Session Information
Session Type: Abstract Submissions (ACR)
Background/Purpose:
A strong genetic contribution to knee osteoarthritis (OA) is widely recognized although few loci have been robustly associated with knee OA susceptibility. To identify genes associated with knee OA, we performed a meta-analysis of genome wide association (GWA) results from two large studies of knee OA: the Genetic Components of Knee OA (GeCKO) Study, an ancillary study from the Osteoarthritis Initiative (OAI), and the Johnston County (JoCo) Osteoarthritis Project. In addition, we attempted replication of 18 single nucleotide polymorphisms (SNPs) previously reported in the literature to be associated with hip or knee OA by the TREAT-OA Consortium, arcOGEN Study and others.
Methods: Cases with knee OA were Caucasians who had at least one knee with definite radiographic OA (Kellgren-Lawrence [KL] grade >=2) at any available visit. Controls were Caucasians who were free of definite radiographic OA in both knees (KL grade <=1) at all available study visits. In total, data from 2014 cases and 953 controls from the OAI and 653 cases and 823 controls from the JoCo Study were included in the analysis. The OAI and JoCo DNA samples were genotyped using the Illumina 2.5M and Illumina 1M genotyping chip, respectively. In both studies imputation was performed using the 1000 genomes reference panel (June 2011 release) and statistical analysis was performed assuming an additive genetic model and adjusting for participant age and sex. Fixed-effects meta-analysis was conducted using METAL.
Results:
Over 5.8 million SNP variants present in both the OAI and JoCo datasets and having both high quality 1000 genomes imputation (r2>0.3) and minor allele frequency of at least 5% were included in the meta-analysis. Ten variants had a meta-analysis P-value < 1x10-6 with Odds Ratios ranging from 1.27-1.54. The 10 variants were located in three loci: one on chromosome 11p15.4 upstream of the TRIM21 gene (encodes for tripartite motif-containing protein 21, an intracellular antibody effector in the proteolysis pathway), and two in non-genic regions of chromosome 11p15.3 and chromosome 2q35; the closest genes to the latter region being MREG and FN1 (FN1 encodes for fibronectin, a glycoprotein that binds to chondrocytes and is involved in cell adhesion and migration). Similar to previously reported findings (Osteoarthritis Cart 2012;20[Suppl 1]:S46), only the rs143383 SNP in GDF5 (Growth and Differentiation Factor 5) was modestly associated with knee OA (P = 0.006) in the current analysis.
Conclusion:
These data support the polygenic nature of the genetic contribution to knee OA. Further analyses will consist of both de novo and in silico replication in other datasets and populations.
Disclosure:
M. C. Hochberg,
Abbott Laboratories, Astra-Zeneca, Bioiberica S.A., Eli Lilly Inc., Genentech/Roche, Merck Inc., Novartis Pharma A.G., Pfizer Inc., Stryker LLC, Xoma.,
5;
L. Yerges-Armstrong,
None;
C. Lu,
None;
M. S. Yau,
None;
B. D. Mitchell,
None;
J. M. Jordan,
Algynomics, Inc. ,
1,
Johnson and Johnson,
5,
Johnson & Johnson,
2,
Interleukin Genetics, Inc. ,
5,
Eli Lilly and Company,
5,
Mutual Pharmaceutical Company,
5;
Y. Liu,
None;
J. B. Renner,
None;
T. McSherry,
TGen,
3;
D. M. Taverna,
TGen,
3;
D. Duggan,
TGen,
3;
W. J. Mysiw,
None;
R. D. Jackson,
None.
« Back to 2012 ACR/ARHP Annual Meeting
ACR Meeting Abstracts - https://acrabstracts.org/abstract/genome-wide-association-studies-of-knee-osteoarthritis-in-2-large-north-american-cohorts-a-meta-analysis-with-2667-cases/