5-11.8%). Adding Cel-miR-39-3p to blood samples prior to RNA-isolation improved the recoveries (median 105.7%; 95% CI 29.9-219.9%), yet the inter-patient variability remained high (median 7.2-fold). Alike, we observed significant variability in CT-values for miRNA16-5p (range 14.7-fold) thus rendering this internal, blood-borne reference gene unacceptable as comparator. Specifically, 10 out of 15 diagnostic miRNAs failed the criteria R² ≥ 0.8 even though we added a defined copy number of Cel-miR-39-3p. This suggests interference of the spike-in control with individual miRNAs in the assay. Our study highlights current limitations in the quantification of blood-borne miRNAs that is of particularly importance when used for disease diagnostic and therapeutic interventions.