Background Osteoarthritis (OA) is the most common form of arthritis and a major socioeconomic burden. Preventive measures and early treatments are considered most effective in OA, but biomarkers to detect very early stages of OA or to predict the course of the disease are not available. Micro RNAs (miRNAs), which are stable and easily accessible in the peripheral circulation, are emerging as biomarkers for various physiological and pathological conditions.
Objectives To explore the association between serum miRNAs and the development of severe osteoarthritis of the knee and hip joint in the general population.
Methods Using the well-defined Bruneck cohort, we followed 816 Caucasian individuals from 1995 to 2010 and assessed joint arthroplasty as a definitive outcome of severe osteoarthritis of the knee and hip. A microarray screening identified candidate miRNAs in two OA serum pools, which we validated by qPCR in the entire cohort using U6 and Ct average as two independent methods for normalization.
Results During the 15-year follow-up, 67 of a total of the 816 individuals had one or more total joint replacement surgeries for severe knee or hip OA corresponding to an intervention rate of 6.9 per 1000 person-years. The body-mass index was significantly increased in subjects receiving arthroplasty for OA (P=0.002), and there was a trend towards an increased age among individuals with arthroplasty (P=0.053).
The microarray screening identified 12 candidate miRNAs, which we validated by qPCR in the entire cohort. After this validation, cox regression analysis demonstrated that 3 miRNAs were associated with severe knee and hip osteoarthritis. Let-7e was a negative predictor for total joint arthroplasty with an adjusted HR of 0.75 (CI: 0.58 to 0.96; P=0.021) when normalized to U6, and 0.76 (CI: 0.6 to 0.97; P=0.026) after normalization to the Ct-average. miR-454 was inversely correlated with severe knee or hip osteoarthritis with an adjusted HR of 0.77 (CI: 0.61 to 0.97; P=0.028) when normalized to U6. This correlation was lost when data were normalized to Ct-average (P=0.118). Finally, miR-885-5p showed a trend towards a positive relationship with arthroplasty when normalized to U6 (HR of 1.24; CI: 0.95 to 1.62; P=0.107) or to Ct-average (HR of 1.30; CI: 0.99 to 1.70; P=0.056).
MiRNA relevance network analysis further supported a putative role of miRNAs as OA biomarkers, since OA was characterized by substantial edge rewiring around let-7e as well as miR-454 and miR-885-5p.
Conclusions Our study is the first to identify differentially expressed circulating miRNAs in osteoarthritis patients necessitating arthroplasty in a large, population-based cohort. Among these miRNAs, let-7e emerged as potential predictor for severe knee or hip osteoarthritis.
Disclosure of Interest None declared