OP0294 Fracture prediction using a genetic markers algorithm compared to frax in three european cohorts. (15th June 2017)

• Record Type: Journal Article
• Authors: Ferrari, S; Rizzoli, R; Chapurlat, R; Brandi, ML; Martínez, H; Herrero, M; Vergés, J; Artieda, M; Tejedor, D; Martínez, A; Blanch, J; Palacios, S
• Is Part Of: Annals of the rheumatic diseases. Volume 76(2017)Supplement 2
• Page Start: 178
• Page End: 178
• Publication Date: 2017-06-15
• Subjects: Rheumatism -- Periodicals; 616.723005
• DOI: 10.1136/annrheumdis-2017-eular.6316
• Languages: English
• ISSNs: 0003-4967

Abstract

Abstract : Background: Numerous genome-wide association studies (GWAS) and large meta-analyses have started to unravel the multiple gene polymorphisms associated with BMD and/or fragility fractures. However the clinical utility of these genetic markers for fracture prediction remains to be established. Objectives: To develop a DNA genotyping tool for predicting osteoporotic fractures in postmenopausal women. Methods: 768 SNPs previously associated with osteoporosis phenotypes were identified in silico through the NHGRI GWAS catalog and BoneKey Genetics website. They were genotyped on an Illumina GoldenGate assay in 1649 post-menopausal women aged 45+ yrs belonging to three osteoporotic fractures cohorts from Switzerland, Italy and France. SNPs potentially associated (p<0.10) with prevalent and incident clinical fragility fractures in one or more of the cohorts, or in the cohorts together, were then combined in a genetic risk score (GRS). GRS association with fragility fractures was tested by forward logistic regressions adjusting for age and FN BMD. The ability of GRS for fracture prediction was evaluated by the area under the ROC curve (AUC) in the three cohorts combined, as well separately (for internal replication). For comparison, fracture probabilities were computed using FRAX clinical risk factors (without BMD) without and with the addition of GRS. Results: The average prevalence of fragility fractures in the three cohorts was 25% (range 22 to 28%), of which half were major fractures (FRAX definition). After QC filtering, 632 SNPs in 1625 individuals were correctly genotyped, of which 73 were potentially associated with fractures in one or more cohorts. In single and multiple regression models, GRS was significantly associated with fractures (OR 1.09, CI 1.07–1.12, p<0.0001). The GRS AUC for fracture prediction was significant (0.65) and highly consistent among the three cohorts. GRS predicted major fractures as well as FRAX clinical risk factors without BMD (AUC 0.63 vs 0.58, p=0.08), and when combined with clinical FRAX, the AUC was significantly improved (0.67, p=0.0106). Conclusions: SNPs previously associated with osteoporosis phenotypes through large GWAS and meta-analyses can be replicated for association with fragility fractures in post-menopausal women from three European countries. Our results provide a proof-of-principle that a genetic risk score (GRS) based on these SNPs represents an independent risk factor for fractures and could be developed into a genetic algorithm to improve the prediction of fragility fractures, either alone or together with FRAX. Disclosure of Interest: None declared