TY - JOUR T1 - Genome-wide association meta-analysis of knee and hip osteoarthritis uncovers genetic differences between patients treated with joint replacement and patients without joint replacement JF - Annals of the Rheumatic Diseases JO - Ann Rheum Dis SP - 384 LP - 392 DO - 10.1136/ard-2022-223199 VL - 82 IS - 3 AU - Cecilie Henkel AU - Unnur Styrkársdóttir AU - Gudmar Thorleifsson AU - Lilja Stefánsdóttir AU - Gyda Björnsdóttir AU - Karina Banasik AU - Søren Brunak AU - Christian Erikstrup AU - Khoa Manh Dinh AU - Thomas Folkmann Hansen AU - Kaspar René Nielsen AU - Mie Topholm Bruun AU - Joseph Dowsett AU - Thorsten Brodersen AU - DBDS Genomic Consortium AU - Thorgeir E Thorgeirsson AU - Kirill Gromov AU - Mikael Ploug Boesen AU - Henrik Ullum AU - Sisse Rye Ostrowski AU - Ole Birger Pedersen AU - Kári Stefánsson AU - Anders Troelsen Y1 - 2023/03/01 UR - http://ard.bmj.com/content/82/3/384.abstract N2 - Objectives Osteoarthritis is a common and severe, multifactorial disease with a well-established genetic component. However, little is known about how genetics affect disease progression, and thereby the need for joint placement. Therefore, we aimed to investigate whether the genetic associations of knee and hip osteoarthritis differ between patients treated with joint replacement and patients without joint replacement.Methods We included knee and hip osteoarthritis cases along with healthy controls, altogether counting >700 000 individuals. The cases were divided into two groups based on joint replacement status (surgical vs non-surgical) and included in four genome-wide association meta-analyses: surgical knee osteoarthritis (N = 22 525), non-surgical knee osteoarthritis (N = 38 626), surgical hip osteoarthritis (N = 20 221) and non-surgical hip osteoarthritis (N = 17 847). In addition, we tested for genetic correlation between the osteoarthritis groups and the pain phenotypes intervertebral disc disorder, dorsalgia, fibromyalgia, migraine and joint pain.Results We identified 52 sequence variants associated with knee osteoarthritis (surgical: 17, non-surgical: 3) or hip osteoarthritis (surgical: 34, non-surgical: 1). For the surgical phenotypes, we identified 10 novel variants, including genes involved in autophagy (rs2447606 in ATG7) and mechanotransduction (rs202127176 in PIEZO1). One variant, rs13107325 in SLC39A8, associated more strongly with non-surgical knee osteoarthritis than surgical knee osteoarthritis. For all other variants, significance and effect sizes were higher for the surgical phenotypes. In contrast, genetic correlations with pain phenotypes tended to be stronger in the non-surgical groups.Conclusions Our results indicate differences in genetic associations between knee and hip osteoarthritis depending on joint replacement status.Data are available in a public, open access repository. Summary results from the meta-analyses will be available at https://www.decode.com/summarydata at the time of the publication. Due to data privacy concerns, individual level genotypic and/or phenotypic data cannot be shared. Access to individual level data can be obtained by applying to each data source individually, given that relevant approvals are acquired. ER -