Cindy Boer

182 | Chapter 4.1 Discussion Osteoarthritis is the leading cause of joint disability and pain worldwide, but to date no curative treatment options exist. By conducting a large-scale GWAS meta-analysis, we have identified 52 new signals implicated in osteoarthritis risk, bringing the total to 148. We have identified multiple risk variants that traverse osteoarthritis phenotypes, indicating partly shared underpinning mechanisms and pathways relating to bone and cartilage development regardless of joint affected. We have further identified clear dif- ferences between osteoarthritis affecting the weight-bearing and non weight-bearing joints. For the first time, we have been able to establish genetic links between the dis- ease and its main symptom, pain. By integrating fine-mapping, colocalization and causal inference analyses with transcriptomics and proteomics studies in primary disease tis- sue, we have identified likely effector genes that represent high-value targets for thera- peutics. These findings enhance our understanding of the genetic aetiology of disease, shed novel biological insights, and provide a stepping stone for translating genetic as- sociations into osteoarthritis drug development. Methods GWAS cohorts, phenotype definition and meta-analysis Genome-wide association analysis for osteoarthritis was performed across 21 cohorts ( Supplementary Table 1 ), for a total of 826,690 individuals (177,517 osteoarthritis patients). We defined 11 stratified osteoarthritis phenotypes: osteoarthritis at any site, osteoarthritis of the hip and/or knee, knee osteoarthritis, hip osteoarthritis, total joint replacement, total knee replacement, total hip replacement, hand osteoarthritis, finger osteoarthritis, thumb osteoarthritis and spine osteoarthritis ( Figure 1 , Supplemen- tary Table 1 , Supplementary Information ). Osteoarthritis was defined by either a) self-reported osteoarthritis, b) clinical diagnosed, c) ICD10 codes or b) radiographic as defined by the TREAT-OA consortium[78], depending on the data available in the cohort ( Supplementary Table 1 ). GWAS analysis were performed by each cohort, and adjusted for cohort specific covariates ( Supplementary Table 1 ). GWAS summary sta- tistics from all cohorts were collected and checked to contain all the data needed for the meta-analysis. The quality control (QC) was performed centrally by using EasyQC[79]. Briefly, missing data, mono-allelic SNVs, nonsensical values (P-value>1, infinite beta’s etc.) and duplicates were removed from the data. We excluded variants with poor impu- tation quality (R2<0.3), if the effective sample size was <20 and if the minor allele count was <6 . Allele coding was harmonized across cohorts (A/T/C/G or I/D). Allele frequen-