Diagnostic Performance of a Panel of miRNAs (OsteomiR) for Osteoporosis in a Cohort of Postmenopausal Women

Abstract

A specific signature of 19 circulating miRNAs (osteomiRs) has been reported to be associated with fragility fractures due to postmenopausal osteoporosis. However, it is unknown whether osteoporotic fractures or low BMD phenotypes are independently contributing to changes in osteomiR serum levels. The first aim was to characterize the abundance, sensitivity to hemolysis, and correlation of osteomiR serum levels, the second objective to evaluate the diagnostic accuracy of osteomiRs for osteoporosis according to the WHO criteria and on basis of major osteoporotic fracture history. Fifty postmenopausal women with osteoporosis (with or without fragility fracture) and 50 non-osteoporotic women were included in this cross-sectional study. The diagnostic performance of osteomiRs for osteoporosis based on the WHO definition or fracture history was evaluated using multiple logistic regression and receiver-operator curve (AUC) analysis. The osteomiR® signature is composed of four clusters of miRNAs providing good performance for the diagnosis of osteoporosis in postmenopausal women defined by WHO criteria (AUC = 0.830) and based on history of major osteoporotic fractures (AUC = 0.834). The classification performance for the WHO criteria and for fracture risk is driven by miR-375 and miR-203a, respectively. OsteomiRs, a signature of 19 emerging miRNA bone biomarkers, are measurable in human serum samples. They constitute a panel of independent bone and muscle biomarkers, which in combination could serve as diagnostic biomarkers for osteoporosis in postmenopausal women.

Keywords: Bone biomarker; Fragility fracture: circulating microRNA; Postmenopausal osteoporosis; osteomiR.

Fig. 1

Quality control assessment of osteomiR analysis by RT-qPCR. a Spike-in controls added to the RNA extraction (red), cDNA synthesis (purple), and PCR reaction (turquoise) indicate the analytical variability. b The ratio (delta Cq-value) between miR-23a/miR-451a is an established indicator of hemolysis

Fig. 2

Effect of hemolysis on osteomiR levels in human serum. a Heatmap illustrating the increase in osteomiR levels between non-hemolytic (red group) to low (blue), moderate (green), and severely (purple) hemolytic samples. b miR-451a and two other exemplary osteomiRs that increase in serum with increasing hemolysis. c All six osteomiRs that are not sensitive towards hemolysis. One-way ANOVA with Tukey post hoc tests. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001

Fig. 3

Abundance and correlation of osteomiR levels in serum of postmenopausal females. a Cq-values for 19 osteomiRs with respect to the limit of quantification (LoQ, Cq 37) and limit of detection (LoD, Cq 40) of the assay. b Unsupervised clustering of 19 osteomiRs based on Euclidean distance and ward.D2 clustering is shown as circular dendrogram. Four distinct clusters were identified and labeled 1–4

Fig. 4

Diagnostic performance of osteomiRs based on the WHO and fracture-based definitions of osteoporosis. a The probability score (p-score) obtained from a multiple logistic regression model including 19 osteomiRs is shown for the WHO definition. b Classification performance of the p-score for WHO definition based on ROC analysis. c and d show the same analyses for the fracture-based definition of osteoporosis